exploreknowledge-sharingstrategyandevolutionary...

Research ArticleExplore Knowledge-Sharing Strategy and EvolutionaryMechanism for Integrated Project Team Based on EvolutionaryGame Model

Yanchao Du 1 Hengyu Zhou 1 Yongbo Yuan 1 and Xiaoxue Liu 2

1Department of Construction Management Dalian University of Technology Linggong Road No 2 116024 Dalian China2School of Economics andManagement Tongji University Street Address Zhongtian Building 2005 Room No 1063 Siping RoadYangpu District Shanghai China

Correspondence should be addressed to Yongbo Yuan yongbodluteducn

Received 1 March 2019 Revised 22 May 2019 Accepted 11 June 2019 Published 27 June 2019

Academic Editor Bo Xia

Copyright copy 2019 Yanchao Du et al is is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Integrated Project Delivery (IPD) has become increasingly popular in the architecture engineering and construction industriesHowever the current practice status by the construction industry fails to deliver the desired results In that backdrop how topromote cooperation within and improve the overall performance of integrated project team has received wide attention Hereinknowledge-sharing plays a critical role in cooperation and overall performance However to the best of our knowledge theresearch on knowledge-sharing strategy interaction and evolutionary mechanism is rare To make up for the deficiency of thestudies existing a novel model is proposed by taking advantage of evolutionary game theory to capture the interaction behavior ofknowledge-sharing and explore its evolutionary mechanism Six parameters of knowledge stock knowledge-sharing degreeheterogeneous knowledge proportion synergy effect knowledge absorption coefficient and knowledge-sharing cost efficient thatare critical to knowledge-sharing are extracted and definede payoff matrix is constructed by analyzing the benefits and costs ofknowledge-sharing en a replicator dynamic system is established based on payoff matrix to determine the evolutionarytendency of knowledge-sharing behavior Finally numerical simulations are conducted to explore the influences of all parameterson the knowledge-sharing strategye findings in this research reveal that strategy interaction behavior is significantly influencedby proportion of strategy of choosing to share knowledge in both game players e authors also find that strategy interactionbehavior has a strong negative correlation with knowledge-sharing cost efficient but has a positive correlation with knowledgestock heterogeneous knowledge proportion degree of knowledge-sharing knowledge absorption coefficient and synergetic effectcoefficientis research can provide the evolutionarymechanism and broaden our understanding of relationship between projectperformance and knowledge-sharing and can offer valuable guidance on improving cooperation and performance ofproject teams

1 Introduction

Since the 1970s the global and the Chinese constructionindustry have been flourishing New project delivery ap-proaches such as design-bid-build (DBB) design-build(DB) and construction management at risk (CMR) havebeen constantly springing up [1] which are widely appliedand popularized in practice However productivity has notbeen improved dramatically and it is hard to satisfy thestakeholdersrsquo performance expectations [2] According to an

industry report published by Construction ManagementAssociation of America (CMAA) 30 of constructionprojects are over the schedule or exceed budget In fact non-value-added activities done by the traditional project teamassume most of the time in the construction process [3]Researchers recommend that a better collaborative andcoordinated project delivery method should be developed toovercome these drawbacks [4]

Integrated project delivery (IPD) has been recognized asa new delivery method integrating people systems business

HindawiAdvances in Civil EngineeringVolume 2019 Article ID 4365358 23 pageshttpsdoiorg10115520194365358

structures and practices into a single process whereinproject participants including the owner designer generalcontractor and special subcontractor can work collabora-tively at the early design stage IPD increases values to theowner decreases waste and maximizes project benefits byutilizing multiparty contract to share benefits and risksHowever according to the investigation carried by Kelly [5]the project performance that adopted the integrated projectdelivery approach did not achieve desired results in practiceOne of the significant reasons is that the cooperative be-havior among integrated project teammembers is hamperedto some extent thereby leading to an overall performancedecline e failure in performance improvements thusmakes it imperative for research to address this issue in orderto effectively facilitate better cooperation

In a knowledge-based economic society organizationalor a project team competitiveness arise from intangiblerather than tangible resources [6] In this economic contextknowledge is defined as the skills objectives and experiencesthat together create a framework for assessing and makinguse of information exchanged in an explicit or implicitmanner [7] Most researchers have argued that the com-petitiveness of an organization relies on the ability to createvaluable knowledge and share it with all project teammembers within this organization in order to infuse thisknowledge into products services and systems [8] us asuccessful knowledge-sharing strategy requires effectivecooperative relationships among integrated project teammembers is is critical for facilitating cooperation andproject performance [9] Without high-level knowledge-sharing a project team leader becomes unable to integrate allthe advantages that stakeholders offer (eg rich experiencestrong organization management skills and information-processing capability) for efficiently and effectively exe-cuting large and complex projects [10]

Cooperation is a common phenomenon in nature andhuman society It is called the third evolutionary principlebesides selection and mutation As an internal driving forceof species evolution and social development it promotes theformation of complex life systems and social organizationsFrom the perspective of individual survival and develop-ment cooperative behaviors will reduce their own benefitsand weaken their competitive advantages From the per-spective of society or team development cooperative be-havior is beneficial to the overall interest although it mayreduce individual interest In construction industry the bestpractice is to realize the overall performance and balanceindividual benefit In traditional project deliveries imple-mentation process is poorly integrated and information isasymmetric and the first consideration of each participant isnot to maximize the benefits of the overall project but tomaximize individual interest and avoid risks which usuallyleads to an overall project significant loss Hence it is es-sential to adopt the idea of IPD to solve this dilemma To thisend we should first capture the dynamic behavior of beingcooperative and make clear how cooperative behaviorspreads and maintains stability in an IPD team namelyevolutionary mechanism from the perspective of knowl-edge-sharing to provide theoretical foundation for

controlling and optimizing cooperative behavior in an IPDteam

In the construction industry the best practice is to re-alize the overall performance while balancing individualbenefits e implementation process in traditional projectdelivery is poorly integrated and crucial information isasymmetric Often each participant does not prioritizemaximization of the projectrsquos benefits Instead she or hechooses to maximize individual interest and avoid riskleading to significant project lossesese counterproductivebehaviors make it necessary to adopt IPD as a solution Tothis end we first capture the dynamic behavior of co-operation and clarify how cooperative behaviors spreadwithin and support the stability of an IPD team that is weidentify the evolutionary mechanism We then adopt theperspective of knowledge-sharing to provide a theoreticalfoundation for controlling and optimizing cooperative be-haviors in the IPD team

Despite the importance of knowledge-sharing in co-operative behaviors of the integrated project team very fewstudies have focused on improving cooperation from thisperspective to explain the dynamics of group behavior andhow cooperative behavior spreads and maintains stability inthe integrated project team namely evolutionary mecha-nism erefore our goal is to explore knowledge-sharingstrategies and reveal the evolutionary mechanism that un-derpins the integrated project teamWe analyze the resultingevolutionary tendency along with the variations in relevantfactors

Our study contributes to both theory and practice Firstwe propose an intrinsic evolutionary mechanism forknowledge-sharing strategies Second we provide valuablereferences for integrated project team leaders to broadentheir understanding of how cooperation is influenced bydifferent factors which would allow them to take corre-sponding measures to improve team performance

e remaining paper is organized as follows In Section 2we present the systematic overview of the extent literature onIPD approaches and knowledge-sharing In Section 3 weexplore the reasons for employing evolutionary game modelIn Section 4 we propose a novel evolutionary game model tocapture the dynamic behavior of knowledge-sharing InSection 5 we conduct numerical simulations to validate ourresults of theoretical analysis and discuss the influences ofmodel parameters on the evolutionary tendency of the in-teraction behavior In the last section we conclude the paperand present the implication of our work

2 Literature Review

21 Integrated Project Delivery Compared with other in-dustries the architecture engineering and construction(AEC) industry lacks a reliable reputation as a leading actorin quality productivity and time and cost managementbecause of its increasing complexity and multidisciplinarynature Most researchers have claimed that the projectsuccess largely depends on the complexity of a project whichthen directly influences the overall project performance [11]erefore it is important to recognize this feature Gidado

2 Advances in Civil Engineering

[12] analyzed project complexity from two aspects (1) themanagerial which is related to new workflow reengineeringand building information technology and (2) the techno-logical which is related to the technological complexity ofimplementing a single work e results showed that themanagerial style and information technology of traditionalproject delivery cannot satisfy the requirements of costcontrol quality improvement and schedule optimizationdue to complexity of construction project e constructionindustry calls for a renewed project delivery approach tofurther improve the overall project performance Giventhese requirements many scholars and practitioners havebegun to focus on the project complexity including themanagerial and technological is gave rise to IPD andbuilding information modeling (BIM)

IPD encourages key stakeholders to collaborate as anintegrated project team to maximize project benefits It issimilar in the function to the Japanese Toyota productionsystem [13] An IPD prototype was first applied in BP oildrilling platform achieving limited success through projectalliance [14] Soon afterward this approach was introducedinto the Wandoo Project and East Spar Project in Australiaand the Comprehensive Medical Project in California Inboth cases it boosted productivity and project performance[15] Since then IPD has received more attention fromexperts and practitioners in AEC industry In order topromote the application of IPD Consensus DOCS analliancing institute founded by 22 leading engineering as-sociations published a series of contracts to guide theimplementation of integrated project delivery IPD is dis-tinct from traditional delivery in the following ways [16] (1)all key stakeholder participating in the early design phase (2)all key stakeholders signing multiparty relationship agree-ments (3) a sharing mechanism for benefits and risks exists(4) there is collaborative decision-making and integratedtarget control and (5) all key stakeholders are subject toliability waivers e American Institute of Architects (AIA)also asserts some benefits if those principles mentionedabove can be executed by the key participants which includethe following

(i) Reduce or eliminate conflicts that often occur inthe traditional project team

(ii) Control risk effectively among key participants(iii) Facilitate cooperative relationships among in-

tegrated project team members(iv) Shorten the project development period from

design to handover(v) Decrease requests for interpretation sent by the

prime contractor and subcontractor(vi) Reduce waste by better and comprehensive

schedule(vii) Contribute to employing cutting-edge technology

such as BIM(viii) Enhance knowledge-sharing in the integrated

project team(ix) Reduce project management cost

(x) Improve project quality through better construc-tion management

Investing the Orlando Utilities Commission North Planproject Matthews and Howell [17] found 10 cost savingsbelow the $6 million GMP owing to the collaboration in theintegrated team Coincidentally a report published in the2006 AIA integrated practice conference emphasized theachievement of 40 Australian projects owing to a projectalliance (PA) that based on IPD principles e AIAcooperated with University of Minnesota to investigate 12large and complex IPD projects ey concluded that theprojects despite challenges in the implementation processstill satisfied the requirements of an aggressive schedule andbudget goals [18]

Despite the emergence of IPD as an advanced deliveryapproach the extent literature has reported some obstaclesin its practices including lack of specific tools for itsimplementation poor communication protocols in-formation barriers weak benefits distribution and in-sufficiently trust-based relationship is has prevented IPDfrom being employed on a large scale [19 20] AIA as aforerunner in IPD has acknowledged that these barriers andchallenges form the concept of IPD to field application Legaland financial issues including liability and risk-sharingwere identified active and stronger collaboration can beaccomplished through contractual and organizational Mostrecently to overcome IPD dependence on ldquobig roomrdquocollaboration requiring all stakeholders to be present all thetime Ma et al [21] set up a collaboration platform for IPDteam e results showed that the dedicated platform cansignificantly reduce the difficulty associated with IPDimplementation and thus prompt knowledge transfer be-tween project team members However it lacks analysis onhow to integrate scatted and heterogeneous informationfrom project team members and does not consider theinfluence of information sharing mechanism on the archi-tecture of platform IPD as a delivery method introducesBIM as a forceful technology tool With a growing trendtoward the application of building information model (BIM)technology in large and complex projects an increasingnumber of experts have paid great attention to investigatethe influence on project performance with the utilization ofBIM and IPD Using structural equation modeling (SEM) toexplore the questionnaire responses of more than 100 BIM-enabled projects Chang et al [22] demonstrated that theacceptability of IPD can be influenced positively All keyparticipants in a project team can exchange design andconstruction information to reduce waste and change orderson the BIM platform [23] BIM technology is perfectlysuitable to acts as a powerful tool for the integrated projectteams [16] Moreover the influence of organizational cultureon integrated project teams will be effective once BIM isimplemented As for this question Howard contended thatBIM services offered to IPD will facilitate collaboration inintegrated project teams making the project goalrsquos link toreality clearer [24]

Researchers have analyzed the performance of tradi-tional project delivery systems Indeed there is increasing

Advances in Civil Engineering 3

proof that a delivery method with high-level collaborationyields better performance than a delivery method with low-level collaboration However the AEC industry has not fullyexploited the potential value of integrated project deliverysystem For example although 84 of AIA members learnabout the concept of IPD less than 40 have a deeperunderstanding of it and only 13 have executed it [25] eefficiency and effectiveness of project performance not onlydepend on the quantity but also the quality of knowledge-sharing interactions between team members erefore it isessential to promote collaboration for IPD team from theperspective of knowledge-sharing

22 Knowledge and Knowledge-Sharing

221 Knowledge Knowledge is regarded as a key factor infavor of enhancing competitive advantage for organizationand project team [26] Organization and project team with alot of valuable knowledge are expected to achieve out-standingly [19] Importantly it has replaced traditional re-sources such as capital and labor as the most importantresource in construction [20] Polanyi [27] divided knowledgeinto two types namely explicit and tacit e former refers toknowledge that can be transmitted by a database photographor mathematic formula via computer programming tech-nology On the other hand the latter derives from individualobservations and perception of experience expertise insightsand talents none of which can be depicted accurately [28]Since then many scholars have followed such a division andstudied explicit knowledge and tacit knowledge respectivelyin many fields like education psychology project manage-ment informationmanagement and so on [29] Hansen et al[30] divided knowledge within an organization into twoforms codification and personalization Codification refers tothe organization converting personal knowledge into theform of explicit knowledge In this approach two codingsystems of knowledge-sharing are presented herein work-flow and database e organization encodes knowledge intoworkflow and information flow for easy use by all teammembers Personalization which is usually used to solvestrategic and low-repeatability problems refers to connectingmembers who do not possess kinds of knowledge with themembers who do possess it in the form of person-person orperson-linkage-person Project teammembers always hope toobtain the tacit knowledge owned by the seasoned experts forimproving personal ability and skills However suchknowledge must be transformed into explicit knowledge firstwhich is also a subject of academic investigation Couchmanand Fulop [31] took advantage of the nature language pro-cessing technology to analyze records and documents pro-vided by enterprises ey thus proposed a framework toextract tacit knowledge wherein their results revealed thattacit knowledge can be translated into explicit knowledgemore efficiently using this framework Nonaka [32] critiquedthis approach stating that the authors focused more onaccessing and acquiring explicit knowledge while thisknowledge only accounts for only a small portion of allpossible knowledge However Wu et al [33] proposed a

different viewpoint that is to say explicit knowledge is one ofthe most important resources to promote project perfor-mance when considering rapid development of imageidentification technology It is believed that a dispute re-garding which type of knowledge is more valuable wouldlikely miss the essence of the question without factoring inknowledge-sharing In fact both tacit and explicit knowledgeare mutually dependent they collectively strengthen thesignificance of knowledge tacit knowledge constitutes thebackground required for specifying the structure to exploitand explain the explicit knowledge erefore we do notdistinguish the two concepts but treat them as a wholenamely knowledge

222 Knowledge-Sharing Knowledge-sharing is generallyimplemented through various channels within or across anorganization in the form of documentation databasescommunication and group discussion [34] It is a complexprocess that involves knowledge collection absorption andtransformation it is considered critical to improve teamflexibility [35] Issa and Haddad [36] claimed that knowl-edge-sharing as a significant component of knowledgemanagement reflects the provision or receipt of work in-formation skills and feedback with respect to a product[37] Navimipour and Charband [38] believed that knowl-edge-sharing plays an important role in a project teambecause it offers a link between the member and the projectteam by cutting down cost and enhancing the projectperformance e authors conducted a comprehensive andsystematic review of the knowledge-sharing mechanism byscreening 28 out of 71 papers identified According to thefindings knowledge-sharing mutually benefitting re-lationship sense of self-worth and external motivation arecrucial factors of an academicianrsquos opinion towardsknowledge-sharing and work efficiency innovation andorganizational learning can be impacted positively byknowledge-sharing Moreover the authors also found thatwhen a trust-based relationship is established among teammembers knowledge-sharing behavior occurs more quicklyHowever they neglected to monitor the influence of ab-sorption capability on the effect of knowledge-sharing owingto limited resources from data bases Annadatha [39] ex-amined social-cultural factors such as trust shared goalsclose relationship and shared language as important factorstoward knowledge-sharing in a virtual project team by socialnetwork analysis approach e results showed that trustclose relationship and language importantly affect theoutcome of knowledge-sharing while shared goals do nothave significant impact on knowledge-sharing However asan important social-cultural factor knowledge-sharingwillingness is not checked Farajpour [40] further developeda four-leveled hierarchical inference system which iscomposed of six fuzzy rule bases to evaluate the informationdegree in the supply chain e results showed that theinformation sharing degree is influenced by willingnessimportantly However this research does not address howwillingness affects information sharing behavior dynami-cally from theoretical foundation perspective


Khatri et al [41] pointed out that the team membersshould assist coworkers in realizing their potential and goingbeyond their limitations in that context the trust re-lationships between team members would be generated andthey are inclined to share knowledge with each other [42]however they did not distinguish between the types ofknowledge such as heterogeneous knowledge or commonknowledge Further Yuan et al [43] found that projectcommitment directly influenced explicit knowledge-sharingand mutual trust however its influence on tacit knowledge-sharing was absent Zhang and He [44] conducted asyste-matic literature review to detect 31 factors influencingimplicit-knowledge-sharing in an IPD team by using aquestionnaire ey collected data from 300 respondents tothe factors influencing implicit-knowledge-sharing impor-tantly through significant analysis Subsequently factoranalysis was adopted to gather the similar factors such asswift trust (ST) personal benefits (PBs) lack of self-effi-ciency (LSE) identification-based trust (IBT) and in-formation-based trust (INBT)e authors later made use ofpath analysis to analyze the interrelationship between thesefive critical factors and their influence on implicit-knowl-edge-sharing e factor analysis revealed that PB is posi-tively correlated with ST and IBT IBT can shorten self-efficiency which affects tacit knowledge-sharing negativelyis study displayed a deep analysis of what sort of trustimpacts implicit-knowledge-sharing besides making clearthe close relationship between trust and implicit-knowledge-sharing It also found that ST influences INBT and IBTimportantly which in turn is positively correlated withimplicit-knowledge-sharing However the authors did notconsider the influence of knowledge stock on tacit knowl-edge-sharing behavior

Zareie et al [45] investigated the influence of electronicenvironment knowledge (EEK) on the environment be-havior (EB) by collecting data from 330 students e authorestablished a structural model of the key factors affectingenvironment behavior Subsequently the model was vali-dated using the smart PLS 20 e results proved that EEKinfluenced the EB directly in education Later the authorstudied the relationship of e-learning system between theemployeersquos commitments with the same methode resultsshowed that employeersquos commitments are positively andcrucially affected by learnerrsquos satisfaction readily availabletraining material personalized autonomous learning andwork efficiency and subsequently positively relate to theemployeersquos competitive advantage [6] Further learnersrsquo oremployeersquos satisfaction was positively and vitally influencedby technology education content motivation and attitude[46]en a comprehensive literature review on knowledge-sharing mechanism in education was conducted by Char-band and Navimipour [47] revealing how competitive ca-pability creativity learning effect and interaction behaviorcan be enhanced or optimized by knowledge-sharingHowever whether the findings can be applied to the con-structions industry is doubtful because the knowledge-sharing mechanism in the education field has a great dif-ference from construction project team especially for anintegrated project team First educational organizations are

mostly nonprofit while the construction project teamoperates for profit Moreover projects are established to onlycomplete a specific project task after which the projectteams are dissolved that is to say they are temporary Inconstruction the composition of the project team membersis not immutable but constantly adjusted as the projectprogresses or changes Hence the project team is charac-terized by openness e project team also often consists ofmembers with different majors and experiences and theconstruction project is comprehensive complex Finallysuch projects have a significant constraint of quality du-ration cost environment capital management technologysafety and other objectives

An important development in knowledge-sharing hasbeen the growth of cloud computing In order to enablecloud users acquire necessary human expertise at any lo-cation and share their own experience and knowledge in thecloud Navimipour et al [9] proposed an expert cloud-basedframework and in their study this system enhanced theorganizationrsquos performance and customer satisfactionHowever the findings did not consider human resources indifferent regional contexts Whether these findings can begeneralized requires further investigations Later regardingarchitectural problems and component analysis of expertclouds Navimipour et al [48] used NuSMV model checkerArgo UML and Rebeca Verifier tools to extract the checkingattributes in the form of LTL and CTL formulas of controlbehaviors and validate the attributes automatically and theresults indicated that the system was reliable However thelayers and related components still need to be improved inthe future and the algorithm needs to be developed for thewhole components not only the specified properties ofcontrolled behavior Subsequently Fouladi and Navimipour[49] proposed a cloud-based knowledge-sharing frameworkutilizing the quality control (QC) criteria e authors usedthese criteria of QC to develop a ranking diagram of humanresources on basis of trust reputation cost and expertisewhich is offered for members to choose the required humanresources To rank the human resources in the expert cloudan AHP-based method was introduced to assign weights tofeatures by especially considering the interdependenceamong features e authors found that compared with theprevious studies a hierarchical structure improved thequality and speed rating of human resources Howeverdynamic and interactive nature of the relationship variablesin the cloud was not captured According to the descriptionof researches above we know that these researches on theknowledge-sharing so far mainly focused on the technologyand assessment mechanism perspective than the theoreticallevel to explore knowledge-sharing behaviors in the projectteam

Knowledge-sharing behavior within a project team doesnot occur spontaneously Knowledge-sharing not only haseconomic benefit but also has social relationship benefit[50] Appropriate reward or motivation is believed to play animportant role in knowledge-sharing Further the socialexchange and social capital theories highlight that organi-zational rewards such as promotion bonus and high salarycan promote knowledge contribution with greater frequency


in the knowledge management system [51] However lack ofsufficient benefits is one of the most important reasonshindering knowledge-sharing in the information context soit is critical to design a perfect incentive mechanism topromote it [52] Gupta and Govindarajan [53] sated that achallenging rewards system should be developed to en-courage employees to participate in sharing activities bylearning it from the case of Nucor Steel However it onlyoffered the suggestions according to results extracted fromthe practice perspective with no theoretical analysis of theoperational mechanism of cost of knowledge-sharing be-havior Jewels [54] studied the why the individuals serving inIT project teams are easily motivated to or suppressed to-wards participating in sharing expertise e key findingsrevealed that an individualrsquos inclination to share expertiseand skills with others is a function of the associated benefitsand costs However how the function of benefits and costacts on the knowledge-sharing behavior is not discussed Byanalyzing the benefits and cost of knowledge-sharing be-tween a traditional pyramid organization structure and anetworked enterprise structure Liu [55] maintained that it isnecessary for enterprises to flatten the organizationalstructure in order to improve coordination among teammembers However while direct benefits were consideredhere the potential impact of the synergy benefit from thecoordination among project team members was notNavimipour and Soltani [56] then investigated the influenceof key factors including cost technology acceptance andsatisfaction of employee on the effectiveness of electroniccustomer relationship management (E-RCM) by establish-ing SEM and it was tested based on questionnaire datacollected e results indicated that the effectiveness ofE-RCM is affected positively by customer cost However thecross-sectional data were collected at a single point of timeand hence they cannot dynamically capture the relation-ships among influence factors at a different stage Accordingto aforementioned researches we can conclude that benefitsand cost affect the behavior of knowledge-sharing greatlyHowever these literatures describe the benefit and cost onlyas a total concept respectively while exploring theknowledge-sharing behavior statically from macro aspects

Recently with the increase of project scale complexityand dynamic property in AEC industry more and moreteam leaders across modern project organizations areseeking the way to master and thus make advantage of theenormous multidisciplinary knowledge [19] Organizationsin the AEC industry presents apparent knowledge-intensivecharacteristic highly depending on the rich knowledge stockand expertise derived from project team members [57 58]Based on the data collected from the respondents Cooke[59] made use of structural equation modeling to study therelationship between knowledge-sharing and project com-plexity e results showed that knowledge-sharing canmitigate the impact of project complexity Ribeiro [60] alsoproposed the similar point of view that project teammembers in the AEC organizations need to actively shareknowledge and expertise to deal with complicated anddifficult tasks Xia and Chan [61] argued that the con-struction complexity affects the success of a project and

developed a composite complexity index for measuring theconstruction complexity through tapping expert knowledgein construction field and thus help stakeholders and con-structors take appropriate actions to improve the projectperformance Owning to the importance of knowledge-sharing behavior organization leaders should be consciousof the tremendous advantages brought by knowledge-sharing behavior [62] and try to study the measures that canimprove the project performance from the perspective ofproject teammembers [34] China as the largest engineeringconstruction country in the world has many buildings andinfrastructures in progress A vast amount of knowledge canbe repeatedly applied to different types of constructionprojects despite the unique disposable characteristic of theconstruction project [63] e construction managementorganizations with the temporary feature easily lead toplenty of knowledge loss when project teams are disbandederefore it is necessary for stakeholders to share their ownvaluable knowledge with other stakeholders in the projectteam to avoid the ldquoreinvention of the wheelrdquo in the other newprojects [64]

e systematic review of the extant literature reveals thatwith the emergence of large and complex engineeringprojects the ldquolooserdquo collaboration pattern under traditiondelivery has not been able to meet the needs of efficientproject management On the other hand the IPD approachbased on the lean construction theory becomes the researchfocus of the construction industry Most researchers haverecognized the importance of knowledge-sharing to projectteams and presented its effectiveness to improve projectperformance However few have focused on the selectionstrategies of knowledge-sharing and thus explored how thesestrategies change dynamically over time in the IPD teamerefore the objective our objective is to capture the in-teraction behavior of knowledge-sharing among IPD teammembers and explore its evolutionary tendency associatedwith respect to influence factors

3 Research Method

Cooperative and competitive relationships exist amongintegrated project team members Whether or not theproject team members choose to share knowledge is acomplicated and dynamic game process erefore to ex-plore the knowledge-sharing strategy and evolutionarymechanism for IPD team the most important work is todevelop an algorithm that can capture dynamically thestrategy interaction behavior Currently according to theavailable studies the methods to focus on knowledge-sharing mechanism such as SEM case study classic gametheory and social network analysis that we will analyzerespectively in the following can merely capture the staticstrategy interaction behavior at a point of time whichcannot predict the knowledge-sharing behavior evolutionarytrend over time Leveraging the evolutionary game theory tostudy knowledge-sharing is helpful to dynamically analyzethe process of strategy selection from the micro perspective

SEM is a method capable of measuring the underlyinglatent constructs identified by factor analysis and assessing


the path of the hypothesized relationships between theconstructs to address the complicated behavioral relation-ships [65] However while SEM has been popular for itsadvantages it had been criticized for incredible conclusionowning to its indiscriminate utilization [66] Also Cliff [67]questioned the improper use of SEM and pointed out itspotential drawbacks First of all the data obtained by re-searchers cannot completely confirm or deny the correctnessof a model because the model is artificial and can beredefined in many ways Next chronological evidence doesnot represent cause and effect and the naming of latentvariables is a subjective process other than an objective factquite apart from the trap of nominalistic fallacy existing inthe estimation of latent variables Furthermore the re-searchers may unconsciously apply excessive modificationprogramming in the model for obtaining the ideal modelnamely overfitting problem which is really hard to controlin data processing leading to larger distortions betweenresults and the real situation [68] Recently Xiong et al [69]analyzed the limitations and risks in SEM applicationsthrough systematic literature review of 84 articles in solvingconstruction research problem and found that sample sizegoodness of fit measures and especially construct validityare significant problems need to address Although aguideline framework is given to solve these issues howeverthe substantial weakness that it only reflects statically asituation at a point of time and is unable to capture thedynamic interaction behavior cannot be solved

Due to the number of variables processed by SEM andthe complex relationships among variables a large samplesize must be used to maintain an unviolated statistical hy-pothesis erefore the influence of sample size is importantissue for SEM Breckler [70] conducted an analysis of 72SEM empirical studies in the field of personality and socialpsychological e sample size is between 40 and 8650 witha median of 198 One-fourth of the studies has a sample sizesmaller than 500 while 20 had smaller than 100 eauthor concluded that a sample size larger than 200 can beregarded as a medium-sized sample while that below 200 isdiscouraged when pursuing stable SEM results Conse-quently reliable results are impossible with a small samplesize besides the overfitting problem which is hard tocontrol Such SEM drawbacks can easily lead to faultyconclusions In fact SEM is mainly suitable for research onthe relationships among constructs from macro perspective[71] and cannot explore the mechanism of action of theconstructs from the inner nature of things However thenovel evolutionary game model we proposed can avoid thisdilemma because it can simulate IPD teammember strategyinteraction behaviors determined by the key factors is isdone by setting sufficient units representing the sample sizeof the team members In other words the sample size doesnot affect this method

Some researchers for example Yoo [72] and Annadatha[39] also use case study and social networks analysis re-spectively to investigate the knowledge-sharing mechanismamong project team members e case study is an effectiveapproach to provide empirical evidence indicating how keyfactors affect the outcome of knowledge-sharing behavior

[73] However it can only reflect the external phenomenonbut not the internal mechanism Further social networkanalysis emphasizes interpersonal relationship relationshipconnotation social network structure and social phenom-enon which can assist in regulating the project team andidentify knowledge-sharing barriers [74] However this typeof analysis also cannot determine the dynamic variationtrend of knowledge-sharing behavior

Classic game theory is often used by some scholars inknowledge-sharing mechanism for example Luo and Yin[75] is theory was initially exploited to analyze strategicinteractions in the economic sphere It has been frequentlyapplied in similar fields over the past decades with en-couraging achievements [76] In most researches to confirmthe model prediction to theoretical analysis the classic gametheory assumes that the game players should exhibit high-level rationality and possess complete information of theother game players Further mistakes are not allowedthroughout the game process [77] However this extremelystrict assumption cannot be satisfied inmost cases because itis nearly impossible for each game player to be entirely awareof the complete information on competitors [78] Addi-tionally the decision-making is also influenced significantlyby the knowledge level of the game player In fact the gameplayers who are limited by rationality often make dynamicchanges to the strategies by comparing the payoff with othercompetitors [79] Hence the assumption of super rationalitylimits the application of Classic game theory

To solve this problem mentioned above the evolutionarygame theory was developed by replacing high-level rationalitywith bounded rationality assumption allowing informationincompleteness and asymmetry [80] is theory deals withthe emergence transformation diffusion and stabilization ofbehavior forms It integrates the idea of evolutionary biologyand rational economics combining game analysis withevolutionary dynamics to provide a powerful analytical toolfor the study of the cooperative behaviors and a replicatorsystem as the core of the evolutionary system is utilized toexpress the evolution mechanism of the game players [81]is characteristic of the evolutionary game theory effectivelyexplains the behavior of long-term economic and transactionrelations among large populations well So far economistsand sociologists have made remarkable achievements byemploying evolutionary game theory to analyze the factorsthat influence the formation of social habits and social systemand it has been used to explain the process of formation [82]

According to a previous description the most importantfinding obtained is that evolutionary game theory is asuitable method with the characteristics to predict theevolutionary trend of strategy interaction behaviors andconsequently based on its unique advantage we propose anovel evolutionary game model to study the knowledge-sharing strategies and the evolutionary mechanism

4 Evolutionary Game Model

41 Model Parameters Based on a comprehensive andsystematic literature review we can identify that the factorsof knowledge stock degree of knowledge-sharing


heterogeneous knowledge synergy effect knowledge ab-sorption capability and cost control capability of knowl-edge-sharing have been rarely studied systematically inknowledge-sharing fields However these factors play acritical role in affecting the cooperative behavior Hence thefollowing model variables are presented based on criticalanalysis as follows

Ki knowledge stock refers to the total amount ofknowledge of an organization which reflects its competi-tiveness and potential to deal with tasks Researchers havedifferent points of view on knowledge but reach a broadconsensus Knowledge is a necessary and sustainable re-source and has become the most important element in allwalks of life [52] In this study we hold that the definition ofknowledge should be developed not only according toeconomic significance but also easy operability observ-ability and measurability A successful organization orenterprise in the construction industry can always create andspread knowledge among project team members andeventually embody this knowledge in products and services[83] e traditional project management system based oncapital and labor can no longer effectively explain thephenomenon of performance improvement of the projectteam [84] Knowledge as a resource like other capital can beproduced and exchanged Its accumulation is a prerequisitefor successful project management However how to pre-serve knowledge and make use of knowledge resourcescreated by project team members are challenges we face atpresent which involves concept of knowledge stock Wethus introduce it as a parameter into the evolutionary gamemodel

ηi degree of knowledge-sharing refers to willingness toshare the knowledge provider in the integrated project teame stronger the willingness to share the higher the degreeof knowledge-sharing e willingness to share knowledgemainly reflects the extent to which the project teammembersare inclined to share and provide relevant knowledge andexperience with the other team members through variousways and activities in the workplace [85] According to thesocial cognition theory [86] behavior can be best explainedby the continuous interaction between cognition and en-vironmental factors In order to achieve a sense ofachievement in work strong intrinsic motivation plays animportant role in triggering the willingness to share [87 88]Intrinsic motivation refers to the satisfaction brought byteam members towards the work It is derived from theparticipantsrsquo strong pursuit and willingness to challenge thework Team members with intrinsic motivation are curiousand enthusiastic about their work and they tend to exploreunusual solutions spontaneously and actively Such be-haviors indicate that these project teammembers can show ahigh level of creativity [89 90] However only depending onintrinsic motivation to improve the degree of knowledge-sharing is not sufficient Without trust-based business re-lationship and rewards the effect of willingness will beundermined [91] Hence we can find that the degree ofknowledge-sharing is affected by three factors includingintrinsic motivation trust-based relationship and incentivemechanism e extant literature has rarely focused on the

relationships between the willingness and knowledge-sharing In this paper we deliberately introduce degree ofknowledge-sharing as a parameter in the evolutionary gamemodel to fill this knowledge gaps

ui Heterogeneous knowledge proportion ui refers tothe ratio of the amount of complementary knowledge to theamount of shared knowledge e larger the heterogeneousknowledge proportion the stronger the complementarityShi et al [92] divided knowledge into complementary andhomogeneous knowledge types according to the similarity ofthe knowledge structuree former is the inherent attributeof knowledge It is the core knowledge that team membersdistinguish from the others and can bring competitive ad-vantages for team members [93 94] e project organi-zation exists as a system and comprises smaller subsystemsAccording to system theory the components within thesystem are interrelated It is this close connection that allowsthe personnel to consider the system as a whole while payingattention to the interaction behaviors among its distinctparts Traditionally in order to improve labor efficiencyorganizations are operated through strict division of laborHowever strict specialization not only improves labor ef-ficiency of workers but also disrupts internal connectionWhile managers do help coordinate the overall organiza-tional efficiency declines due to the tacit feature of knowl-edge and information distortion that occur duringtransmission at is true especially when it comes to theutilization and innovation of knowledge [95] Specializationhas its limitations and specialization itself is subject todiminishing returns Overspecialization leads to losses be-cause the division of labor and knowledge must be co-ordinated wherein the latter requires coordination becausethey are complementary

For an enterprise or organization knowledge comple-mentarity could increase returns because the fusion ofdiscrete and complementary knowledge can bring aboutinnovation possibly leading to cost reduction quality im-provement and profit increase [96] is is particularlyimportant in the information economy where knowledgeassets are increasingly replacing physical assets Henceheterogeneous knowledge is an important parameter that weintroduce into the evolutionary game model

βi Synergy effect coefficient βi refers to the knowledge-sharing effect that reflects the capability of integrating thecomplementary knowledge existing among different projectteam members It is originally a physical and chemicalphenomenon wherein when two or more components areadded or mixed together the resulting effect is greater thanthe sum of the various components when utilized aloneatis ldquo1 + 1gt 2rdquo Ever since the concept of synergy effect wasintroduced into the field of enterprise management it hasbeen become an important element of enterprise or orga-nization for efficiency improvement e synergy effect isdivided into four types sales operation investment andmanagement synergies On the other hand Chen and Liu[97] divided it into five types from the perspective of en-terprise resources strategic cultural human resourcesupply chain and financial synergies While difference existsbetween two classification methods they both proved that


synergy effect is important to increase the benefits to par-ticipants in collaboration e extant literature has revealedthat the knowledge synergy between organizations iscomplex and not a simple linear manner in knowledgetransfer In this study based on the emergency theory ofpartner selection combined with the research of Montoyaet al and Zeng et al [98 99] and the management practicederived from Chinese enterprise we take into considerationthe influence on the synergetic effect coefficient from theperspective of synergy management system including cul-tural synergy technical synergy and organization synergye synergetic effect is an important factor influencing theproject performance thereby enabling it as a parameter inthe evolutionary game model

αi Knowledge absorption coefficient αi refers to thecapability of knowledge receivers to assimilate the knowl-edge that transmitted by the knowledge providers andconvert it into their own competitive knowledge is playsan important role in improving the knowledge-sharingbenefit of the knowledge providere higher the knowledgeabsorption coefficient the higher the capability of assimi-lating knowledge e absorptive capacity theory argues thatthe knowledge absorptive capacity is the ability of an in-dividual or organization to identify the value of new outsideinformation and then absorb and apply it for a specificbusiness purpose e individual knowledge absorptivecapability determines whether knowledge receiver can up-date their own knowledge system in a timely and accuratemanner [100] us it can be inferred that the innovationperformance of project team members is closely related toindividual knowledge absorption capability from the per-spective of absorptive capacity theory which itself has beenwidely used in the study of enterprise innovation perfor-mance [101] However most studies have mainly focused onknowledge absorption capacity at the organizational levelbut rarely from the individual level us we chooseknowledge absorption coefficient as an important parameterinvolved in the evolutionary game model

Ci Knowledge-sharing cost coefficient Ci reflects theability of the knowledge provider to control costs duringknowledge-sharinge higher the value of Ci the weaker ofthe cost control capability Cost control positively takes aseries of prevention and adjustment measures to ensure therealization of cost management goals is is carried out byan enterprise or project team with the scope of its functionsand powers and according to preestablished cost objectivesIt fully makes use of the principles of system engineering tocalculate and supervise all kinds of costs incurred in theproduction and operation process Cost control is also aprocess of discovering weakness tapping internal potentialand finding all possible ways to reduce cost e con-struction project is the basic part of construction industry inwhich cost control is particularly important Scientific or-ganization and implementation of cost control can promoteenterprises to improve operation and management helpchange the operation mechanism and enhance the com-petitiveness of the project teammemberse importance ofcost control capability cannot be overstated because onlywhen the construction cost of the project is reduced to a

reasonable range can the enterprise and project team im-prove benefits and maintain virtuous development ere-fore the knowledge-sharing cost efficient as an importantparameter is introduced into the evolutionary game model

42 Hypothesis H1 e game is regarded as an un-observable system that comprises two game players Group 1and Group 2

First of all although the IPD team is comprised ofmultidisciplinary stakeholders to simplify the study withoutloss of generality two groups Group 1 and Group 2 areselected randomly as the game players Each group can playthe role of the knowledge receiver or provider

Next all players obtain the information about them-selves and the others before making a decision In otherwords owing to the incomplete information characteristicthe two groups select their strategies simultaneously andneither can observe the other playerrsquos choice and whetherthe payoff obtained from the other part is attractive or notwhile making decisions erefore this game is an un-observed system that is in alignment with the boundedrationality assumption of the evolutionary

H2 e pure strategy set for each group is set to beldquosharerdquo and ldquonot sharerdquo strategies with the proportion of xyand 1minus x1minus y for Group 1 and Group 2 respectively

H3 e Revenue obtained from the ldquonot sharerdquo strategyis assumed to be Ri of which i (1 2)

H4 e value-added benefit including direct and syn-ergetic benefits will be obtained by two groups when bothgroups adopt the ldquosharerdquo strategy

Direct benefit refers to the benefit acquired by theknowledge receiver absorbing knowledge from knowledgeprovider For the knowledge receiver such as Group 1 thedirect benefit is subject to the degree of knowledge-sharingη2 the heterogeneous knowledge proportion u2 knowledgeabsorption capability α1 and the knowledge providerrsquosknowledge stock K2 namely α1μ2K2η2 Similarly the directbenefit for Group 2 is α2μ1K1η1

For one participant the amount of knowledge fusion isthe sum of knowledge shared by this participant andknowledge by absorbing the other participantrsquos knowledgeKiηi + αiμiKiηi Specifically the amount of knowledge fu-sion for Group 1 and Group 2 is K1η1 + α1μ2K2η2 andK2η2 + α2μ1K1η1 respectively Synergetic benefit is subjectto the synergetic capability coefficient βi and the amount ofknowledge fusion Kiηi + αiμiKiηi erefore for Group 1and Group 2 the synergetic benefit is β1(K1η1 + α1μ2K2η2)and β2(K2η2 + α2μ1K1η1) respectively

H5 e knowledge-sharing cost is subject to theknowledge-sharing cost coefficient Ci and the amount ofknowledge that can be shared by the participant Kiηi

According to the hypothesis mentioned above the payoffmatrix (Table 1) can be obtained In the evolutionary gametheory payoff matrix also known as ldquowinning matrixrdquo isused to describe the strategies and payments of two or moregame players It comprises alternative action plans naturalstates profit and loss (or utility) values As described inTable 1 in the scenario (A1 B1) both participants obtain


direct and synergetic benefits Meanwhile it costs C1K1η1and C2K2η2 respectively to perform strategy (A1 B1)erefore the total payoff acquired is R1 + α1μ2K2η2+β1(K1η1 + α1μ2K2η2)minusC1K1η1 and R2 + α2μ1K1η1 + β2(K2η2 + α2μ1K1η1)minusC2K2η2 respectively If Group 1chooses to share knowledge and Group 2 does not to shareknowledge namely strategy (A1 B2) the former will bearcost with no benefit whereas the latter will obtain directbenefit with no cost erefore the total payoff for bothparticipants is R1 minusC1K1η1 and R2 + α2μ1K1η1 Similarlythe payoff is R1 + α1μ2K2η2 and R2 minusC2K2η2 for Group 1and Group 2 in scenario (A2 B1) respectively If two groupsall choose not to share knowledge they will not obtain anydirect and synergetic benefits Hence the payoff for Group 1and Group 2 is R1 R2 respectively

43 Evolutionary Game Model e expected benefit ofGroup 1 that agrees to choose strategy A1 is Ea

1

Ea1 y R1 + α1μ2K2η2 + β1 K1η1 + α1μ2K2η2( 1113857minusC1K1η1( 1113857

+(1minusy) R1 minusC1K1η1( 1113857

(1)

e expected benefit of Group 1 that does not agree tochoose strategy A2 is En

1

En1 y R1 + α1μ2K2η2( 1113857 +(1minusy)R1 (2)

e expected average benefit of Group 1 under the mixedstrategy is E1

E1 xEa1 +(1minus x)E

n1 yα1μ2K2η2 + xβ1α1μ2K2η2

+ xyβ1K1η1 minus xC1K1η1(3)

According to the evolutionary game theory if the benefitof a strategy adopted by one game player is higher than theaverage population the strategy diffuses within the entirepopulation en the frequency of the game playerschoosing the strategy will improve within the populatione dynamic differential equation which is referred to as thereplication dynamic equation can be utilized to depict thefrequency of the strategy adopted within the population Onthe basis of equations (1) and (3) the replication dynamicequation of Group 1 is

fA dx

dt x E

a1 minusE1( 1113857 x(1minus x)

middot β1y K1η1 + α1μ2K2η2( 1113857minusC1K1η11113858 1113859

(4)

Similarly the expected benefit of Group 2 that agrees tochoose strategy B1 is Ea

2

Ea2 x R2 + α2μ1K1η1 + β2 K2η2 + α2μ1K1η1( 1113857minusC2K2η2( 1113857

+(1minusx) R2 minusC2K2η2( 1113857

(5)

e expected benefit of Group 2 that does not agree tochoose strategy B2 is En

2

En2 x R2 + α2μ1K1η1( 1113857 +(1minus x)R2 (6)

e expected average benefit of Group 2 is

E2 yEa2 +(1minusy)E

n2 xα2μ1K1η1 + xyβ2α2μ1K1η1

+ yxβ2K2η2 minusyC2K2η2(7)

According to equations (5) and (7) the replicationdynamic equation that is constructed is

fB dy

dt y E

a2 minusE2( 1113857 y(1minusy)

middot β2x K2η2 + α2μ1K1η1( 1113857minusC2K2η21113858 1113859

(8)

From equations (4) and (8) a two-dimensional non-linear dynamic system can be acquired

dx

dt x(1minus x) β1y K1η1 + α1μ2K2η2( 1113857minusC1K1η11113858 1113859 0

dy

dt y(1minusy) β2x K2η2 + α2μ1K1η1( 1113857minusC2K2η21113858 1113859 0

⎧⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎩

(9)

Based on the solution of equation (9) we can obtain fivelocal equilibrium points (LEP) from the nonlinear dynamicsystem A(00) B(01) C(10) D(11) E(xlowast ylowast) wherein

xlowast C2K2η2

β2 K2η2 + α2μ1K1η1( 1113857

ylowast C1K1η1

β1 K1η1 + α1μ2K2η2( 1113857

⎧⎪⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎪⎩

(10)

Full advantage of the Jacobian matrix (J) is taken toqualitatively analyze the five LEP e Jacobian matrix thatcan be obtained from the game is

J

zfA

zx

zfA

zy

zfB

zx

zfB

zy

⎡⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣

⎤⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦

(11)

erefore

Table 1 Payoff matrix

Group 1Group 2

B1(y) B2(1minus y)

A1(x)R1 + α1μ2K2η2 + β1(K1η1 + α1μ2K2η2)minusC1K1η1 R1 minus C1K1η1R2 + α2μ1K1η1 + β2(K2η2 + α2μ1K1η1)minusC2K2η2 R2 + α2μ1K1η1

A2(1minus x) R1 + α1μ2K2η2 R2 minusC2K2η2 R1 R2


zfA

zx (1minus 2x) yβ1α1μ2K2η2 + yβ1α1K1η1 minusC1K1η1( 1113857

zfA

zy x(1minus x)β1 α1μ2K2η2 + K1η1( 1113857

zfB

zx y(1minusy)β1 α1μ2K2η2 + K1η1( 1113857

zfB

zy (1minus 2y) xβ2α2μ1K1η1 + xβ2α2K2η2 minusC2K2η2( 1113857

(12)

Based on Friedmanrsquos theory we can conclude that A BC and D are local pure strategy equilibrium points Addi-tionally E is a local equilibrium point that has a mixedstrategy With respect to the stability of the five points it canbe acquired by analyzing the determinant (det(J)) and trace(tr(J)) of the Jacobian matrix For any pointdet(J) (zfAzx)lowast (zfBzy)minus (zfAzy)lowast (zfBzx) andtr(J) (zfAzx) + (zfBzy) Specifically the values of det(J)and tr(J) of the Jacobian in the five points mentioned aboveare presented in Table 2

For a certain point that satisfies the evolutionary gamestable state (ESS) it is subject to det(J)gt 0 and tr(J)lt 0 and itpossesses the asymptotic stabilityereby the correspondingstrategy portfolio is regarded as a stable equilibrium solutionIf det(J) ge 0 and tr(J)gt 0 the point is defined as unstable Ifdet(J)gt 0 and tr(J) 0 the point is defined as neutral Ad-ditionally if det(J)lt 0 the point is defined as a saddle point

44ModelAnalysis In terms of the range of values for xlowast andylowast four scenarios are derived from the evolutionary gameEach equilibrium point and its stability decision processunder the corresponding scenario are presented in Table 3

5 Simulation of the Model

For the purpose of further understanding the theoreticalresults analyzed in Section 4 we develop a program tosimulate the evolutionary process of interaction behaviorsamong integrated project team members e simulation isconducted using MATLAB 2016a e RungendashKuttamethod embedded in the simulation platform is introducedto settle the differential equation group of the dynamicreplicator system

Function ode45 which is built in MATLAB platform isused to obtain the simulation solutions Parameter settingshould be completed before implementing the simulatione evolutionary time span is limited to the range of values[0 10] and the initial point is assumed to be P(x0 y0)wherein 0le x0 y0 le 1 According to the model parametersmentioned in Section 41 the base values are set toK1 20000 K2 10000 η1 05 η2 05 β1 01β2 01 α1 02 α2 05 μ1 02 and μ2 05

51 Verification for Evolutionary Tendency With the othermodel variables in line with the base values the parameters

C1 and C2 are set to 005 under conditions 0ltxlowast lt 1 and0ltylowast lt 1 at is to say the evolutionary tendency ofknowledge-sharing is subject to the conditionsC2K2η2 lt β2(K2η2 + α2μ1K1η1) and C1K1η1 lt β1(K1η1 +

α1μ2K2η2) In this scenario 100 array points are selectedrandomly as the initial evolutionary points that represent theproportion of the integrated project teammembers who choosethe ldquosharerdquo strategy in Group 1 and Group 2 respectively eevolutionary outcome is described in Figure 2(a) We canconclude that the general evolutionary path is attracted to thetwo stable points namely point (00) and point (11) which isin accordance with 1(a)

With the other model variables consistent with the basevalues the parameters C1 and C2 are set to 005 and 05respectively under the condition xlowast gt 1 and 0ltylowast lt 1 atis to say the evolutionary tendency of knowledge-sharing issubject to the condition that C2K2η2 gt β2(K2η2 + α2μ1K1η1)and C1K1η1 lt β1(K1η1 + α1μ2K2η2) In this scenario 100array points are selected randomly as the initial evolutionarypoints that represent the proportion of the integrated projectteam members who choose the ldquosharerdquo strategy in Group 1and Group 2 respectively e evolutionary result isdepicted in Figure 2(b) We can see that the general evo-lutionary path is attracted to stable points (00) which isconsistent with 1(b)

With the other model variables consistent with the basevalues the parameters C1 and C2 are set to 05 and 005respectively under the conditions 0ltxlowast lt 1 and ylowast gt 1atis to say the evolutionary tendency of knowledge-sharing issubject to the conditions C2K2η2 lt β2(K2η2 + α2μ1K1η1)and C1K1η1 gt β1(K1η1 + α1μ2K2η2) en 100 array pointsare selected randomly as the initial evolutionary points thatrepresent the proportion of integrated project team mem-bers who choose the strategy ldquosharerdquo in Group 1 and Group2 respectively e evolutionary result is shown inFigure 2(c) We can see that the general evolutionary path isattracted to the stable point (00) which is in line with 1(c)

With the other model parameters remaining the same asthe base values the parameters C1 andC2 are set to 05 underthe condition xlowast gt 1 ylowast gt 1 at is to say the evolutionarytendency of knowledge-sharing is subject to the conditionsC2K2η2 gt β2(K2η2 + α2μ1K1η1) and C1K1η1 gt β1(K1η1 +

α1μ2K2η2) en 100 array points are selected randomly asthe initial evolutionary points that represent the proportionof integrated project team members who choose the strategyldquosharerdquo in Group 1 and Group 2 respectively e evolu-tionary result is illustrated in Figure 2(d)We can see that thegeneral evolutionary path is convergent to the stable point(00) which is in line with 1(d)

In Figures 2(b)ndash2(d) we can see that the integratedproject team members in both Group 1 and Group 2 willeventually choose strategy (A2 B2) under scenario 2 sce-nario 3 and scenario 4 Hence the evolution trend is certainand clear In Figure 2(a) we can see that the two conditionsexist under condition 0lt xlowast lt 1 and 0ltylowast lt 1 When theinitial points are located in the upper right area the be-haviors of both groups either converge to point (00) in-dicating that both groups will choose the ldquonot sharerdquostrategy or converge to the point (11) indicating that both


groups will choose the ldquosharerdquo strategy when the initialpoints are located in the lower left Hence the evolutionaryoutcome is uncertain in scenario 2 and we will furtherdiscuss the impact of the model parameters on the evolu-tionary tendency of interaction behaviors among integratedproject team members in the following section

52 Influence of x and y e proportion of the integratedproject team members with different strategies in Group 1and Group 2 is studied firstly To set the simulation pa-rameters x and y reasonably and ensure the reliability andextensibility of this research findings we conducted semi-structured interviews through Internet search and socialnetwork relations us 20 experts with rich experience inproject team management are chosen ese experts aredivided as follows 15 with three to five years of workexperience 35 with five to ten years of work experienceand 50 with more than ten years of work experiences We

sorted out and adopted the suggestions proposed and ac-cepted by the most experts

Most experts argued that team members within alearning project team would show a strong tendency tocooperate Within this the proportion of the integratedproject teammembers who would prefer to share knowledgeplays an important role in cooperation ey also suggestedthat it was necessary to distinguish between different effortlevels referring to the proportion of team members pre-ferring to share knowledge at the early stage of organizingthe project team In terms of expertsrsquo suggestions the nu-merical values from 0 to 1 with an addition of 01 wereprovided for the experts to identify the appropriate range ofdifferent effort level According to the results of semi-structured interviews a value equal to or greater than 08 wasconsidered the high effort level and a value equal to orgreater than 05 and less than 08 is regarded as the moderateeffort level and a value equal to or less than 05 is regarded asthe low effort level Hence in this paper three values of 02

Table 2 Determinant and trace of Jacobian

LEP e determinant and trace of J

A(00)det(J) C1K1η1C2K2η2

tr(J)minus(C1K1η1 + C2K2η2)

B(01)det(J) C2K2η2(β1K1η1C2K2η2 + β1α1μ2K2η2C2K2η2 minusC1K1η1)

tr(J) (β1K1η1 + β1α1μ2K2η2 minusC1K1η1) + C2K2η2

C(10)det(J) C1K1η1(β2K2η2 + β2α2μ1K1η1 minusC2K2η2)

tr(J) C1K1η1 + (β2K2η2 + β2α2μ1K1η1 minusC2K2η2)

D(11)det(J) [C1K1η1 minus β1(K1η1 + α1μ2K2η2)][C2K2η2 minus β2(K2η2 + α2μ1K1η1)]

tr(J) [C1K1η1 minus β1(K1η1 + α1μ2K2η2)] + [C2K2η2 minus β2(K2η2 + α2μ1K1η1)]

E(xlowast ylowast)det(J) (minus(β1K1η1 + β1α1μ2K2η2 minusC1K1η1)(β2K2η2 + β2α2μ1K1η1 minusC2K2η2)

C1K1η1C2K2η2)(β1(K1η1 + α1μ2K2η2)β2(K2η2 + α2μ1K1η1))

tr(J) 0

Table 3 e equilibrium stability of the nonlinear dynamic system for four scenarios

Scenarios Range of values Points det(J) tr(J) Equilibrium results Phase diagrams

Scenario 1 0ltxlowast lt 1 0ltylowast lt 1

(00) + minus ESS

Figure 1(a)(01) + + Unstable(10) + + Unstable(11) + minus ESS

(xlowast ylowast) + 0 Saddle

Scenario 2 xlowast gt 1 0ltylowast lt 1

(00) + minus ESS

Figure 1(b)(01) minus Uncertain Saddle(10) + + Unstable(11) minus Uncertain Saddle

(xlowast ylowast) + 0 Neutral

Scenario 3 0ltxlowast lt 1 ylowast gt 1

(00) + minus ESS

Figure 1(c)(01) + + Unstable(10) minus Uncertain Saddle(11) minus Uncertain Saddle


Scenario 4 xlowast gt 1 ylowast gt 1

(00) + minus ESS

Figure 1(d)(01) minus Uncertain Saddle(10) minus Uncertain Saddle(11) + + Unstable

(xlowast ylowast) minus 0 Saddle


05 and 08 are adopted to represent the low eort levelmoderate eort level and high eort level respectively tosimulate the impact of x and y on the evolutionary trend

With the other model parameters remaining the samewith the base values by setting C1 005 and C2 005Figures 3(a)ndash3(c) elaborate the evolutionary trend of Group1 under the condition that the xed y values are 02 05 and08 which respectively represents the low eort moderateeort and high eort levels of Group 2 Comparing withFigures 3(a)ndash3(c) we can observe that the probability ofchoosing the ldquosharerdquo strategy in Group 1 even if at the loweort level improves obviously with the increase in theproportion of teammembers choosing B1 in Group 2 When80 (ie high-level eort) of integrated project teammembers in Group 2 choose the ldquosharerdquo strategy all theintegrated project team members in Group 1 eventuallychoose to share their knowledge with the other teammembers in Group 2 as shown in Figure 3(c) When theproportion of the integrated team members in Group 2 is atmoderate eort level (ie y 05) the integrated projectteam members with moderate and high eort levels (iex 05 and x 08) in Group 1 are eventually attracted tostrategy A1 as shown in Figure 3(b) Furthermore thesmaller the initial value of x is the faster the evolutiontendency of team members in Group 1 converges to 0 if theteam members of Group 2 are at the low eort level (iey 02) as shown in Figure 3(a) Similarly Figures 4(a)ndash4(c)

illustrate the evolutionary trend of Group 2 under thecondition that the xed x values are 02 05 and 08 Weobserve the same results as in Figure 3

Based on the analysis of the simulation results above wecan conclude the rst interesting nding that if the in-tegrated project team members in both groups are atmoderate- or high-eort level the interaction behaviors ofboth groups would evolve to the combined ldquoshare sharerdquostrategy which means that they will choose to shareknowledge with each other e second nding is obtainedthat as long as one of the two game players (Group 1 andGroup 2) is at high eort level initially it will drive the othergame player to evolve to the ldquosharerdquo strategy and both gameplayers will share their own knowledge ultimately

53 Inuence of Special Model Parameters

531 Inuence of Heterogeneous Knowledge ProportionWe rst analyzed the inuence of the heterogeneousknowledge proportion (μi) on two game players whichincreases from 02 to 06 with a step length of 02 by settingC1 005 and C2 005 and remaining the other parametersconsistent with the base values e initial point is assumedto be (04 06) e simulation results are shown inFigures 5(a) and 5(b)e interesting nding is obtained thatwith the increase of the heterogeneous knowledge

B

A

D

C

E (xlowast ylowast)

(a)

B

A

D

C(b)

B

A

D

C

(c)

B

A

D

C

(d)

Figure 1 Phase portraits


proportion both game players eventually evolve to thecombined strategy ldquoshare sharerdquo and the larger the values ofμi are the faster the values of x and y converge to 1 at iswhen heterogeneous knowledge proportion increases theprobability of adopting the combined strategy ldquoshare sharerdquoincreases for both game players

532 Inuence of Knowledge Absorption Coecient Wethen analyzed the impact of the knowledge absorption ca-pability (αi) on the choice of the combined strategy for bothgame players By setting C1 005 and C2 005 the pa-rameter αi increases from 02 to 08 with a step length of 02We leave the remaining parameters unchanged and theinitial point is assumed to be (04 06) e simulationoutcomes are shown in Figures 5(c) and 5(d) We canconclude that with increase of αi the proportion of choosingthe strategy ldquosharerdquo by both game players eventually con-verges to 1 and the larger the values of αi the faster the rateof convergence us as αi increases the probability of

adopting combined strategy ldquoshare sharerdquo improvesgradually

533 Inuence of Synergetic Eshyect Coecient Next theinuence of the heterogeneous knowledge proportion (βi)on the evolution results of the interaction behavior of thetwo game players was analyzed by setting C1 005 andC2 005 and then leaving the remaining parameters un-changed e parameter βi increases from 02 to 08 with astep length of 02e initial point is assumed to be (04 06)e simulation results are shown in Figures 5(e) and 5(f)e interesting nding can be obtained that with the in-crease of heterogeneous knowledge proportion both gameplayers eventually evolve to the combined strategy ldquosharesharerdquo and the larger the values of βi the faster the evolutionresults of the interaction behavior converge to 1 at iswhen the synergetic eect coecient increases the proba-bility of adopting combined strategy ldquoshare sharerdquo will beimproved for both game players

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)

FIGURE 2 Evolutionary paths of scenarios 1sim4 (a) scenario 1 (b) scenario 2 (c) scenario 3 (d) scenario 4


534 Inuence of Knowledge-Sharing Cost CoecientNext we evaluate the inuence of the knowledge-sharingcost coecient (Ci) on the interaction behavior of both gameplayers First the inuence of parameter C1 on both gameplayers is analyzed wherein C1 increases from 003 to 006with a step length of 001 while C2 008 e other modelparameters have the same base values and the initial point isassumed to be (04 06) e simulation result is shown inFigure 5(g) Second we analyze the inuences of parameterC2 on both game players wherein C2 increases from 003 to006 with a step length of 001 and C1 008 Remain theother model parameters the same with base values and theinitial point is assumed to be (04 06) e simulation resultis shown in Figure 5(h) e evolution results indicate thatwhen Ci is small the interaction behavior of both gameplayers will eventually evolve to the combined strategyldquoshare sharerdquo However as the parameter Ci increases andreaches a certain level (eg Ci 005 or 006) the interactionbehaviors of both game players will abandon the ldquosharerdquostrategy and eventually choose the combined strategy ldquonotshare not sharerdquo which the project leader does not seek

535 Inuence of Knowledge Stock By setting C1 005 andC2 005 the value of knowledge stock (Ki) increases from

16000 to 22000 with a step length of 2000 keeping the othermodel variables the same as the base valuese initial point is(04 06) We then studied the impact of the numericalvariation of the knowledge stock on the interaction behaviorsof both game players From the simulation results as shownin Figures 5(i) and 5(j) we can observe that the larger thevalue of Ki is the faster the evolution results converge to 1at is with the increase of Ki the probability of choosingcombined strategy ldquoshare sharerdquo improves gradually

536 Inuence of Degree of Knowledge-Sharing Finally theinuence of the degree of knowledge-sharing (ηi) on theinteraction behavior of both game players was analyzed Bysetting C1 005 and C2 005 and keeping the other modelvariables the same as the base values the values of η1 increasefrom 02 to 08 with a step length of 02 From the simulationresults as shown in Figures 5(k) and 5(l) we can see thatwith the increase of degree of knowledge-sharing the rate ofevolution converging to 1 improves gradually at isimproving the value of ηi benets the probability of adoptingcombined strategy ldquoshare sharerdquo for both game players

54 Result Discussion We validated the novel evolutionarygamemodel established in this study by conducting detailed

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)

Figure 3 Evolutionary results of x inuenced by the dierent initial value of y where (a) y 02 (b) y 05 and (c) y 08

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)

Figure 4 Evolutionary results of y inuenced by the dierent initial value of x where (a) x 02 (b) x 05 and (c) x 08


0

01

02

03

04

05

06

07

08

09

1

x (μ1 = 02)y (μ1 = 02)x (μ1 = 04)y (μ1 = 04)x (μ1 = 06)y (μ1 = 06)x (μ1 = 08)y (μ1 = 08)

2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1

x (α1 = 02)y (α1 = 02)x (α1 = 04)y (α1 = 04)x (α1 = 06)y (α1 = 06)x (α1 = 08)y (α1 = 08)

2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1

x (β1 = 02)y (β1 = 02)x (β1 = 04)y (β1 = 04)x (β1 = 06)y (β1 = 06)x (β1 = 08)y (β1 = 08)

2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1

x (C1 = 03)y (C1 = 03)x (C1 = 04)y (C1 = 04)x (C1 = 05)y (C1 = 05)x (C1 = 06)y (C1 = 06)

5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1

x (K1 = 16000)y (K1 = 16000)x (K1 = 18000)y (K1 = 18000)x (K1 = 20000)y (K1 = 20000)x (K1 = 22000)y (K1 = 22000)

2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1

x (η1 = 02)y (η1 = 02)x (η1 = 04)y (η1 = 04)x (η1 = 06)y (η1 = 06)x (η1 = 08)y (η1 = 08)

5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)

Figure 5 Inuences of the evolutionary game model parameters Inuence of (a) parameter μ1 (b) parameter μ2 (c) parameter α1(d) parameter α2 (e) parameter β1 (f ) parameter β2 (g) of parameter C1 (h) parameter C2 (i) parameter K1 (j) parameter K2 (k)parameter η1 and (l) parameter η2


computer simulation experiments using MATLAB Wemainly focused on the interaction behavior of knowledge-sharing Our simulation results successfully captured theinteraction behaviors between two groups accurately andthus predicted the evolution tendency with respect to timeDifferent from the classical game model that requires ab-solute rationality and complete information the evolu-tionary game theory relaxes the two limitations therebymaking the simulation results conform to actual situationse simulation results reveal some valuable findings relevantto integrated project team leaders allowing them to pro-foundly understand knowledge-sharing as well as how toimprove overall project performance

(1) e evolution path results show that as long as thereis a game player whose knowledge-sharing cost isgreater than the synergetic benefits as shown inFigures 2(b)ndash2(d) both game players will eventuallyabandon knowledge-sharing thereby leading tofailure of cooperation Knowledge-sharing is oftenregarded as a function of cost and benefit [102]However synergetic benefit generated from thecollaboration is excluded when analyzing the in-fluence of benefit on knowledge-sharing behavioris may lead to an unreliable result Compared toprior studies we distinguish the types of direct andsynergetic benefits and demonstrate that synergeticrather than direct benefit plays a key role inknowledge-sharing

(2) e evolutionary tendency of interaction behaviorsis affected greatly by the initial proportion ofchoosing knowledge-sharing strategy as shown inFigures 3 and 4 If both game players are at mod-erate- or high-effort level they will eventually chooseto share knowledge with each other Additionallyeven if one game player is at low-effort level it can bedriven by the other game player being at high-effortlevel to adopt the ldquosharerdquo strategy us the gameplayer with smaller proportion of preferring toknowledge-sharing will be positively influenced bythe other game player with higher proportion ofknowledge-sharing Ultimately both game playerschoose to share knowledge However both gameplayers will abandon the share strategy when they areall at low-effort level Although all kinds of criticalfactors including culture environment close re-lationship trust leadership behavior team membersattitude intrinsic and extrinsic motivation and so on[63 83 103ndash105] affect knowledge-sharing withinproject teams or organizations however the influ-ence of project team members structure on strategyinteraction behavior has never been discussed inprior studies is finding fixed the knowledge gapand provides a bright insight for project team leadersto improve performance

(3) Knowledge stock is proved an important indicatorfor learning organization e real purpose ofbuilding a learning organization in the knowledgeeconomy is to improve team membersrsquo knowledge

stock [106] However most studies only point outthe importance of knowledge stock lacking a deeperunderstanding of the influences of knowledge stockon knowledge-sharing mechanism within an orga-nization rough simulation results we found thatknowledge stock positively promote knowledge-sharing Although Knowledge integration couldincrease benefit [107] however most studies do notelaborate what kind of knowledge fusion generatesbenefit and its impact on knowledge-sharing be-haviore finding in this research suggests that onlyheterogeneous knowledge fusion brings about addedbenefit and reveals further that it has a positivecorrelation with knowledge-sharing behaviorA great number of studies have paid attention to therole of team membersrsquo willingness in knowledge-sharing behavior However most researchersmainly focus on how to improve willingness toshare knowledge from static perspective [108]wherein the data are collected only at a point oftime And thus exploring dynamically the re-lationship between team membersrsquo willingness andevolutionary trend of knowledge-sharing behavioris absent We get out of this dilemma and suc-cessfully reveal that degree of knowledge-sharingpositively affects knowledge-sharing from the dy-namic perspective e concept of knowledge ab-sorption capacity defined as the ability to identifydigest and utilize knowledge in the external envi-ronment is believed to affect companyrsquos competi-tive advantages greatly [109] However mostresearchers only study it at the organizational levellacking an explanation of influence of knowledgeabsorption capacity on knowledge-sharing behav-ior We introduce this concept at the individuallevel and explore the relationship between them andfound that knowledge absorption capacity has apositive correlation with knowledge-sharing be-havior within project teamSynergetic management is regarded as an importantpart and provides a theoretical fundament for or-ganizations to improve management efficiencythrough information sharing and complementarycooperation [110] Based on the extant literaturehowever influence of synergetic effect on knowl-edge-sharing is rarely involved which is a knowl-edge gap need to address rough our numericalsimulation results we found that raising the valueof synergetic effect coefficient can improve theprobability of knowledge-sharing under certainrestrictions e convergence rate depends on thesize of the parameters at is the larger the valuesof the parameter are the faster the interactionbehavior converges to the combined ldquoshare sharerdquostrategy

(4) e knowledge-sharing cost coefficient plays anegative role in promoting knowledge-sharing esimulation results as shown in Figures 5(g) and 5(h)


indicate that with the increase of the knowledge-sharing cost coefficient the evolution rate of in-teraction behavior converging to 1 slows down andmoves away from the combined strategy ldquosharesharerdquo When the value of the knowledge-sharingcost coefficient reaches a certain level the evolutiontrend reverses en the larger the cost coefficient isthe faster the evolution results converge to 0 at isboth game players will increasingly give up theirmotivation for knowledge-sharing thereby exhibit-ing passive non-cooperative attitude By comparisonof the simulation results of the parameters above weobserve that a small variation in the parameter of theknowledge-sharing cost coefficient will change theevolutionary direction for example when C1 in-creases from 004 to 005 and the evolution resultsconverge from the combined strategy ldquoshare sharerdquoto the combined strategy ldquonot share not sharerdquo atis the two game players abandon knowledge-shar-ing leading to the death of cooperation erefore itcan be concluded that the evolutionary tendency ofthe interaction behavior is most sensitive to theparameter of knowledge-sharing cost coefficientDespite analysis of influences of cost on knowledge-sharing behavior in prior studies to our bestknowledge very few studies assess the degree ofimportance of cost in knowledge management fieldis is a very important problem need to addressedbecause it can help project managers allocate rea-sonable resources to improve performance withinthe project team and reduce unnecessary waste Ourresearch work can fix this knowledge gap and pro-vide valuable guidance for project manager to un-derstand deeply the key role of cost in knowledgemanagement

6 Conclusion and Implications

e IPD method is an emerging research domain that hasreceived extensive attention from stakeholders in the AECindustry Our goal is to study knowledge-sharing strategiesand the evolutionary trend of strategic interaction amongintegrated project team members An innovative modelwhich takes benefits and costs into consideration isestablished by taking advantage of the evolutionary gametheory to capture the dynamic behavior of knowledge-sharing and thus reveal the cooperation tendency amonggame players Moreover simulation experiments are carriedto verify the results derived from the theoretical analysis tostudy the influence of model variables on the evolution trendof interaction behavior Our research findings indicate thatthe knowledge stock heterogeneous knowledge proportiondegree of knowledge-sharing knowledge absorption co-efficient and synergetic effect coefficient all contribute toknowledge-sharing A strong negative correlation was foundbetween knowledge-sharing cost coefficient and strategyinteraction From the theoretical analysis and simulationresults we propose that the following implications can be

obtained for integrated project managers to improve thecooperation and overall project performance

(1) For the integrated project managers the priority is tomake sure the knowledge-sharing cost is not greaterthan the indirect benefits or it will directly lead to afailure of cooperation Further the cost-controlability should be improved because the evolutionresult is most affected by the parameter Ci accordingto the numerical simulation outcome In terms ofpresent application most completed integratedprojects lack a dynamic monitoring mechanism ofcost leading to low efficiency of cost control Spe-cifically due to the defects of traditional constructiontechnology dynamic monitoring cannot fully coverthe construction project resulting in the lack of acost control plan is adversely affects projectmanagement and even delays the progress of thewhole project e real-time cost control needs to bebased on a big data model that includes quantity ofinformation at all stages of the project As anemerging technology BIM can act as an in-formation-integrated platform to satisfy the in-formation requirements of the cost dynamicalcontrol erefore perfect cost dynamic monitoringmechanism combined with BIM technology shouldbe established for the project manager so as toimprove the cost-control ability

(2) Party selection for IPD is important for the ownerAccording to the simulation results we now knowthat improving the parameters of the knowledgestock heterogeneous knowledge proportion andknowledge absorption coefficient will promote theknowledge-sharing between both game playerserefore the owner should consider the stake-holdersrsquo experience of implementing IPD projects toensure they have rich knowledge stock Further it isimportant to investigate whether the stakeholdershave operated different types of projects so as toensure significant knowledge difference amongthem Knowledge absorption capability is closelyrelated to stakeholdersrsquo characteristics Based on thetheory of learning organizations compared withtraditional organizations the team members inlearning organization have better learning ability andknowledge absorption ability Learning organiza-tions are an important feature for a successful en-terprise and has been increasingly by all walks of lifeerefore the owner should develop an index systemset to judge and select the stakeholders who have theoptimal knowledge absorption capability

(3) More attention should be paid to knowledge-sharingwillingness as it plays a positive role in promotingknowledge-sharing For project managers it is es-sential to establish rights and a liability distributionmechanism in order to further arouse the intrinsicmotivation of each project team member Accordingthe current practice of IPD serious lack of trust-


based relationship and inappropriate rewards hinderthe willingness for knowledge-sharing amongstakeholders Hence the project manager shouldspeed up the development of trust and incentivemechanisms to remove these barriers

(4) Synergetic capability building is another importantwork for the project managers As discussed in Section41 the synergetic coefficient is subject to three factorssynergy organization synergy culture and synergytechnology Project managers in synergy organiza-tion aspect should carry out broad business processreengineering to adjust to the requirements of IPDand BIM is would allow free flow of informationand knowledge within the project team Regardingsynergy culture project managers should convenecoordination meetings to align goals motivation andthe ideal of value and thus eliminate conflicts causedby cultural differences among different stakeholdersFinally regarding synergy technology project man-agers should require stakeholders to adopt a buildinginformation platform that supports the open in-formation standards eg Industry Foundation Class(IFC) to meet the requirements of information ex-change throughout the life cycle of a project esynergetic effect can be improved through thesemeasures

Nevertheless there are some limitations to this studythat should be investigated in the follow-up research Firstthe data of projects adopting IPD are difficult to obtainHence our findings were mainly derived through numericalsimulation is may neglect some actual and valuable in-formation that is difficult to obtain in literature reviews Infuture research work an empirical analysis could be con-ducted to further supplement the research results Secondwe did not consider the risk factors in the evolutionary gamemodel because the risks vary greatly depending on the typesof projects us we seek to investigate the risk types indetail and establish a risk knowledge base to offer morevaluable suggestions

Despite these limitations in this study our researchfindings still have important contributions in both theory andpractice eoretically our research work can provide theevolutionary mechanism behind knowledge-sharing strate-gies and broaden our understanding of relationship betweenproject performance and knowledge-sharing Practically weoffer valuable references for integrated project team leaders tobroaden their understandings of how cooperation is influ-enced by different factor and thus taking correspondingmeasures to improve project team performance

Data Availability

e data used to support the findings of this study are in-cluded within the article

Conflicts of Interest

e authors declare that they have no conflicts of interest

Acknowledgments

e authors appreciate those experts who assisted them inthe semistructured interviews and also gratefully acknowl-edge the experts for their constructive comments andsuggestions is work was supported by the ldquoFundamentalResearch Funds for the Central Universitiesrdquo (DUT18JC44)

References

[1] H A Mesa K R Molenaar and L F Alarcon ldquoExploringperformance of the integrated project delivery process oncomplex building projectsrdquo International Journal of ProjectManagement vol 34 no 7 pp 1089ndash1101 2016

[2] D R Hale P P Shrestha G E Gibson andG C Migliaccio ldquoEmpirical comparison of designbuildand designbidbuild project delivery methodsrdquo Journal ofConstruction Engineering and Management vol 135 no 7pp 579ndash587 2009

[3] W A Lichtig ldquoe integrated agreement on project per-formancerdquo Construction Lawyer vol 26 no 3 pp 1ndash8 2006

[4] P Mitropoulos and C B Tatum ldquoManagement-driven in-tegrationrdquo Journal of Management in Engineering vol 16no 1 pp 48ndash58 2000

[5] D J Kelly Investigating the Relationships of Project Per-formance Measures with the Use of Building InformationModeling (BIM) and Integrated Project Delivery (IPD)Eastern Michigan University Ypsilanti MI USA 2015

[6] B Zareie and N J Navimipour ldquoe effect of electroniclearning systems on the employeersquos commitmentrdquo In-ternational Journal of Management Education vol 14 no 2pp 167ndash175 2016

[7] Z Soltani and N J Navimipour ldquoCustomer relationshipmanagement mechanisms a systematic review of the stateof the art literature and recommendations for future re-searchrdquo Computers in Human Behavior vol 61 pp 667ndash688 2016

[8] M Alvesson ldquoSocial identity and the problem of loyalty inknowledge-intensive companiesrdquo Journal of ManagementStudies vol 37 no 8 pp 1101ndash1124 2010

[9] N J Navimipour A M Rahmani A H Navin andM Hosseinzadeh ldquoExpert cloud a cloud-based frameworkto share the knowledge and skills of human resourcesrdquoComputers in Human Behavior vol 46 pp 57ndash74 2015

[10] X Zhang P O D Pablos and Z Zhou ldquoEffect of knowledgesharing visibility on incentive-based relationship in elec-tronic knowledge management systems an empirical in-vestigationrdquo Computers in Human Behavior vol 29 no 2pp 307ndash313 2013

[11] E Ochieng and L Hughes ldquoManaging project complexity inconstruction projects the way forwardrdquo Architectural En-gineering Technology vol 2 no 1 pp 1-2 2013

[12] K I Gidado ldquoProject complexity the focal point of con-struction production planningrdquo Construction Managementand Economics vol 14 no 3 pp 213ndash225 1996

[13] A L Aryani M Suzila K Narimah and M S FathildquoBuilding information modeling (BIM) application inMalaysian construction industryrdquo International Journal ofConstruction Engineering and Management vol 2 no 4pp 1ndash6 2013

[14] W S Matthew ldquoProject alliancing a relational contractingmechanism for dynamic projectsrdquo Lean Construction Jour-nal vol 2 no 1 2005


[15] C omsen and J Darrington Managing Integrated ProjectDelivery Construction Management Association of America(CMAA) McLean VA USA 2010

[16] D C Kent and B Becerik-Gerber ldquoUnderstanding con-struction industry experience and attitudes toward integratedproject deliveryrdquo Journal of Construction Engineering andManagement vol 136 no 8 pp 815ndash825 2010

[17] O Mathhews and G A Howell ldquoAn integrated projectdelivery an example of relational contractingrdquo Lean Con-struction Journal vol 2 no 1 pp 46ndash61 2005

[18] AIA (American Institute of Architects) IPD Case StudiesUniversity of Minnesota Minneapolis MN USA 2012

[19] S M Allameh J K Pool A Jaberi and F M SoveinildquoDeveloping a model for examining the effect of tacit andexplicit knowledge-sharing on organizational perfor-mance based on EFQM approachrdquo Journal of Science andTechnology Policy Management vol 5 no 3 pp 265ndash2802014

[20] J Liebowitz and I Megbolugbe ldquoA set of frameworks to aidthe project manager in conceptualizing and implementingknowledge management initiativesrdquo International Journal ofProject Management vol 21 no 3 pp 189ndash198 2003

[21] Z Ma D Zhang and J Li ldquoA dedicated collaborationplatform for integrated project deliveryrdquo Automation inConstruction vol 86 pp 199ndash209 2018

[22] C Y Chang W J Pan and R Howard ldquoImpact of buildinginformation modeling implementation on the acceptance ofintegrated delivery systems structural equation modelinganalysisrdquo Journal of Construction Engineering and Man-agement vol 143 no 8 article 04017044 2017

[23] J Ma Z Ma and J Li ldquoAn IPD-based incentive mechanismto eliminate change orders in construction projects inChinardquo KSCE Journal of Civil Engineering vol 21 no 7pp 2538ndash2550 2017

[24] Q Wu A Research on the Influencing Path of Trust onManagement Performance of Integrated Project DeliveryTeam Tianjin University Tianjin China 2012

[25] A S Hanna ldquoBenchmark performance metrics for in-tegrated project deliveryrdquo Journal of Construction Engi-neering and Management vol 142 no 9 article 040160402016

[26] G Radaelli M Mura N Spiller and E Lettieri ldquoIntellectualcapital and knowledge sharing the mediating role oforganisational knowledge-sharing climaterdquo KnowledgeManagement Research amp Practice vol 9 no 4 pp 342ndash3522011

[27] M Polanyi ldquoTacit knowing its bearing on some problems ofphilosophyrdquo Review of Modern Physics vol 34 no 4pp 601ndash615 1962

[28] J-H Woo M J Clayton R E Johnson B E Flores andC Ellis ldquoDynamic knowledge map reusing expertsrsquo tacitknowledge in the AEC industryrdquo Automation in Construc-tion vol 13 no 2 pp 203ndash207 2004

[29] G Anand P T Ward and M V Tatikonda ldquoRole of explicitand tacit knowledge in six sigma projects an empiricalexamination of differential project successrdquo Journal of Op-erations Management vol 28 no 4 pp 303ndash315 2010

[30] M T Hansen N Nohria and T Tierney ldquoWhatrsquos yourstrategy for managing knowledgerdquo Harvard Business Re-view vol 77 no 2 p 106 1999

[31] P K Couchman and L Fulop ldquoExamining partner expe-rience in cross-sector collaborative projects focused on thecommercialization of RampDrdquo Innovation Organization ampManagement vol 11 no 1 pp 85ndash103 2009

[32] I Nonaka ldquoA dynamic theory of organizational knowledgecreationrdquoOrganization Science vol 5 no 1 pp 14ndash37 1994

[33] Z Wu Z Zhong W Liu W Zhang and X Yao ldquoCon-struction of multi-layer semantic database based onmpeg-7rdquo Computer Science vol 39 pp 532ndash535 2012

[34] S Fernie S D Green S J Weller and R NewcombeldquoKnowledge sharing context confusion and controversyrdquoInternational Journal of Project Management vol 21 no 3pp 177ndash187 2003

[35] P Hendriks ldquoWhy share knowledgee influence of ICTonthe motivation for knowledge sharingrdquo Knowledge andprocess management vol 6 no 2 pp 91ndash100 1999

[36] R R A Issa and J Haddad ldquoPerceptions of the impacts oforganizational culture and information technology onknowledge sharing in constructionrdquo Construction In-novation vol 8 no 3 pp 182ndash201 2008

[37] J N Cummings ldquoWork groups structural diversity andknowledge sharing in a global organizationrdquo ManagementScience vol 50 no 3 pp 352ndash364 2004

[38] N J Navimipour and Y Charband ldquoKnowledge-sharingmechanism and techniques in project teams literature re-view current trendsrdquo Computers in Human Behaviorvol 62 pp 730ndash742 2016

[39] J V Annadatha Sociocultural Factors and Knowledge-sharing Behaviors in Virtual Project Teams Robert MirrisUniversity Moon PA USA 2012

[40] F Farajpour M T Taghavifard and A Yousefli ldquoIn-formation sharing assessment in supply chain hierarchicalfuzzy rule-based systemrdquo Journal of Information amp Knowl-edge Management vol 17 no 1 article 1850002 2018

[41] N Khatri G D Brown and L L Hicks ldquoFrom a blameculture to a just culture in health carerdquo Health Care Man-agement Review vol 34 no 4 pp 312ndash322 2009

[42] M Chiregi and N J Navimipour ldquoA new method for trustand reputation evaluation in the cloud environments usingthe recommendations of opinion leadersrsquo entities and re-moving the effect of troll entitiesrdquo Computers in HumanBehavior vol 60 pp 280ndash292 2016

[43] M Yuan X Zhang Z Chen D R Vogel and X ChuldquoAntecedents of coordination effectiveness of software de-veloper dyads from interacting teams an empirical in-vestigationrdquo IEEE Transactions on Engineering Managementvol 56 no 3 pp 494ndash507 2009

[44] L Y Zhang and J He ldquoCritical factors affecting tacit-knowledge-sharing within the integrated project teamrdquoJournal of Management in Engineering vol 32 no 2 article04015045 2016

[45] B Zareie N J Navimipour and Y Charband ldquoKnowledge-sharing mechanisms and techniques in project teams lit-erature review classification and current trendsrdquo Computersin Human Behavior vol 62 pp 730ndash742 2016

[46] N J Navimipour and B Zareie ldquoA model for assessing theimpact of e-learning system on employeesrsquo satisfactionrdquoComputers in Human Behavior vol 53 pp 475ndash485 2015

[47] Y Charband and N J Navimipour ldquoKnowledge sharingmechanisms in the educationrdquo Kybernetes vol 47 no 7pp 1456ndash1490 2018

[48] N J Navimipour A H Navin and A M RahmanildquoBehavioral modeling and automated verification of acloud-based framework to share the knowledge and skillsof human resourcesrdquo Computers in Industry vol 68pp 65ndash77 2015

[49] P Fouladi and N J Navimipour ldquoHuman resources rankingin a cloud-based knowledge sharing framework using the


quality control criteriardquo Kybernetes vol 46 no 5pp 876ndash892 2017

[50] Z H Yang and Q F Shi ldquoOpportunity cost analysis of tacitknowledge-sharing in colleges and universitiesrdquo Science andTechnology Management Research vol 27 no 5 pp 234ndash236 2007

[51] A Kankanhalli B C Y Tan and K K Wei ldquoContributingknowledge to electronic knowledge repositories an empir-ical investigationrdquoMIS Quarterly vol 29 no 1 pp 113ndash1432005

[52] S Karkoulian L C Messarra and R Mccarthy ldquoe in-triguing art of knowledge management and its relation tolearning organizationsrdquo Journal of Knowledge Managementvol 17 no 4 pp 511ndash526 2013

[53] A K Gupta and V Govindarajan ldquoKnowledge manage-mentrsquos social dimension lessons from nucor steelrdquo SloanManagement Review vol 42 no 1 pp 71ndash80 2000

[54] T J Jewels Motivators and Inhibitors to Knowledge-Sharingin IT Teams Queensland University of Technology BrisbaneAustralia 2006

[55] L P Liu ldquoCost benefit and incentive mechanism ofknowledge-sharing in enterprisesrdquo Commercial Researchvol 8 pp 98ndash101 2006

[56] N J Navimipour and Z Soltani ldquoe impact of electronicenvironmental knowledge on the environmental behaviorsof peoplerdquo Computers in Human Behavior vol 55pp 1052ndash1066 2016

[57] P Zhang and F F Ng ldquoAttitude toward knowledge sharingin construction teamsrdquo Industrial Management amp DataSystems vol 112 no 9 pp 1326ndash1347 2012

[58] W Teerajetgul and C Charoenngam ldquoFactors inducingknowledge creation empirical evidence from ai con-struction projectsrdquo Engineering Construction and Archi-tectural Management vol 13 no 6 pp 584ndash599 2006

[59] T Cooke ldquoCan knowledge sharing mitigate the effect ofconstruction project complexityrdquo Construction Innovationvol 13 no 1 pp 5ndash9 2013

[60] F L Ribeiro ldquoEnhancing knowledge management in con-struction firmsrdquo Construction Innovation vol 9 no 3pp 268ndash284 2009

[61] B Xia and A P C Chan ldquoMeasuring complexity forbuilding projects a delphi studyrdquo Engineering Construc-tion and Architectural Management vol 19 no 1 pp 7ndash242012

[62] H S Robinson P M Carrillo C J Anumba and A M Al-Ghassani ldquoKnowledge management practices in largeconstruction organisationsrdquo Engineering Construction andArchitectural Management vol 12 no 5 pp 431ndash4452005

[63] G D Ni Q B Cui L H Sang W S Wang and D C XialdquoKnowledge-sharing culture project-team interaction andknowledge-sharing performance among project membersrdquoJournal of Management in Engineering vol 34 no 2 article04017065 2018

[64] I Ruuska and M Vartiainen ldquoCharacteristics of knowledgesharing communities in project organizationsrdquo InternationalJournal of Project Management vol 23 no 5 pp 374ndash3792005

[65] H Doloi K C Iyer and A Sawhney ldquoStructural equationmodel for assessing impacts of contractorrsquos performance onproject successrdquo International Journal of Project Manage-ment vol 29 no 6 pp 687ndash695 2011

[66] H Baumgartner and C Homburg ldquoApplications of struc-tural equation modeling in marketing and consumer

research a reviewrdquo International Journal of Research inMarketing vol 13 no 2 pp 139ndash161 1996

[67] N Cliff ldquoSome cautions concerning the application of causalmodeling methodsrdquo Multivariate Behavioral Researchvol 18 no 1 pp 115ndash126 1983

[68] R C MacCallum and J T Austin ldquoApplications of structuralequation modeling in psychological researchrdquo Annual Re-view of Psychology vol 51 no 1 pp 201ndash226 2000

[69] B Xiong M Skitmore and B Xia ldquoA critical review ofstructural equation modeling applications in constructionresearchrdquoAutomation in Construction vol 49 pp 59ndash70 2015

[70] S J Breckler ldquoApplications of covariance structure modelingin psychology cause for concernrdquo Psychological Bulletinvol 107 no 2 pp 260ndash273 1990

[71] K Nusair and N Hua ldquoComparative assessment of struc-tural equation modeling and multiple regression researchmethodologies E-commerce contextrdquo Tourism Manage-ment vol 31 no 3 pp 314ndash324 2010

[72] D K Yoo ldquoInnovation its relationships with a knowledge-sharing climate and interdisciplinary knowledge integrationin cross-functional project teamsrdquo in Proceedings of theHawaii International Conference on System Science IEEEHawaii USA 2015

[73] N Siggel ldquoPersuasion with case studiesrdquo Academy ofManagement Journal vol 50 pp 20ndash24 2007

[74] R B Mcevily ldquoNetwork structure and knowledge transferthe effects of cohesion and rangerdquo Administrative ScienceQuarterly vol 48 no 2 pp 240ndash267 2003

[75] P L Luo and X H Yin ldquoIncentive framework for knowl-edge-sharing based on benefit game modelrdquo RampD Man-agement vol 21 no 2 pp 24ndash29 2009

[76] D-F Li ldquoLinear programming approach to solve interval-valued matrix gamesrdquo Omega vol 39 no 6 pp 655ndash6662011

[77] CWang LWang J Wang S Sun and C Xia ldquoInferring thereputation enhances the cooperation in the public goodsgame on interdependent latticesrdquo Applied Mathematics andComputation vol 293 pp 18ndash29 2017

[78] Y Chen S Ding H Zheng Y Zhang and S Yang ldquoEx-ploring diffusion strategies for mHealth promotion usingevolutionary game modelrdquo Applied Mathematics andComputation vol 336 pp 148ndash161 2018

[79] J Jin J Zhang and Q H Zhang ldquoEvolutionary game theoryandmodeling of economic behaviorrdquoDe Economist vol 146no 1 pp 59ndash89 1998

[80] D Li JMa Z Tian andH Zhu ldquoAn evolutionary game for thediffusion of rumor in complex networksrdquo Physica A StatisticalMechanics and its Applications vol 433 pp 51ndash58 2015

[81] D Friedman ldquoEvolutionary games in economicsrdquo Econo-metrica vol 59 no 3 pp 637ndash666 1991

[82] D Friedman and K C Fung ldquoInternational trade and theinternal organization of firms an evolutionary approachrdquoJournal of International Economics vol 41 no 2 pp 113ndash137 1996

[83] J H Humphreys Z Ma and L Qi ldquoKnowledge sharing inChinese construction project teams and its affecting factorsan empirical studyrdquo Chinese Management Studies vol 2no 2 pp 97ndash108 2008

[84] B Jones ldquoIntegrated project delivery (IPD) for maximizingdesign and construction considerations regarding sustain-abilityrdquo Procedia Engineering vol 95 pp 528ndash538 2014

[85] S Wang and R A Noe ldquoKnowledge sharing a review anddirections for future researchrdquo Human Resource Manage-ment Review vol 20 no 2 pp 115ndash131 2010


[86] C-M Chiu M-H Hsu and E T G Wang ldquoUnderstandingknowledge sharing in virtual communities an integration ofsocial capital and social cognitive theoriesrdquo Decision SupportSystems vol 42 no 3 pp 1872ndash1888 2006

[87] L Lu K Leung and P T Koch ldquoManagerial knowledgesharing the role of individual interpersonal and organi-zational factorsrdquo Management and Organization Reviewvol 2 no 1 pp 15ndash41 2006

[88] T Dewett ldquoLinking intrinsic motivation risk taking andemployee creativity in an RampD environmentrdquo RampD Man-agement vol 37 no 3 pp 197ndash208 2007

[89] R Eisenberger and J Aselage ldquoIncremental effects of rewardon experienced performance pressure positive outcomes forintrinsic interest and creativityrdquo Journal of OrganizationalBehavior vol 30 no 1 pp 95ndash117 2009

[90] M Baer and G R Oldham ldquoe curvilinear relation betweenexperienced creative time pressure and creativity moder-ating effects of openness to experience and support forcreativityrdquo Journal of Applied Psychology vol 91 no 4pp 963ndash970 2006

[91] Q Huang R M Davison and J Gu ldquoe impact of trustguanxi orientation and face on the intention of Chineseemployees and managers to engage in peer-to-peer tacit andexplicit knowledge sharingrdquo Information Systems Journalvol 21 no 6 pp 557ndash577 2011

[92] J G Shi L G Lin and D Z Tang ldquoKnowledge-sharingincentive model of project team members based onknowledge complementary effectrdquo Science and TechnologyManagement Research vol 24 pp 129ndash135 2014

[93] M Wang and C Shao ldquoSpecial knowledge sharing incentivemechanism for two clients with complementary knowledgea principal-agent perspectiverdquo Expert Systems with Appli-cations vol 39 no 3 pp 3153ndash3161 2012

[94] P Lee N Gillespie LMann and AWearing ldquoLeadership andtrust their effect on knowledge sharing and team perfor-mancerdquo Management Learning vol 41 no 4 pp 473ndash4912010

[95] X Z Jia G B Sun and Z R Ye ldquoKnowledge comple-mentarity and the limit to specialization an new classicalmodelrdquo Journal of Technical Economics amp Managementvol 4 p 135 2010

[96] X Z Jia and B Z Ye ldquoResearch on innovation based onknowledge complementarity a profit modelrdquo Economy andManagement vol 24 no 10 pp 33ndash38 2010

[97] Z J Chen and X Liu ldquoEstablish evaluation model for parentndashsubsidiary companies synergy effectsrdquo Economic Manage-ment vol 32 no 10 pp 51ndash56 2010

[98] P V Montoya R S Zarate and L A G Martın ldquoDoes thetechnological sourcing decision matter Evidence fromSpanish panel datardquo RampD Management vol 37 no 2pp 161ndash172 2007

[99] S X Zeng X M Xie and C M Tam ldquoRelationship betweencooperation networks and innovation performance ofSMEsrdquo Technovation vol 30 no 3 pp 181ndash194 2010

[100] WM Cohen and D A Levinthal ldquoAbsorptive capacity a newperspective on learning and innovationrdquo Strategic Learning ina Knowledge Economy vol 35 no 1 pp 39ndash67 1990

[101] V Scuotto M D Giudice and E G Carayannis ldquoe effectof social networking sites and absorptive capacity on SMESrsquoinnovation performancerdquo Journal of Technology Transfervol 42 no 2 pp 409ndash424 2017

[102] T JewelsMotivators and Inhibitors to Knowledge Sharing inIT Project Teams Queensland University of TechnologyBrisbane Australia 2006

[103] C H Huang and I C Huang ldquoe moderating effect of co-workersrsquo reactions on social ties and knowledge sharing inwork teamsrdquo International Journal of Learning and In-tellectual Capital vol 6 no 2 pp 156ndash169 2008

[104] Z K Ding and F F Ng ldquoKnowledge sharing among ar-chitects in a project design team an empirical test of theoryof reasoned action in Chinardquo Chinese Management Studiesvol 2 no 2 pp 130ndash142 2009

[105] L Y Qi K Y Wang and Z Z Ma ldquoAntecedent factors ofknowledge sharing in project teams evidence from chineseconstruction sectorrdquo in Proceedings of the InternationalConference on E-Businessamp E-Government IEEE GuangzhouChina May 2010

[106] Y Liu and K J Zeng ldquoResearch on knowledge stock oflearning organization based on system dynamicsrdquo Scienceand Technology Management Research vol 38 no 13pp 182ndash191 2018

[107] B P Cozzarin and J C Percival ldquoComplementarities be-tween organizational strategies and innovationrdquo Economicsof Innovation and New Technology vol 15 no 3 pp 195ndash217 2006

[108] A Wiewiora B Trigunarsyah G Murphy and V CoffeyldquoOrganizational culture and willingness to share knowledgea competing values perspective in Australian contextrdquo In-ternational Journal of Project Management vol 31 no 8pp 1163ndash1174 2013

[109] A Fosfuri and J Tribo ldquoExploring the antecedents of po-tential absorptive capacity and its impact on innovationperformancerdquo Omega vol 36 no 2 pp 173ndash187 2008

[110] V D Mathijs ldquoInterfirm networks in periods of techno-logical turbulence and stabilityrdquo Research Policy vol 43no 10 pp 1666ndash1680 2014


International Journal of

AerospaceEngineeringHindawiwwwhindawicom Volume 2018

RoboticsJournal of

Hindawiwwwhindawicom Volume 2018


Active and Passive Electronic Components

VLSI Design



Shock and Vibration


Civil EngineeringAdvances in

Acoustics and VibrationAdvances in



Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawiwwwhindawicom

Volume 2018

Hindawi Publishing Corporation httpwwwhindawicom Volume 2013Hindawiwwwhindawicom

The Scientific World Journal

Volume 2018

Control Scienceand Engineering

Journal of



Journal ofEngineeringVolume 2018

SensorsJournal of



RotatingMachinery


Modelling ampSimulationin EngineeringHindawiwwwhindawicom Volume 2018


Chemical EngineeringInternational Journal of Antennas and

Propagation




Navigation and Observation


Hindawi

wwwhindawicom Volume 2018

Advances in

Multimedia

Submit your manuscripts atwwwhindawicom

structures and practices into a single process whereinproject participants including the owner designer generalcontractor and special subcontractor can work collabora-tively at the early design stage IPD increases values to theowner decreases waste and maximizes project benefits byutilizing multiparty contract to share benefits and risksHowever according to the investigation carried by Kelly [5]the project performance that adopted the integrated projectdelivery approach did not achieve desired results in practiceOne of the significant reasons is that the cooperative be-havior among integrated project teammembers is hamperedto some extent thereby leading to an overall performancedecline e failure in performance improvements thusmakes it imperative for research to address this issue in orderto effectively facilitate better cooperation

In a knowledge-based economic society organizationalor a project team competitiveness arise from intangiblerather than tangible resources [6] In this economic contextknowledge is defined as the skills objectives and experiencesthat together create a framework for assessing and makinguse of information exchanged in an explicit or implicitmanner [7] Most researchers have argued that the com-petitiveness of an organization relies on the ability to createvaluable knowledge and share it with all project teammembers within this organization in order to infuse thisknowledge into products services and systems [8] us asuccessful knowledge-sharing strategy requires effectivecooperative relationships among integrated project teammembers is is critical for facilitating cooperation andproject performance [9] Without high-level knowledge-sharing a project team leader becomes unable to integrate allthe advantages that stakeholders offer (eg rich experiencestrong organization management skills and information-processing capability) for efficiently and effectively exe-cuting large and complex projects [10]

Cooperation is a common phenomenon in nature andhuman society It is called the third evolutionary principlebesides selection and mutation As an internal driving forceof species evolution and social development it promotes theformation of complex life systems and social organizationsFrom the perspective of individual survival and develop-ment cooperative behaviors will reduce their own benefitsand weaken their competitive advantages From the per-spective of society or team development cooperative be-havior is beneficial to the overall interest although it mayreduce individual interest In construction industry the bestpractice is to realize the overall performance and balanceindividual benefit In traditional project deliveries imple-mentation process is poorly integrated and information isasymmetric and the first consideration of each participant isnot to maximize the benefits of the overall project but tomaximize individual interest and avoid risks which usuallyleads to an overall project significant loss Hence it is es-sential to adopt the idea of IPD to solve this dilemma To thisend we should first capture the dynamic behavior of beingcooperative and make clear how cooperative behaviorspreads and maintains stability in an IPD team namelyevolutionary mechanism from the perspective of knowl-edge-sharing to provide theoretical foundation for

controlling and optimizing cooperative behavior in an IPDteam

In the construction industry the best practice is to re-alize the overall performance while balancing individualbenefits e implementation process in traditional projectdelivery is poorly integrated and crucial information isasymmetric Often each participant does not prioritizemaximization of the projectrsquos benefits Instead she or hechooses to maximize individual interest and avoid riskleading to significant project lossesese counterproductivebehaviors make it necessary to adopt IPD as a solution Tothis end we first capture the dynamic behavior of co-operation and clarify how cooperative behaviors spreadwithin and support the stability of an IPD team that is weidentify the evolutionary mechanism We then adopt theperspective of knowledge-sharing to provide a theoreticalfoundation for controlling and optimizing cooperative be-haviors in the IPD team

Despite the importance of knowledge-sharing in co-operative behaviors of the integrated project team very fewstudies have focused on improving cooperation from thisperspective to explain the dynamics of group behavior andhow cooperative behavior spreads and maintains stability inthe integrated project team namely evolutionary mecha-nism erefore our goal is to explore knowledge-sharingstrategies and reveal the evolutionary mechanism that un-derpins the integrated project teamWe analyze the resultingevolutionary tendency along with the variations in relevantfactors

Our study contributes to both theory and practice Firstwe propose an intrinsic evolutionary mechanism forknowledge-sharing strategies Second we provide valuablereferences for integrated project team leaders to broadentheir understanding of how cooperation is influenced bydifferent factors which would allow them to take corre-sponding measures to improve team performance

e remaining paper is organized as follows In Section 2we present the systematic overview of the extent literature onIPD approaches and knowledge-sharing In Section 3 weexplore the reasons for employing evolutionary game modelIn Section 4 we propose a novel evolutionary game model tocapture the dynamic behavior of knowledge-sharing InSection 5 we conduct numerical simulations to validate ourresults of theoretical analysis and discuss the influences ofmodel parameters on the evolutionary tendency of the in-teraction behavior In the last section we conclude the paperand present the implication of our work

2 Literature Review

21 Integrated Project Delivery Compared with other in-dustries the architecture engineering and construction(AEC) industry lacks a reliable reputation as a leading actorin quality productivity and time and cost managementbecause of its increasing complexity and multidisciplinarynature Most researchers have claimed that the projectsuccess largely depends on the complexity of a project whichthen directly influences the overall project performance [11]erefore it is important to recognize this feature Gidado

































3 Research Method







































1



(1)


1







fA dx

dt x E



(4)


2



(5)


2



E2 yEa2 +(1minusy)E




fB dy

dt y E



(8)


dx


dy


⎧⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎩

(9)


xlowast C2K2η2

β2 K2η2 + α2μ1K1η1( 1113857

ylowast C1K1η1

β1 K1η1 + α1μ2K2η2( 1113857

⎧⎪⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎪⎩

(10)


J

zfA

zx

zfA

zy

zfB

zx

zfB

zy

⎡⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣

⎤⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦

(11)

erefore


Group 1Group 2

B1(y) B2(1minus y)




zfA


zfA


zfB


zfB


(12)
































tr(J) 0




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS










B

A

D

C

E (xlowast ylowast)

(a)

B

A

D

C(b)

B

A

D

C

(c)

B

A

D

C

(d)







0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)








10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia

































3 Research Method







































1



(1)


1







fA dx

dt x E



(4)


2



(5)


2



E2 yEa2 +(1minusy)E




fB dy

dt y E



(8)


dx


dy


⎧⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎩

(9)


xlowast C2K2η2

β2 K2η2 + α2μ1K1η1( 1113857

ylowast C1K1η1

β1 K1η1 + α1μ2K2η2( 1113857

⎧⎪⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎪⎩

(10)


J

zfA

zx

zfA

zy

zfB

zx

zfB

zy

⎡⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣

⎤⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦

(11)

erefore


Group 1Group 2

B1(y) B2(1minus y)




zfA


zfA


zfB


zfB


(12)
































tr(J) 0




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS










B

A

D

C

E (xlowast ylowast)

(a)

B

A

D

C(b)

B

A

D

C

(c)

B

A

D

C

(d)







0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)








10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





































1



(1)


1







fA dx

dt x E



(4)


2



(5)


2



E2 yEa2 +(1minusy)E




fB dy

dt y E



(8)


dx


dy


⎧⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎩

(9)


xlowast C2K2η2

β2 K2η2 + α2μ1K1η1( 1113857

ylowast C1K1η1

β1 K1η1 + α1μ2K2η2( 1113857

⎧⎪⎪⎪⎪⎪⎪⎨

⎪⎪⎪⎪⎪⎪⎩

(10)


J

zfA

zx

zfA

zy

zfB

zx

zfB

zy

⎡⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢⎣

⎤⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥⎦

(11)

erefore


Group 1Group 2

B1(y) B2(1minus y)




zfA


zfA


zfB


zfB


(12)
































tr(J) 0




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS










B

A

D

C

E (xlowast ylowast)

(a)

B

A

D

C(b)

B

A

D

C

(c)

B

A

D

C

(d)







0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)








10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


zfA


zfA


zfB


zfB


(12)
































tr(J) 0




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS










B

A

D

C

E (xlowast ylowast)

(a)

B

A

D

C(b)

B

A

D

C

(c)

B

A

D

C

(d)







0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)








10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


















tr(J) 0




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS




(00) + minus ESS










B

A

D

C

E (xlowast ylowast)

(a)

B

A

D

C(b)

B

A

D

C

(c)

B

A

D

C

(d)







0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)








10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia








B

A

D

C

E (xlowast ylowast)

(a)

B

A

D

C(b)

B

A

D

C

(c)

B

A

D

C

(d)







0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)








10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia






0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(a)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(b)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(c)

0

01

02

03

04

05

06

07

08

09

1

Y

02 04 06 08 10X

(d)








10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia







10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25

Time30 35 40 45 50

x = 02x = 05x = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

x = 02x = 05x = 08

(c)


Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(a)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(b)

Time

10908

Prop

ortio

n 07060504030201

00 5 10 15 20 25 30 35 40 45 50

y = 02y = 05y = 08

(c)



0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(a)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(b)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(c)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(d)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(e)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(f )

Figure 5 Continued


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(g)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 40 450Time

(h)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(i)

0

01

02

03

04

05

06

07

08

09

1


2 4 6 8 10 12 14 16 18 200Time

(j)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(k)

0

01

02

03

04

05

06

07

08

09

1


5 10 15 20 25 30 35 400Time

(l)






















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia





















Data Availability




Acknowledgments


References























































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia









































































































RoboticsJournal of




VLSI Design



Shock and Vibration







Journal of



Volume 2018



Volume 2018


Journal of




SensorsJournal of



RotatingMachinery





Propagation






Hindawi


Advances in

Multimedia


exploreknowledge-sharingstrategyandevolutionary...

Documents