a gis based tool to determine optimal routes

UNLV Retrospective Theses & Dissertations

1-1-2005

A GIS based tool to determine optimal routes A GIS based tool to determine optimal routes

Rajeev Naroor University of Nevada, Las Vegas

Follow this and additional works at: https://digitalscholarship.unlv.edu/rtds

Repository Citation Repository Citation Naroor, Rajeev, "A GIS based tool to determine optimal routes" (2005). UNLV Retrospective Theses & Dissertations. 1892. http://dx.doi.org/10.25669/qt8d-kp3f

This Thesis is protected by copyright and/or related rights. It has been brought to you by Digital Scholarship@UNLV with permission from the rights-holder(s). You are free to use this Thesis in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s) directly, unless additional rights are indicated by a Creative Commons license in the record and/or on the work itself. This Thesis has been accepted for inclusion in UNLV Retrospective Theses & Dissertations by an authorized administrator of Digital Scholarship@UNLV. For more information, please contact [email protected].

http://library.unlv.edu/

http://library.unlv.edu/

https://digitalscholarship.unlv.edu/rtds

https://digitalscholarship.unlv.edu/rtds?utm_source=digitalscholarship.unlv.edu%2Frtds%2F1892&utm_medium=PDF&utm_campaign=PDFCoverPages

http://dx.doi.org/10.25669/qt8d-kp3f

mailto:[email protected]

A GIS BASED TOOL TO DETERMINE OPTIMAL ROUTES

by

Rajeev Naroor

Bachelor of Technology Kerala University

1998

A thesis submitted in partial fulfillment of the requirements for the

Master of Science Degree in Engineering Department of Civil and Environmental Engineering

Howard R. Hughes College of Engineering

Graduate College University of Nevada, Las Vegas

December 2005

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

UMI Number: 1435623

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy

submitted. Broken or indistinct print, colored or poor quality illustrations and

photographs, print bleed-through, substandard margins, and improper

alignment can adversely affect reproduction.

In the unlikely event that the author did not send a complete manuscript

and there are missing pages, these will be noted. Also, if unauthorized

copyright material had to be removed, a note will indicate the deletion.

UMIUMI Microform 1435623

Copyright 2006 by ProQuest Information and Learning Company.

All rights reserved. This microform edition is protected against

unauthorized copying under Title 17, United States Code.

ProQuest Information and Learning Company 300 North Zeeb Road

P.O. Box 1346 Ann Arbor, Ml 48106-1346


Thesis ApprovalThe Graduate College University of Nevada, Las Vegas

The Thesis prepared by

___________________ RAJEEV NAROOR

2 7 , SeptembeigQ 0 5

Entitled

A G IS BASED TOOL TO DETERMINE OPTIMAL ROHTRS

is approved in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE IN ENGINEERING

_____

Exam ination C om m ittee M em ber

Examination C om m ittee M em ber

/O f C ---Graduate CoUegelFaculty Representative

Exam ination C om m ittee Chair

Dean o f the G raduate College

11


ABSTRACT

A GIS Based Tool To Determine Optimal Routes

Rajeev Naroor

Shashi S. Nambisan, Ph.D., P.E., Examination Committee Chair Professor, Department of Civil & Environmental Engineering

University o f Nevada, Las Vegas

Travel in urban and suburban areas is an important activity in an individual’s daily

time budget. Each trip is preceded by a planning process for selecting a “desirable” or an

optimal route from the origin to destination. “Optimal” can take many forms, such as

shortest time, shortest distance, least total cost, or safest/most reliable path. Reliable paths

are of interest because of the impacts of congestion on travel tomes, and the resultant

variability in travel times. A model to help identify optimal travel paths hased on travel

time variability is developed in this research.

Congestion on streets is one of the prime reasons for the variability of travel time on

streets. In order to find an “optimal” route (based on minimizing expected travel time

variability) from a source to destination factors influencing congestion were studied. Data

were collected for these factors. Djikstra’s algoritlim was used to develop a route finding

application. The application considers factors related to recurrent and non recurrent

congestion and their impacts on estimated travel times. Data for such factors influencing

111


congestion were processed and used as inputs to the application. The application was

tested on the street network in the Las Vegas valley as a case study.

IV


TABLE OF CONTENTS

ABSTRACT.............................................................................................................................. iii

TABLE OF CONTENTS.......................................................................................................... v

LIST OF TABLES...................................................................................................................vii

LIST OF FIGURES........................... viii

ACKNOWLEDGEMENTS..................................................................................................... ix

CHAPTER 1 INTRODUCTION.................................. 1Motivation..............................................................................................................................1Research Objective, Tasks and Scope................................................................................3Study A rea............................................................................................................................ 4Layout o f the Thesis.............................................................................................................6

CHAPTER 2 LITERATURE REVIEW.............................................................................7Congestion.............................................................................................................................7Reliability and Vulnerability................................................ 10Network Analysis...............................................................................................................12GIS in Network Analysis....................................................................................... 15Application Development in Arc Info.............................................................................. 17

CHAPTER 3 METHODOLOGY..................................................................................... 21Travel Time Analysis......................................................................................................... 21

Recurring Congestion Analysis.................................................................................... 23Analysis o f non recurring delays..................................................................................25

Creating Buffers and Spatial Overlay...............................................................................25Delay due to Work Zones..................................................................................................28Dijkstra's Algorithm........................................................................................................... 29

Description of Dijkstra’s Algorithm............................................................. 31Application Development.................................................................................................32

CHAPTER 4 ANALYSIS, CASE STUDY & RESULTS.............................................. 34Recurrent Delay..................................................................................................................34Non Recurrent Delay......................................................................................................... 38Interface.............................................................................................................................. 49


CHAPTER 5 SUMMARY, CONCLUSIONS AND RECOMMENDATIONS..........60

REFERENCES.................................................... 62

VITA......................................................................................................................................... 66

VI


LIST OF TABLES

Table 1 Time Periods Over A Day...............................................................................24Table 2 Illustration Of Posted Speed And Travel Time............................................. 36Table 3 Estimated Prevailing Vehicle Speeds (Mph) During Different Time

Periods On A Sunday, M onday....................................................................37Table 4 Sample Crash Database...................................................................................40Table 5 Time Of Lane Blockage (Minutes).......................................... 41Table 6 Percentage Time Of Lane Blockage..............................................................41Table 7 Median Value Of Lane Blockage...................... 42Table 8 Lane Blockage Estimate For Street Network................................................43Table 9 Reduced Capacity (Vehicle Per Hour).......................................................... 45Table 10 Traffic Delay Due To Recurrent Congestion (In Seconds).........................46Table 11 Traffic Delay Due To Non Recurrent Congestion (In Seconds)................. 48Table 12 Summary Of Free Flow Route Result............................................................55Table 13 Summary Of Recurrent Delay Route............................................................56Table 14 Summary Of Non Recurrent Delay Route.................................................... 57Table 15 Summary Of Recurrent + Non Recurrent Delay Route............................... 58Table 16 Summary Of Shortest Distance Route...........................................................59

V ll


LIST OF FIGURES

FIGURE 1 Las Vegas Metropolitan Area....................................................................... 5FIGURE 2 Simple View Of Arc Info 9 Showing The Role Of Arc Objects............. 18FIGURE 3 Real Vs. Geometric Capacity Reduction By % Of Lanes Blocked 27FIGURE 4 A Graph With 6 Vertices And 7 Edges......................................................30FIGURE 5 Variation Of Traffic Demand By Day Of Week....................................... 35FIGURE 6 Crashes Before Buffering........................................................................... 39FIGURE 7 Crashes After Buffering.............................................................................. 39FIGURE 8 Selecting Start Major Street........................................................................ 50FIGURE 9 Selecting Cross Street For Start Location....................... 51FIGURE 10 Selecting Time Period..................................................................................52FIGURE 11 Selecting Day Of The W eek....................................................................... 53FIGURE 12 Free Flow Route Result.............................................................................. 55FIGURE 13 Route With Recurrent Delay.......................................................................56FIGURE 14 Route With Non Recurrent Delay.............................................................. 57FIGURE 15 Route With Recurrent Plus Non Recurrent Delay....................................58FIGURE 16 Shortest Route.............................................................................................. 59

V lll


ACKNOWLEDGEMENTS

The author wishes to express his sincere gratitude to his committee members Dr.

Shashi Nambisan, Dr. Srinivas Pulugurtha, Dr. Mohammed Kaseko and Dr. Wolfgang

Bein for serving in his advisory committee. The author is grateful to Dr. Pulugurtha for

his constant support and guidance throughout his thesis and Masters program. The author

thanks Carlos Moreno of Clark County Department of Public Works for his assistance.

The author also extends his thanks to Brian Hoeft o f Regional Transportation

Commission of Southern Nevada, Las Vegas Nevada for providing the travel demand

model data of the Las Vegas valley. The author acknowledges the support of the

Transportation Research Center at University of Nevada, Las Vegas for the support it has

provided during various stages of his Masters Program. The author would also like to

take this opportunity to thank all the students and staff at TRC, UNLV for their support

and sincere help extended for this research work. At this point the author would like to

thank the TRC at UNLV for financially supporting him with the graduate assistantship

which greatly helped him in the Masters program. The author expresses his sincere

gratitude to Dr. Namhisan, Director of TRC for heing a mentor and constant guide in his

education and personal life.

Last hut not the least the author thanks his family members for the constant support

and encouragement for his educational endeavors.

IX


CHAPTER 1

INTRODUCTION

Motivation

Traveling from one location to another is a very important daily activity which

consumes a significant amount of time for the commuter. Each trip is generally preceded

by a planning process for selecting the optimal route from the origin to the destination.

“Optimal” may be defined in various ways by different travelers; such as shortest time,

shortest distance, least total cost, or safest/most reliable path. The safest or most reliable

path is major concern considering the ever increasing congestion and vulnerability of

urhan/suburban street networks to crashes and other obstructions. Determining such a

path is of significant value not only to individual travelers, hut also to freight and delivery

service organizations in developing realistic and reliable schedules. The focus of this

research is to develop a methodology to determine such routes and to illustrate its

application using a GIS based tool.

Reliability is a term that is familiar to most engineers and transportation planners. The

reliability of a transportation network can be defined as the probability of one or more of

its links either fiinctioning or failing to function. A non-functioning, or at best badly-

functioning, link will impose costs on the user in terms of loss of time.


operational costs or other costs as a result of delays and diversions. Transporters of

perishable goods will also experience a loss o f value (Husdal, 2004). Reliability is then,

as Berdica (2002) describes it: “is the ability to maintain adequate serviceability

according to set operational parameters”. A truck load of perishable goods will have

different requirements for reliability than a truck transporting materials to a construction

site.

Many motorists who live in areas with heavy traffic congestion have noted that they

can “live” with slow moving traffic if it is predictable in its congestion, hut become

increasingly agitated when traffic flow conditions undergo a large swing of operations

i.e., smooth to stop-and-go flows (ORA, KHA and UNLV-TRC, 2003).

Highway incidents (vehicle crashes and disablements) are a major cause of traffic

congestion in urban areas. Previous studies have estimated that around 60% of all traffic

congestion on highways is caused by incidents (Lindley, 1987). Thus, considerable

attention has been directed toward the effective management of incidents. Such

management seeks to mitigate the traffic-related impacts of incidents by minimizing the

incident duration and by diverting traffic away from the area of the incident by providing

pre-trip information to road users and controlling traffic intelligently near the vicinity of

the incident.

In the recent decades road urban transportation systems have undergone considerable

increase in complexity and congestion proclivity. The field of ITS, Intelligent Transport

Systems, aims to develop and apply advanced technologies to make transportation

systems more safe and efficient, while reducing congestion, pollution and environmental

impacts. In working towards this goal, ITS can take many different forms. Use of


transportation Geographical Information Systems (GIS), network analysis, operations

research and artificial intelligence to find safe and reliable routes from an origin to

destination are strategies used to achieve the objective.

Research Objective, Tasks and Scope

The objective of this research is to develop a GIS based application to identify

optimal routes from an origin to destination. Modified travel time is used to identify an

optimal route. Estimates of modified travel time are obtained based on recurring and non

recurring congestion on the road network. Other routing criteria supported include the

quickest and shortest routes. The research presents a technique to determine the modified

travel time found based on historical data analysis of traffic volume and crash data on the

streets. Identifying individual segments or corridors that have relatively higher levels of

congestion can help the road user in making pre-trip decisions related to route choice,

length of the commute, time taken to reach the destination location etc. Djikstra’s

algorithm was used to find the optimal path from an origin to a destination based on the

selected objective routing (Husdal, 2000).

Delays caused due to recurrent and non-recurrent congestion need to be

differentiated. Two approaches for estimating road network reliability are proposed to

quantify delays due to recurrent and non-recurrent congestion. These indices are

expressed as functions of operating speed and travel time respectively.


Study Area

The research methodology and the application were tested on the street network in

Clark County, Nevada (Figure 1). The Las Vegas metropolitan area in the study area is

estimated to have 1.70 million residents (www.Wikipedia.com) and about 37 million

visitors per year.

Las Vegas is a new urban area by most standards. It was built during the first half of

the twentieth century in a vast desert with ample land for sprawl. A majority of the

population and economic growth in this area has occurred over the last 20 years. The

low-density template used to develop the desert city provided a traditional street grid

pattern with major surface arterial streets at every mile, and rights-of-way adequate to

provide six or eight lanes of traffic that generally facilitates posted speed limits o f 45

miles per hour on these arterials. Principal arterial streets and minor arterial streets

accounted for 47 percent of urban vehicle miles of travel in the greater Las Vegas area.

Texas Transportation Institute’s (TTI) Urban mobility study (2003) states that expansion

of the roadway network continues as the area struggles to serve the growing vehicular

demand, hut the length of time spent commuting gradually creeps up as growth in

roadway capacity has not heen able to keep pace with the growth in demands. The greater

Las Vegas area is a non-attainment area for national clean air standards, which casts a

shadow of uncertainty on the area’s ability to continue to increase lane miles indefinitely.


http://www.Wikipedia.com

Las Vegas Valley Street Network

[ g

KAREN

0 3,500 7,000 14,000 Feet1 inch equals 1.193892 miles

FIGURE 1 Las Vegas Metropolitan Area AReproduced witti permission of ttie copyrigfit owner. Furtfier reproduction profiibited witfiout permission.

In 2003, Las Vegas ranked 18th out of the 75 largest metropolitan areas in the US in

TTI’s travel time index. Travel takes 35% longer during peak travel times in Las Vegas

than it does in non-peak times. The average Las Vegas commuter spends 30 hours per

year stuck in traffic. This compares to just six hours of delay annually per person in Las

Vegas in 1982. The portion of daily travel taking place in congested road conditions has

tripled in Las Vegas since 1982, from 10 percent of travel then to 32 percent of travel in

2003.

In Las Vegas in 2003, 78 percent o f travel during peak travel periods was under

congested conditions.

Layout of the Thesis

This thesis report is divided in five chapters. The motivation for this study has been

presented in Chapter 1. Chapter 2 provides a hrief review of the literature review

pertaining to traffic congestion, reliahility, and shortest route algorithms. It identifies the

need for the proposed research objective. The methodology for analyzing the raw data,

developing the application, programming, determining incident frequency, estimating

incident delay, and finally using the processed data in the application are discussed in

Chapter 3. Chapter 4 presents a case study done to test the developed methodology and

the analysis of different route selection options in the application developed. In Chapter

5 the results are summarized along with conclusions and recommendations for future

research.


CHAPTER 2

LITERATURE REVIEW

Congestion

For several decades the growth of traffie volumes has generally outstripped the

growth in road infrastructure in urban areas in the US. The result has been an increase in

traffic congestion. Congestion imposes various costs on travelers: reduced speeds and

increased travel times, a decrease in travel time reliability, greater fuel consumption and

vehicle wear, inconvenience from rescheduling trips or using alternative travel modes,

and (in the longer run) the costs of relocating residences and jobs. The costs of increased

travel times and fuel consumption alone are estimated to amount to hundreds of dollars

per capita per year in the US (Schrank and Lomax, 2004) and comparable values have

heen reported for Europe. Traffic congestion is a consequence of the nature o f supply and

demand. Capacity is time consuming and costly to huild and maintain for long periods of

time. Demand also fluctuates over long periods of time, and transport services cannot be

stored to smooth imbalances between capacity and demand. Congestion delay consists of

recurrent delay plus the additional (non-recurrent) delay caused hy crashes, breakdowns,

and other random events, such as inclement weather and debris. Recurrent delay arises

from fluctuations in demand, the manner in which the roadway is operated, as well as the

physical layout of the roadway. Non-recuiTent delay depends on the nature of the


incident: a crash is likely to cause more delay than a vehicle stopped on the shoulder

of the highway.

In the United States, construction of new highway capacity has not kept pace with

population increases and the resulting increase in demand for highway travel. Between

1980 and 1999, the total length of highways as measured by miles increased by only 1.5

percent, while the total number o f miles of vehicle travel increased by 76 percent

(Lindsey et.al, 2000).

The TTI estimates that in 2003 the 75 largest metropolitan areas experienced 3.6

billion vehicle-hours o f delay, resulting in 5.7 billion gallons (21.6 billion liters) in

wasted fuel and $67.5 billion in lost productivity. Traffic congestion is increasing in

major cities, and delays are becoming more frequent in smaller cities and rural areas

(Lindsey et.al, 2000).

The five areas in the United States with the highest levels of traffic congestion are:

• Los Angeles, CA

• Washington, DC

• Miami-Hialeah, FL

• Chicago, IL

• San Francisco-Oakland, CA


Additionally, residents of Atlanta, Georgia have an average commute of 35 minutes.

This has been attributed to the large migration of people to the city and the fact that only

5 of the 28 counties that make up Metro Atlanta have any type of public transportation.

Due to dramatic population increases, San Diego and Las Vegas have seen their

congestion levels increase by more than 50 percent since 1982.

Basically, traffic congestion results when traffic demand approaches or exceeds the

available capacity of the highway system. While this is a simple concept, it is not

constant. Traffic demands vary significantly depending on the season of the year, the day

of the week, and even the time of day. Also, the capacity, often mistaken as constant, can

change because of weather, work zones, traffic incidents, or other non-recurring events.

It is estimated that roughly half of the congestion experienced hy Hoboken, NJ

residents is what is known as recurring congestion - caused hy recurring demands that

exist virtually every day, where demand for road use approaches or exceeds existing

capacity. The other half is due to non-recurring congestion caused by temporary

disruptions. The four main causes of non-recurring congestion are: traffic incidents

(ranging from disabled vehicles to major crashes), work zones, weather, and special

events. Non-recurring events dramatically reduce available capacity and reliability o f the

entire transportation system.

Due to the unpredictable behavior of congestion it would be of great use to the

commuters to have a tool which would help them decide a priori which route to take to

their destination. As mentioned in the ESRI white paper tool can be developed in ESRI

GIS product to show the route selected by any network analysis algorithm. Various

articles on congestion mention that half of the congestion caused on streets is of recurring


nature and other half non-recurring having an unpredictable nature. The research work

done hy Parimi (2004) on congestion to rank streets on reliability can be used to give a

ranking to each street link in a street network. In turn, these could he used as inputs to

any route finding algorithm to find reliable routes and shortest routes. Existing GIS

software and programming tools can he used to develop a tool to help identify the most

reliable routes.

Reliability and Vulnerability

Reliability needs to be more than simply a quantitative probability calculation related

to the functioning or non-functioning of a link. The probability of a link being present or

not, as is the case with traditional reliability engineering, does not translate into a

probability of the same link being operable or not. Consequently, reliability needs to have

a measure of serviceability in order to be of any value to the transportation planner or

engineer. If serviceability contains a description of whether or to what extent is possible

to use this link, this road, this route, this network, at a given time, then vulnerability can

be defined as a measure of how susceptible this link is to a severe reduction in its

serviceability. Reliability is a complement to vulnerability, namely the ability to maintain

adequate serviceability according to set operational parameters Berdica (2002).

Vulnerability, representing the consequential costs of a lack of reliahility is a quantitative

term that must be included in a cost-benefit analysis.

Reliability is an expression of the probability that links within the network will

function. Reliability may thus be viewed as the degree of stability of the quality of

service that a system offers. Bell and lida (1997) addressed transport network reliability

10


on connectivity reliability (also named terminal reliability) and travel time reliability.

Nicholson et al. (2002) list and discuss encountered reliahility, capacity reliability and

flow decrement reliability as ways of measuring reliability. Whereas prohability or

predictability is a major concern in network reliability studies, the impacts or

consequences of disruptions is the focus of vulnerability studies. Taylor and D’Este

(2003a) note that vulnerability and reliability are two related concepts, but emphasize that

network vulnerability relates to network weaknesses and the economic and social

consequences of network failure, not so much the probability of failure. Berdica (2002)

relates vulnerability to serviceability, namely the possibility to use a link, route or road in

a network at a given time. Vulnerability, then, is the inability to supply adequate

serviceability and serviceability as such is determined hy a set of performance measures.

Taylor and D’Este (2003b) relate vulnerability to the degree of accessibility of a given

node in the network, where accessibility is expressed as the travel cost needed to access

the particular node, comparing optimal and alternative routes or detours.

LLeras-Echeverri and Sanchez-Silva (2001) extend the minimum path calculations as

seen in Bell and lida (1997), to include a progressive failure scenario approach to identify

weak links. By assigning a centroid within the network, and restricting the study to

failures of links in the minimum paths between the centroid, critical links can be quickly

identified for further evaluation.

The term operability needs to be approached differently for different types of roads,

road conditions, goods or transports (Husdal 2004). The probahility of the network of

being available, accessible and expressing the desired level of service may be used as a

measure of reliability. Conversely, the road network’s susceptibility to not being

11


available, accessible and expressing the desired level of service may be construed as a

measure of vulnerability, as defined by Berdica (2002). The notion of susceptibility

implies something more than probable and statistically predictable failure, namely the

influence of external circumstances, circumstances that cannot be predicted, and which

may, but not necessarily will have an influence.

With reliability and vulnerability being two different concepts, a measure of

reliability does not translate into a measure of vulnerability. Vulnerable may not mean

non-reliable and reliable may not mean non-vulnerable. Lessening transport network

vulnerability may be different from improving transport network reliability. There may

be no linear or reciprocal relationship. Nonetheless, if associating reliable with being

operable, and associating vulnerable with being non-operable, then it is warranted to

juxtapose reliability with vulnerability: reliability, in this sense, means no vulnerability or

exhibiting a high degree of operability under any circumstances; vulnerability means

non-reliability or exhibiting a low degree of operability under certain circumstances.

Network Analysis

The analysis of transportation networks is one of many application areas in which the

computation of shortest paths is one of the most fundamental problems. These have been

the subject of extensive research for many years: Deo (1984), Cherkassky et al. (1993),

Zhan et al. (1996 and 1997).

A network consists of arcs, or links, and nodes. The fastest path is calculated as a

function associated with the cost of traveling the link. Even though the literature tends to

group the types of shortest paths problems slightly differently, one can discern, in

12


general, between paths that are calculated as one-to-one, one-to-some, one-to-all, all-to-

one, or all-to-all shortest paths.

Several algorithms and data structures for algorithms have heen put forward since the

classic shortest path algorithm hy Dijkstra (1959). In its modified version, this algorithm

computes a one-to-all path in all directions from the origin node and terminates when the

destination has been reached. The original Dijkstra's algorithm explores an unnecessary

large search area. This lead to the development of heuristic searches, among them the A*

algorithm by Jacob et al. (1998), that searches in the direction of the destination node.

This avoids considering directions with non- favorable results and reduces computation

time. It is noteworthy that the heuristic A* algorithm dominates the research literature

for static networks. Almost all comparison studies involving the A* algorithm conclude

with this algorithm’s superiority over other approaches. Dijkstra-based algorithms, with

various enhancements in their data structure in the form of heaps, buckets and queues, are

also well represented (Husdal, 2000).

Research has also proven that dynamic network fastest path problems can be reduced

to static fastest path problems if continuously varying link travel times are expanded for a

time interval or given an estimated value. The A* and Dijkstra's algorithms are applicable

in both static and dynamic networks (Husdal, 2000).

In order to address the issue of computational time in Djikstra’s algorithm it has to be

mentioned that Djikstra’s algorithm original strategy uses a queue data structure. In the

basic Dijkstra's algorithm (Dijkstra, 1959) the complexity bottleneck is node selection.

In enhanced versions of Dijkstra's algorithm distances are maintained in sorted order.

This avoids scanning all temporarily labeled nodes at every iteration to find the minimum

13


distance label. In turn this reduces the computation time. Heap (or priority queue) data

structures (Ahuja [1]) are implemented in modified Dijkstra algorithms. We can express

the running time of Dijkstra's algorithm on a graph with m edges and n vertices can be

expressed as a function of m and n using the Big O notation. The Big O notation is a

mathematical notation used to descrihe the asymptotic behavior of functions. More

precisely, it is used to describe an asymptotic upper bound for the magnitude of a

function in terms of another, usually simpler, function. In mathematics it is usually used

to characterize the residual term of a truncated infinite series, especially an asymptotic

series. The simplest implementation of the Dijkstra's algorithm stores vertices of set Q in

an ordinary linked list or array, and operation Extract-Min (Q) is simply a linear search

through all vertices in Q. In this case, the running time is O (n2). For sparse graphs, that

is, graphs with much less than n2 edges, Dijkstra's algorithm can be implemented more

efficiently by storing the graph in the form of adjacency lists adjacency list and using a

binary heap or Fibonacci heap Fibonacci heap as a priority queue to implement the

Extract-Min function. With a binary heap, the algorithm requires O ((m+n) log n) time,

and the Fibonacci heap improves this to O (m + n log n). This helps in reducing the

computational time.

To address the issue of space limitation in large networks, the network has to be

considered as sparse graph. The primary property of a graph to consider when deciding

which data structure to use is sparsity, the number of edges relative to the number of

vertices in the graph. A graph where E is close to V2 is a dense graph, whereas a graph

where E = alpha V and alpha is much smaller than V is a sparse graph. For dense graphs,

the adj acency-matrix representation is usually the best choice, whereas for sparse graphs

14


the adjacency-list representation is a better choice. Also the edge-list representation is a

space efficient choice for sparse graphs that is appropriate in some situations. An

adjacency-list representation of a graph stores an out-edge sequence for each vertex. For

sparse graphs this saves space since only O (V + E) memory is required. In addition, the

out-edges for each vertex can be accessed more efficiently. Edge insertion is O (1),

though accessing any given edge is O (alpha), where alpha is the sparsity factor of the

matrix (which is equal to the maximum number of out-edges for any vertex in the graph).

GIS in Network Analysis

In a broad sense a Geographic Information System (GIS) is an information system

specializing in the input, storage, manipulation, analysis and reporting of geographical

(spatially related) information. The wide range of potential applications GIS can be used

for includes transportation issues. The four major components of a GIS, encoding,

management, analysis and reporting, have specific considerations for transportation Chou

(1997):

• Encoding. Deals with issues concerning the representation of a transport system

and its spatial components. To he of use in a GIS, a transport network must be

correctly encoded, implying a functional topology composed of nodes and links.

Other elements relevant to transportation, namely qualitative and quantitative data

must also be encoded and associated with their respective spatial elements. For

instance, an encoded road segment can have data related to its width, number of

lanes, direction, peak hour traffic, etc.

15


• Management. The encoded information is stored in a file structure and can be

organized along spatial (by region, country, census block, etc.), thematic (for

road, transit, terminals, etc.) or temporal considerations (by year, month, week,

etc.). This process is commonly automated and has established conventions. For

instance, many government agencies use specific file formats and deploy their

information along pre-determined spatial (their jurisdiction), thematic (their field

of interest) and temporal (their frequency of data collection) considerations.

• Analysis. Considers the wide array of tools and methodologies available for

transport issues. They can range from a simple query over an element of a

transport system (what is the peak hour traffic of a road segment?) to a complex

model investigating the relationships between its elements (if a new road segment

was added, what would be the impacts on traffic and future land use

developments?).

• Reporting. A GIS would not be complete without all its visualization and

cartographic capabilities. This component is particularly important as it gives the

possibility to convey complex information in a symbolic format. A GIS T

(Geographic Information Systems for Transportation) thus becomes a tool to

inform and convince actors who otherwise may not have the time or the capability

for non-symbolic data interpretation.

Information in GIS is often stored and represented as layers, which are a set of

geographical features linked with their attributes.

In both rural areas and intercity corridors, traffic is disrupted by incidents,

maintenance operations, detours, and congestion on tourist routes, among other causes.

16


To the traveler, congestion means lost time, missed opportunities, frustration, and waste

of personal resources. Hence the need to find a reliable route from a source to destination

would be of great help to commuters. Network and Land data can be easily developed,

maintained and updated in GIS database. Transportation Planning and management needs

accurate and timely spatial and non-spatial information like, network, capacity, speed

restriction etc., to assist planning activities. Most importantly GIS database with the help

of its powerful geographic data analysis engine could serve the need to identify reliable

routes for commuters. Hence research is required to develop application to achieve this

objective.

Application Development in Arc Info

Arc Objects is the development environment of the Desktop ArcGIS applications Arc

Map, Arc Catalog and Arc Scene. It is used to customize and extend ArcGIS using the

embedded Visual Basic for Applications (VBA). Arc Objects is the development

environment introduced with ArcGIS 8. It provides customization for Arc Map, Arc

Catalog and Arc Scene, and is based on Visual Basic for Applications (VBA) that is

embedded in ArcGIS.

There are three levels of customization provided in ArcGIS: (Guide 93, Programming

in ArcGIS using Arc Objects and AML)

• Customize the user interface (does not use Arc Objects directly).

• Write scripts using VBA.

• Use an external development environment to create a stand-alone COM

(Component Object Model) component (using for example Visual Basic, Visual C++ or

17


Visual J++). These can be used to incorporate Arc Objects into non- GIS applications,

but they still need an ArcGIS license to run.

ArcinfoApplications

(Presentation)

Data(Data Management) Data

ArcMap ArcCatalog

ArcObjects

FIGURE 2 Simple View Of Arc Info 9 Showing The Role Of Arc Objects

Arc Info 9 includes three levels of customization, two of which utilize Arc Objects

directly. The first and simplest level of Arc Info customization involves no programming

and, therefore, no direct use of Arc Objects. All users can easily change the look and feel

of the Arc Info applications using standard user interface capabilities. For example,

toolbars can be turned on and off using the customize dialog. Many other application

properties can be changed using such simple menu-driven actions like this (ESRI White

Paper, 2000).

18


The second level involves using the in-built Visual Basic for Applications (VBA)

scripting capabilities to add new menus, tools, and work flows to the Arc Info

applications. The Arc Map and Arc Catalog applications include a fully working VBA

development environment, with an integrated debugger, which provides access to all the

Arc Objects technology. VBA allows the addition of modules, class modules, and user

forms to Arc Info project files. With VBA, it is possible to create extensive and

sophisticated applications based on Arc Objects that run within the Arc Map and Arc

Catalog application frameworks. It is not, however, possible to create new custom feature

classes or build applications that run outside Arc Map and Arc Catalog. VBA is a very

good choice for small-to-medium-size applications that use or extend the existing Arc

Info applications or functions. VBA is useful for many of the same types o f tasks that

ARC Macro Language (AML ™) has been used for in the past (ESRI White Paper, 2000).

A software component is a binary unit of reusable code. The key to the success of

components is that they implement in a very practical way many of the object-oriented

principles now commonly accepted in software engineering. Components encapsulate

behind a well-defined interface the manipulation methods and the data that characterize

each instantiated object. This promotes structured and safe system development.

Components facilitate software reuse because they are self-contained building blocks that

can easily be assembled into larger systems. They also support inheritance and

polymorphism. Inheritance is the ability to reuse functionality in other components by

including reference to another object's state and behavior. For example, a new type of

water valve can easily be created by overwriting or adding a few properties or methods to

an existing type of valve that is similar. Polymorphism describes the process whereby

19


each object has its specific implementation for operations such as draw, create, and

delete. One example of the benefits of polymorphism is that a GIS can have a generic

object creation component that issues requests to be processed in a specific way by each

type of object class factory (e.g., gas pipes, valves, and service lines) (ESRI White Paper,

2000).

Upon studying the benefits of ArcGIS it appears the software would provide the

perfect platform to present the methodology. Also the ability to present the geographical

data in an easy to understand layout would be very helpful for the commuters to use the

product. ArcGIS also has the capability o f showing the maps on internet for a wider

coverage of the population. Understanding the power of ArcGIS and its robustness,

ArcGIS is considered as software for the research.

20


CHAPTER 3

METHODOLOGY

This chapter describes the methodology involved in developing the model for

identifying optimal routes based on alternative routing objectives. It starts with the

analysis of data and a description of the model.

Travel Time Analysis

Travel time is defined as the time taken to travel along a particular segment of a

corridor or length of road, between points of known distance. Knowing the travel speed

on the segment and the distance between known points, the travel time can be computed.

A comparison of the actual travel time with the travel time under ideal conditions

provides a sense of the degree of congestion present on that route. Travel time is a

commonly used measure of system performance by various agencies. The travel time

measure can vary depending on whether stops or time spent in queues or under a certain

speed are included or excluded from the travel time calculations. The typical calculation

includes total travel time where such delays are included. An example calculation of

travel time on the freeway between TMS (Traffic Monitor System) stations can be

estimated by:

Travel time ti = (di)/si -------------- 3.1

21


Where di is the distance between the known points such as the detector station "i" and “j ”

on the road; s, is the speed between the detector station "i" and “j ”.

Delay can be defined as the additional time required to travel a distance due to

impeded travel conditions on the road when compared to the travel time under

unimpeded conditions. Delay measures the degree of congestion indirectly by quantifying

the difference in travel times between conditions that would have allowed free flow

traffic and the prevailing traffic conditions. For freeways it can be computed by finding

the difference in travel times with free flow speed and the actual speed between two

known points. For surface streets delay, it can be measured by taking the difference in

actual travel time and the acceptable travel time between two know points or stations.

The equation for travel time under existing traffic conditions To, is

To = D(j / So -------------- 3.2

Where Da is the distance in miles and So is speed in miles per second.

The equation for computed travel time with free flow speed or acceptable speed limit Tf,

is

Tf = Dd/S f ---------------- 3.3

Sf is the free flow speed.

Delay is then defined as follows:

Delay = T o-T f.-------------- 3.4

This calculation determines delay over a particular segment of the roadway. Delay is

a useful measure as it is relatively easy to compute in comparison with other MOEs

(mesure of effectiveness). It is also widely used and appreciated by the general public and

22


can be compared readily with historic conditions to measure the temporal impacts on a

system.

Recurring Congestion Analysis

The following steps were followed in the analysis of recurring congestion. The traffic

volumes on individual links during different times of the day were obtained from a

regional travel demand model (ORA, KHA and UNLV-TRC, 2003). A typical day was

divided into seven different time periods as shown in Table 1. The daily variations in the

ADT (Average Daily Traffic) values are obtained from the report and this correction is

made to the volumes on different weekdays. For example, on a Sunday the ADT is equal

to 80.1% of AADT (Average Annual Daily Traffic) and on a Monday the ADT is equal

to 107% of AADT. Hourly traffic counts are used to distribute the average daily traffic

value into the different time periods.

ADTi = P i* AADT----------- 3.5

Vij =ADTi * kj -------------3.6

?i = adjustment factor for day of the week

Vij = traffic volume for time period ‘j ’ on weekday ‘i’

kj = adjustment factor for the time period of the day

23


Table 1 Time Periods Over A DayNo. Time Period # Hours Category1 00:00 - 07:00 7 Night, Off - Peak2 07:00 - 09:00 2 Day, Peak3 09:00 - 14:00 5 Day, Off Peak4 14:00-16:00 2 Day, Off Peak5 16:00-18:00 2 Day, Peak6 18:00-20:00 2 Day, Off Peak7 20:00 - 24:00 4 Night, Off-Peak

The “ideal” travel time for each street link was calculated using the posted speed limit

and length of the street link. The volume to capacity ratio is computed for each street

link and the LOS (Level of Service) is assigned based on the volume to capacity ratio for

each link. The “actual” travel speed is assigned to each street link based on the LOS of

the street link. Again, travel time is computed based on this prevailing travel speed.

The “actual” vehicle speed during each time periods can be calculated using the

Bureau of Public Roads relation (Dowling et al., 1997),

Act

\ + a--3.5

Where, Saci is the actual reduced speed caused by congested conditions, Sp is the free

flow speed under normal condition, C is the capacity of the roadway, a and b are

constants. Thill et. al (2001) used values of a = 0.15 and h - A .

Practical capacity is defined in this equation as 80% of the capacity. Free-flow speed

is defined as 1.15 times the speed at the practical capacity. The parameter “a” determines

the ratio of free-flow speed to the speed at capacity. The parameter “b” determines how

24


abruptly the curve drops from the free-flow speed. A high value of “b” causes speed to be

insensitive to v/c until the v/c gets close to 1.0, and then the speed drops abruptly.

Analysis of Non Recurring Delays

Another type of delay encountered by travelers is incident delay. This is the delay that

results from a crash or a disabled vehicle. Incident delay is related to the frequency of

crashes or vehicle breakdowns and how easily those incidents are removed from the

traffic lanes and shoulders.

The yearly crash database prepared and maintained by the Nevada Department of

Transportation Safety Engineering Division in cooperation with Nevada Department of

Motor Vehicles and Public Safety and other state and local law enforcement agencies

(NDOT, 2003) is used for calculating the incident frequency. Other incidents such as

stalled vehicles, vehicle breakdowns are not considered to be in the scope of the study.

Crashes for a five year period (1998 to 2002) were used to find the delay due to non

recurrent congestion caused by crashes in this research study. The process used to

estimate such delays is discussed next.

Creating Buffers and Spatial Overlay

Buffering and overlay are two of the most common operations in cartographic

modeling. A buffer zone is an area that is within a given distance from a map feature.

Points, lines, or polygons can be buffered. Buffers are used to identify areas surrounding

geographic features. However, the crash locations may not be exactly on the link in case

the street segment is not a grid network or if the streets be exactly on the link the use of

25


buffers is appropriate [Pulugurtha (2004)]. In order to get the number o f crashes on a

particular street link buffer analysis for the street network was done for 100 feet to

capture crashes outlying on a particular street link.

Buffer analysis was performed on the crash data. Once the crash data were captured

for each street link, the crash data and the street network data were joined.

The delay calculation for non recurring delay [Parimi (2004)] is found by the

following equation.

...............

The equation indicates that the amount of traffic delay due to an incident is

determined by four variables: the roadway capacity before an incident occurs C (vph),

traffic volume V (vph), the capacity after the occurrence of incident (lane blockage) C r

(vph), and the incident duration T (min). However, there are other influencing factors like

time of day, peak vs. non-peak hour conditions that influence some of these variables,

which in turn, affect incident delay.

The occurrence of an incident on a roadway segment is expected to reduce its

capacity. The estimated reduced capacity C r depends upon the number of lanes blocked

by an incident, n . In general, a non-recurring incident will reduce the roadway capacity

by an amount greater than the reduction in actual space. One of the reasons behind this

capacity reduction is the rubbernecking phenomenon, where the effect of an incident in a

lane translates into neighboring lanes because of the tendency of the drivers to slow down

after spotting the incident. For example, if an incident blocks two out of three lanes, the

theoretical road space reduction is sixty six percent while the observed traffic capacity

reduction is seventy nine percent (Yi, 2002). The ratio of traffic capacity reduction versus

26


road space reduction will always be greater than or equal to 1.0. Presley (1998)

developed a chart for this ratio and basing on three data points in the chart, two equations

(linear and quadratic) are developed as shown in Figure 5.

For simple calculation purposes the linear equation is adopted.

y = 1.7467-0.7679% -------------- 3.7

Where, y = TCr / R S r and % = RSr.

TC r is the traffic capacity reduction ratio, T C r = C r / C and R Sr is the road space

reduction ratio, R S r = n /N, where n is the number of lanes blocked by an incident and N

is the number of lanes in the road segment where the ineident occurred.

22

1.8 y = 0.5744x - 1.5338x+ 1.95^1.6

y = -0.7679x+ 1.74671.4

0.80% 20% 40%

% of Lanes Bbcked60% 80% 100%

♦ Ratio — — linear (Ratio)

FIGURE 3 Real vs. Geometrie Capacity Reduction By % Of Lanes BlockedSource: Calculating Benefits fo r NA VIGA TOR: G eorgia’s Intelligent Transportation System

27


Because y = T C r / R S r and R Sr = n/N, n should not equal to 0. According to

Sullivan (1997), the traffic capacity is reduced to about 77% when n = 0. So the capacity

of the occurrence of an incident can be estimated by the following equation,

Q = C - ( C x y x ( n / w ) ) -------------- 3.8

Q = 0.77 • C [When n = 0]-------------- 3.9

=> Q = C - ( l .7 4 6 7 -0.7679(«/w))x Cx (n/iv) [When n > 0] -------------- 3.10

By using average values for incident duration and incident lane blockage and

estimating residual capacity, traffic delay is found for the different time periods. The time

of blockage would depend on the severity of the crash. A fatal crash is expected to take a

longer time to clear than an injury crash and thus the roadway would operate with less

capacity for a longer duration for a fatal crash than for an injury crash. The effect of an

incident would be more on a single lane roadway when compared to a multilane facility.

Various influencing factors are considered while suggesting the average values of

incident duration and incident lane blockage like the type of incident, time of day, peak

vs. non-peak conditions etc.

Once the delay is calculated for each type of incident the delay for each street link

due to non recurrent delay is assigned based on the reliability index methodology

formulated by Parimi (2004).

Delay Due to Work Zones

Periodically agencies with responsibility for the roadway operations and maintenance

come out with plan on location where and when construction or maintenance of street

links would be carried out. The delay due to work zones occurs due to reduced speed

28


limits and lane closures. The reduction in capacity and throughput can also be calculated

if the reduced speed limit and the number of lanes closed are known and the travel time

for that link can be calculated for each time zone and day of the week. The new travel

time can be then updated in the database of the application. Since in this research the

effect due to crashes is considered mainly at this point the congestion due to work zones

is not included in the analysis.

Travel time estimation as described would provide with delay caused by recurring

and non recurring congestion for each street link. This delay for links can be used in

network analysis to find routes having the smallest delay. An application to do the

network analysis would serve the purpose.

Dijkstra's Algorithm

In mathematics and computer science a graph is the basic object of study in graph

theory. Informally, a graph is a set of objects called vertices connected by links called

edges. Typically, a graph is depicted as a set of dots (the vertices) connected by lines (the

edges). Depending on the application some edges can be directed.

29


FIGURE 4 A Graph With 6 Vertices And 7 Edges

Dijkstra's algorithm, is an algorithm that solves the single-source shortest path

problem for a directed graph with nonnegative edge weights. For example, if the vertices

of the graph represent cities and edge weights represent driving distances between pairs

of cities connected by a direct road, Dijkstra's algorithm can be used to find the shortest

route between two cities. The input of the algorithm consists of a weighted directed graph

G and a source vertex s in G. We will denote V the set of all vertices in the graph G. Each

edge of the graph is an ordered pair of vertices (u,v) representing a connection from

vertex u to vertex v. The set of all edges is denoted E. Weights of edges are given by a

weight function w; E -> [0, o^; therefore w(u,v) is the non-negative cost of moving from

vertex u to vertex v. The cost of an edge can be thought of as (a generalization of) the

distance between those two vertices. The cost of a path between two vertices is the sum

of costs of the edges in that path. For a given pair of vertices s and t in V, the algorithm

30


finds the path from s to t with lowest cost (i.e. the shortest path). It can also be used for

finding costs of shortest paths from a single vertex s to all other vertices in the graph

(www.wikinedia.com).

Description of Dijkstra’s Algorithm

Dijkstra’s algorithm works by keeping for each vertex v the cost d[v] of the shortest

path found so far. Initially, this value is 0 for the source vertex s and infinity for all other

vertices, representing the fact that we do not know any path leading to those vertices.

When the algorithm finishes, d[v] will be the cost of the shortest path from s to v or

infinity, if no such path exists.

The basic operation of Dijkstra's algorithm is edge relaxation; if there is an edge from

u to V, then the shortest known path from s to u can be extended to a path from s to v by

adding edge (u, v) at the end. This path will have length d[u] +w (u, v). If this is less than

d[v], we can replace the current value of d[v] with the new value.

Edge relaxation is applied until all values d[v] represent the cost o f the shortest path

from s to V. The algorithm is organized so that each edge (u, v) is relaxed only once,

when d[u] has reached its final value.

The algorithm maintains two sets of vertices S and Q. Set S contains all vertices for

which the value d[v] is already known as the cost of the shortest path and set Q contains

all other vertices. Set S starts empty, and in each step one vertex is moved from Q to S.

This vertex is chosen as the vertex with the lowest value of d[u]. When a vertex u is

moved to S, the algorithm relaxes every outgoing edge (u, v).

31


http://www.wikinedia.com

Application Development

Network analysis in GIS is often related to finding solutions to transportation

problems. In general, a network is a system of interconnected linear features through

which resources are transported or communication is achieved. The network data model

is an abstract representation of the components and characteristics of real-world network

systems. A network model can be defined as a line graph, which is composed of links

representing linear channels of flow and nodes representing their connections (Lupien et

al., 1987). In other words, a network takes the form of edges (or arcs) connecting pairs of

nodes (or vertices). Nodes can be junctions and edges can be segments of a road or a

pipeline. For a network to function as a real-world model, an edge will have to be

associated with a direction and with a measure of impedance, determining the resistance

or travel time along the network.

Tree-building algorithms find the shortest path from an origin node to all other nodes,

producing a tree of shortest paths with branches emanating from the origin. (Lombard et

al., 1993). The most commonly used tree-building procedure is that originally developed

by Dijkstra (1959), of which to date many modifications and improvements have been

made for specific applications. In order to find a path, the algorithm will build a tree data

structure that represents specific paths through the network. This is often referred to as a

breadth-first search, which fans out to as many nodes as possible before penetrating

deeper into the tree (Dolan, 1993). Starting from one origin node, the search tree builds

branches in all directions, adds up the resistance figures, and keeps only those that

represent the cumulative least cost. For each new set of adjacent nodes the calculations

for all possible edges towards these nodes are repeated till all nodes and edges have been

32


utilized, and the final destination is reached with minimal cost. During the process, edges

may appear in the search tree and then disappear as the calculations discard their value.

GIS programs provide effective decision support through their database management

capabilities, graphical user interfaces and cartographic visualization of complex spatial

and temporal congestion patterns and the resulting shortest paths and reliable route. A

GIS facilitates bringing various types of data together based on the spatial characteristics

of the data. But unlike a static paper map, a GIS program can display many layers of

information that are relevant (www.esri.com). GIS helps to integrate, visualize, manage,

solve, and present the information in a new way. Relationships between the data will

become more apparent and the data will become more valuable. GIS gives us the power

to create maps, integrate information, visualize scenarios, solve complicated problems,

present powerful ideas, and develop effective solutions like never before.

Computing Travel time for each route for different type of delay conditions can be

obtained from the application by summing up the delay in each individual street link.

The route with least delay value is computed by the algorithm implemented in the

application developed and displayed to the user.

33


http://www.esri.com

CHAPTER 4

ANALYSIS, CASE STUDY & RESULTS

This chapter presents the analysis of the methodology to identify the most reliable

route. Delay due to recurrent and non recurrent delays are incorporated in the application

developed based on this methodology to find the reliable route from a source to a

destination. The procedure involved in the calculation of delay is first finding the delay

due to the recurring delay and then finding delay due to non recurring delay and summing

both the delays. The application used information from the street network in the Las

Vegas metropolitan area to illustrate the use of the methodology.

Recurrent Delay

The daily traffic volumes for individual links by the time of the day are obtained from

the RTC {Regional Transportation commission o f Southern Nevada) regional traffic

model. The traffic demands by the day of the week are obtained as described in the

methodology. The daily variation of ADT is shown in Figure 6. Once the traffic demand

for each time period and different day of the week are collected, the recurrent delay can

34


be calculated. The difference between the free flow speed and the actual travel speed is

the recurrent delay for each street link.

The posted speed limit for each street link is used to calculate the free flow travel

time using the equation

Travel time in seconds - (L / S) * 3600/5280 -------------- 4.1

Where L is the length in feets of the street link and S is the posted speed in miles per

hour. Table 2 provides an illustration of the computation of “ideal” or “free flow” travel

times on individual links.

120.0%113.1%

106.5%105.9% 104.7%103.6%

100.0%92.5%

80.0% 73.7%% of AADT

60.0%

40.0%

20.0%

0.0%Thursday Friday SaturdayMonday Tuesday Wednesday

Day of the Week

Sunday

Source: Regional Transportation Commission o f Southern NevadaFIGURE 5 Variation Of Traffic Demand By Day Of Week

35


Table 2 [llustration Of Posted Speed And Travel TimeLink Function Length (ft) Posted speed (mph) Travel time (seconds)

3480 Freeway 320 65 3.33481 Ramp 186 30 4.23484 Collector 88 45 1.33486 Ramp 186 30 4.23493 Collector 1,016 45 15.43495 Collector 710 45 10.73496 Freeway 184 65 1.93498 Collector 3,958 45 59.93503 Collector 3,436 45 52.03507 Collector 5,285 45 80.03509 Collector 2,970 45 45.03523 Ramp 621 30 14.13531 Ramp 963 30 21.93533 Collector 2,950 45 44.73534 Interstate 678 65 7.13536 Interstate 1,599 65 16.73544 Ramp 930 30 21.13558 Ramp 755 30 17.13563 Minor Arterial 9,625 45 145.83565 Freeway 561 65 5.83568 Collector 4,532 45 68.63584 Collector 2,431 45 36.83602 Interstate 22,564 65 236.63610 Collector 1,330 45 20.13618 Collector 1,301 45 19.73625 Collector 1,259 45 19.03626 Collector 1,407 45 21.33647 Collector 1,302 45 19.7

Once the free flow travel time is computed, the travel time during different time

periods and during different days o f the week under prevailing traffic conditions need to

be estimated. To find the travel time during different time periods and different days o f

the week, the volume to capacity ratio has to be found for each of these periods. The BPR

equation is then used to estimate get the prevailing traveling speed for the link during

different time periods. This is illustrated in Table 3.

36


CD■DOQ.CgQ.

■DCD

C/)C/)

8

( O '

Table 3 Estimated Prevailing Vehicle Speeds (mph) During Different Time Periods On A Sunday, Monc

33"CD

CD■ D

OQ.Cao3"Oo

CDQ.

■DCD

C/)C/)

U)-J

ID FUNCTIONAL Traffic speed / Sunday Traffic speed / Monday

^1

i i

&

-4-Z '

(U

V

iAzn

OO

At/3

0 00

1

Tf

«N

i

" 1

i i

a-

Z '

i

so

5

i

eg-

00

5

■S«Deg-

0

1

eg-

Tf

cs

<ueg-

3481 Ramp 65 65 65 65 65 65 65 65 65 65 65 65 65 653484 Collector 30 30 30 30 30 30 30 30 30 30 30 30 30 303486 Ramp 45 45 45 45 45 45 45 45 45 45 45 45 45 453493 Collector 30 30 30 30 30 30 30 30 30 30 30 30 30 303495 Collector 45 45 45 45 45 45 45 45 45 45 45 45 45 453496 Freeway 45 45 45 45 45 45 45 45 45 45 45 45 45 453498 Collector 65 65 65 65 65 65 65 65 65 65 65 65 65 653503 Collector 45 45 45 45 45 45 45 45 45 45 45 45 45 453507 Collector 45 45 45 45 45 45 45 45 45 45 45 45 45 453509 Collector 45 45 45 45 45 45 45 45 45 45 45 45 45 453523 Ramp 45 45 45 45 45 45 45 45 45 45 45 45 45 453531 Ramp 30 30 30 30 30 30 30 30 30 30 30 30 30 303533 Collector 30 30 30 30 30 30 30 30 30 30 30 30 30 303534 Interstate 45 45 45 45 45 45 45 45 45 45 45 45 45 453536 Interstate 65 65 65 65 65 65 65 65 65 65 65 64.5 65 653544 Ramp 65 65 65 65 65 65 65 65 65 65 65 64.5 65 653558 Ramp 30 30 30 30 30 30 30 30 30 30 30 30 30 303563 Minor Arterial 30 30 30 30 30 30 30 30 30 30 30 30 30 303565 Freeway 45 45 45 45 45 45 45 45 45 45 45 45 45 45

ay

After the actual time according to the time period and day of the week are calculated,

the actual travel time of vehicle in each time period can be recalculated using the

following equation

Actual Travel time in seconds - (L / Saci) * 3600/5280-------------- 4.2

Where L is the length of the street link in feets and Saci is the actual speed in mph.

The recurrent delay for each street link is obtained by subtracting free flow travel time

from the actual travel time. This delay is used as a link attribute for each link to select the

route due to recurrent delay.

Non Recurrent Delay

In order to estimate non recurrent delay, historical traffic crash data are used. There

are three types of crashes that occur on streets involving vehicles: Fatal Crash, Injury

crash and PDO (Property Damage Only) Crash. Crash data from 1998 to 2002 were

obtained and overlaid on the street network as shown in Figure 7. It could be seen there

are lot of crashes that are not exactly over the major street network but they are close to

individual street links. In order to capture these crashes a buffer is created for 100 feet on

the street network and all the crashes falling outside the buffer are discarded as shown in

Figure 8 (these would be crashes on the “minor” streets). For each street link the crashes

are then split based on their time of occurrence, severity and day of the week. The

resulting database for the Las Vegas valley street network has 5,349 rows and 149

columns wide. Each row represents one street link. A sample of this crash database is

shown in Table 4.

38


W A SN IN

mMEADOW S

•A L TA

CHARLEST

FIGURE 6 Crashes Before Buffering

• w a s h i n S t o n►— Ü

K GRAGSON

MEADOW S

a.ALTA

CHARLESTON

FIGURE 7 Crashes After Buffering

39


T S > 2 "0 3 A v a sa n i

TS~03~8IAVaS3ai

i s ' s r o i A v a s a n i

TS“9 r h i A v a s a n i

i s > r 6 A v a s a a i

is~ 6~A A vasanx

i s “^“ T A v a sa n x

T S > r“G2AVaN0P\[

IS“03“ 8IAV(IMOW

lS "8l"9 lA V aN 0M

TS~9rhTAVaM0W fS (N (N

IS“ l7r6AV(IMOM

IS“ 6“ /-AV<IN0IA[

tsT ia v c m o e j CN

IS > 3 “03AVCIN[aS

IS“'02“ 8lA V Cim S

IS~8r9IAVClM lS

IS“ 9rhIA V O N nS

i s ^ r ô A v o m s

IS"6“ /,AVOMnS <N

tsT ia v o n a s

S G q S B J J I B f O l r- <N (N m 00(N

<N T f m 00 CN CN

S5[un m Tf o T f >0 (N(N CN

IT)fS (Nr -(N

om

CNm

mm

î1t8

IO3(DO

ImB

*

40


As the lane blockage duration varies by severity, time of the day, weather condition

etc., arbitrary values are proposed to illustrate the methodology and these values are

showed in Table 4 (Parimi 2004). These values are expressed as percentage time; each

lane is blocked and is shown in Table 5.

Table 5 Time Of Lane Blockage (minutes)Lane Blockage in minutes (Day, Peak)

Fatality Injury PDOLanel Lane2 Lane3 Lanel Lane2 Lane3 Lanel Lane2 Lane3

44 14 0 34 4 0 12 0

Lane Blockage in minutes (Day, Off-Peak)Fatality Injury PDO

Lanel Lane2 Lane3 Lanel Lane2 Lane3 Lanel Lane2 Lane336 10 0 25 0 0 7 0

Lanel(lockage in minutes (Night, Off-Peak)Fatality Injury PDO

Lanel Lane2 Lane3 Lanel Lane2 Lane3 Lanel Lane2 Lane343 19 0 30 5 0 17 0

Table 6 Percentage Time Of Lane Blockage% Time of Lane Blockage (Day, Peak)

Fatality Injury PDOLanel Lane2 Lane3 Lanel Lane2 Lane3 Lanel Lane2 Lane3100.00% 31.82% 0.00% 100.00% 11.76% 0.00% 50.00% 0.00%

% Time of Lane Blockage (Day, Off-Peak)Fatality Injury PDO

Lanel Lane2 Lane3 Lanel Lane2 Lane3 Lanel Lane2 Lane3100.00% 27.78% 0.00% 100.00% 0.00% 0.00% 3&89% 0.00%

% Time of Lane Blockage (Night, Off-Peak)Fatality Injury PDO

Lanel Lane2 Lane3 Lanel Lane2 Lane3 Lanel Lane2 Lane3100.00% 44.19% 0.00% 100.00% 16.67% 0.00% 77.27% 0.00%

41


Based on the values identified in Table 3 and 4, the average values of lane blocked

are found for fatal, injury and PDO crashes under different conditions. These average

lane blockage values are estimated depending on the percentage time each lane is

blocked. For example, a fatal crash during a day peak hour, will have a lane blockage

value of 1 + 31.82% x 1 + 0.00% x l) = i j i g . The calculation is illustrated

Table 6; the results are shown in Table 7.

m

Table 7 Median Value Of Lane BlockageNumber of Lanes Blocked (n)

Fatality Injury PI)0Peak Off-peak Peak Off-peak Peak Off-peak

Day 1.318 1.278 1.118 1.000 0.500 0.389Night 1.442 1.167 0.773

The lane blockage values are then entered for each street link in the Las Vegas street

network for all links having more than one lane. A sample of this table is shown in Table

7.

42


CD■DOQ.C

gQ.

■DCD

C/)C/)

8

( O '

3.3"CD

CD■DOQ.Cao3"Oo

CDQ.

■DCD

C/)C/)

4U)

Table 8 Lane Blockage Estimate For Street Network

a

w

gc/3

IH1

B§c/3

jc/3

C/3jTT

C/3 0 5 "

c/3

o '

c/3

ic/3

3480 TONOPAH Freeway 1.442 1.3182 1.278 1.278 1.3182 1.278 1.4423481 XXX Ramp 1.000 1.000 1.000 1.000 1.000 1.000 1.0003484 KYLE CANYON Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.0003486 XXX Ramp 1.000 1.000 1.000 1.000 1.000 1.000 1.0003493 KYLE CANYON Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.0003495 KYLE CANYON Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.0003496 TONOPAH Freeway 1.442 1.3182 1.278 1.278 1.3182 1.278 1.4423498 KYLE CANYON Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.0003503 SKYPOINT Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.0003507 OSO BLANCA Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.000

IRON3509 MOUNTAIN Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.0003523 XXX Ramp 1.000 1.000 1.000 1.000 1.000 1.000 1.0003531 XXX Ramp 1.000 1.000 1.000 1.000 1.000 1.000 1.0003533 DURANGO Collector 1.000 1.000 1.000 1.000 1.000 1.000 1.0003534 1 15 Interstate 1.442 1.3182 1.278 1.278 1.3182 1.278 1.4423536 1 15 Interstate 1.442 1.3182 1.278 1.278 1.3182 1.278 1.4423544 XXX Ramp 1.000 1.000 1.000 1.000 1.000 1.000 1.0003558 XXX Ramp 1.000 1.000 1.000 1.000 1.000 1.000 1.000

After the lane blockage is computed the next step is to find the reduced capacity due

to a crash which is computed using the equation

C „ = C - [ - 0.7679(«/n)+ 1,7467)xC x(n/v) -------------- 4.3

By using average values for incident duration from Table 5 and incident lane

blockage from Table 8, the residual capacity using equation 4.3 is calculated. A sample

of the result is shown in Table 9. Traffic delay is found for different time periods using

equation

_ t; - . ( c - c . ) . ( f - c , ) _________

44


I>N&u"S

Ia\

H

IS P Z O IA VO NO H

o> NOCOOnVOCN OVOo

VOCNr-VO inCN

rn

inCNr-in00\o

inP!

VO(Nr-VOVOOv

VOVOOvVOCNr-

ooVOrn

VOOvTprri

oVOOoVOCO

oVOp-osrri

VOP!

VO

I NO VO O VO in in in VO VO VO VO VO o VO O o o o VO inCO (N VO CN <N (N CN 00 <N IN VO VO IN o ov VO VO VO o rsiS O Z 8TAVONOPV

ON o r- Ocn

r- VO r- ov Ov p" VOrri rri

o CO p-rri

P- r-

I NO VO o VO in in in in VO VO VO VO VO o VO o o o o VO inON CO CN VO CN <N (N CN 00 <N IN VO VO IN o OV VO VO VO o IN (NS 81 QTAVCINOW

ON r- o N- r-rn

r- VO r- r- Ov Ov p" VOrri

TPcri

o o p-rri

P r-

I r-' T)- VO o vo in in in in VO VO VO VO vo o VO o o o o VO inON CN VO CN (N (N <N 00 <N IN VO VO IN o Ov VO VO VO o IN <NS 91 blA V O N O W

On O o t-'m

VO f" Ov ov p VOrri rri

o o p' VOcri

P~ r-

VO o VO in in in in VO VO VO VO VO o VO o o o o VO inON CN VO CN (N <N CN 00 <N IN VO VO IN o ov VO VO VO o IN <NIS V I 6 A V 0 N 0 M ON r' o C" r-

rnr- VO r- t" Ov Ov p" VO

rri mo CO p- CO

rriP- r-

m NO VO o vo in m in in VO VO VO VO VO o VO o d o o VO inIS“ 6“ /.AV(IN[0W

NO CO CN VO CN (N <N CN 00 (N IN VO VO IN o ovVO VO VO o IN (NON o r-m

<o r' r- ovov p- VOiri m

o o COcri

r>

NO VO o VO in in in in VO VO VO VO VO o VO o o o o VO inis“z.“iavcin:ow

ON CO CN VO CN (N CN CN 00 (N IN VO VO IN o Ov VO VO VO o IN (N'r-4

ON o N- r-m

r> \o t" Ov ov p~ VOrri cri

o CO p- COrri

P- r-

I o o VO o VO in o in in VO VO o o VO o o o CO o o VO ino in CN VO CN <N o CN CN (N IN VO VO IN o o VO VO VO o IN (Ns v z o r A v o m s rn o IN o N- o

rni> l> r~- o o p- VO

rriVOrri

o CO p" COrri

P~ o

I o o VO o VO in o in in VO vo o o VO o o o o o o VO ino VO CN VO CN <N o CN CN <N IN VO VO IN o o VO VO VO o IN <Ns “ oz:“ 8 iA V ( im s

ro o IN o N-rn

o r- o o p- VOcri

VOcri

o o p- COcri

P- r-

I O VO VO o vo in o in in VO VO o o VO o o o o o o VO ino <N CN VO CN <N o CN CN IN IN VO VO P\l o o VO VO VO o IN <NS 81 9IAV(INflS o N o r- V:

rnr- r- r- r- o o p" VO

criVOrri

o (O p" COrri

P" c

I o VO VO o VO in o in in VO VO o o VO o o o o o o VO ino CN CN VO CN (N o CN CN IN IN VO VO IN o o VO VO VO o IN (NS 91 t^IAV O N nS

fO o N- o C-- r-rn

r- r- r- o o p- VOrri

VOrri

o CO p~COrri

P~ r-

o O VO o vo in O in in VO VO o o VO o o o o o o VO mi s > r 6 A v a m s

o VO CN VO CN rsi o CN CN IN IN VO VO IN o o VO VO VO o IN (Nrn o N- o N- VDrn

o r- o o p~VOrri

VOcri

o (O p COrri

P

o o VO o VO in o in in VO VO o o VO o o o o o o VO ino VO CN VO CN <N o CN CN IN IN VO VO IN o o VO VO VO o IN (NIS 6 /.AVQ M flS rn o N o C r- VOrn

O r-* o o p- VOrri

VOcri

CO CO p COrri

o o Vi o V) IT) o in in VO VO o o VO o o o CO o o VO ini s “ A ~ iA v a N n s

o VO CN VO CN r\| o CN CN IN IN VO VO P4 o o VO VO VO o IN <Nm (O N o N r- VOrn

o r- o o p~ VOrri

VOrri

CO C) COcri

P

o NO CO mVO 00 m r-- o\ ro m "Tp VO -4" 00 m VO ooax CO CO CO CO N" Tf in VO VO 00000000ovo mIN IN IN IN

<N CN CN CN CN CN CN CN CN IN CN CN CN CN CN CN CN CN IN CN CN (N

45


Iu

I

IS'tZ 'O ZA V O K O MooTt l>CO 00CN CNCO 00CN 00CN

OOG\ 00CNT—Hcn 00CN OOCN 00cn 00cn 00CN cn Os CNcn CNcn Ocn

00 r-' 00 CN 00 00 00 00 00 00 00 00 00 cn o\ CN CN o cncn CN CO CN CN CN CN cn CN CN m cn CN Os cn cn cn ON

00 r- 00 CN 00 00 00 00 00 OO 00 00 00 cn C\ CN CN O cnIS“ 8 r 9 IA V a N O W

m (N CO CN CN ON CN cn CN CN cn cn CN Ch C\ cn cn cn On

00 00 00 (N 00 OO 00 00 OO OO 00 00 00 cn o\ CN CN o cnIS“ 9T >lA V O N O k\[

m CN CO CN CN ON CN cn CN CN cn cn CN G \ cn cn cn On

oo oo OO <N OO OO OO OO OO OO oo 00 OO cn CN CN o cnIS > r 6 A V a M 0 W

m CN CO CN CN ON CN cn CN CN cn cn CN o\ cn cn cn On

o r - 00 CN 00 00 00 OO OO OO oo 00 00 cn CN CN o cnlS “ 6“ Z.AVOM01A[

i n m CN CO CN CN ON CN cn CN CN cn cn CN G\ cn cn cn ON

00 r - 00 CN 00 00 00 00 00 00 00 00 OO cn Cn CN CN o cnIS“ Z."lAVONOW

m CN CO CN CN ON CN cn CN CN cn cn CN C \ Cn cn cn cn ON

CN 00 CN 00 00 cn 00 00 00 00 CN CN 00 cn cn CN CN o cnTS>2~0ZA V (IN [nS

CO CN CO CN CN ON CN CN CN CN cn cn CN C\ cn cn cn ON

<N 00 CN 00 00 cn 00 00 00 OO CN CN OO cn cn CN CN o cnT S ~ 02~ 8 lA V O m S

CO CN CO CN CN ON CN CN CN CN cn cn CN Cn cn cn cn ON

Tf 00 CN 00 00 cn 00 00 00 00 CN CN OO cn cn CN CN o cnIS 81 91A V (IM nS CO CN CO CN CN ON CN CN CN CN cn cn CN o\ Cn cn cn cn ON

TT 00 CN 00 00 cn 00 00 OO OO CN CN 00 cn cn CN CN o cnTS” 9 r i7 lA V (M a S

CO CN CO CN CN ON CN CN CN CN cn cn CN CN cn cn cn On

fN 00 CN 00 00 cn 00 00 00 00 CN CN 00 cn cn CN CN o cni s > r 6 A v a N n s

T f CO CN CO CN CN ON CN CN CN CN cn cn CN Ch Cn cn cn cn ON

(N OO CN 00 00 cn 00 00 00 00 CN CN 00 cn cn CN CN o cni s “ 6“ /.A v a N a s

CO CN CO CN CN o CN CN CN CN cn cn CN On Cn cn cn cn On

0 4 00 04 00 00 cn 00 OO 00 00 r 4 00 rn cn CN o c^»

IS“ Z.“ lA V O M aSTt CO CN CO CN CN On CN CN CN CN cn cn CN Cn cn cn cn ON

o Tt cn m NO 00 cn r - o\ cn cn Tt NO 00 cn i n

a im CO CO CO N" N" N" m in in 00 00 00 00 ON oCN CN CNCN <N <N CN CN CN CN CN CN CN CN CN CN CN CN CN (N (N CN CN

§Q

IQ

Io(D

g

46


The average incident delay for a roadway segment can be calculated based on the

estimated incident delay caused by a specific type of incident and the correspondent

incident frequency. The average delay per vehicle can be found by dividing the delay by

the flow on the street during the time period. After computing delay for each street link

for different time periods and different types of crashes, the delay for each street link is

found by summing delay due to each type of crash. This computed delay used as an input

for the application to find the best route using the non recurrent delay.

Once the recurrent delay and non recurrent delay are computed the sum of the both

the delays is used to find the route based on the recurrent and non recurrent delay. These

are the optimal routes that are displayed in GIS using the model.

47


CD■DOQ.C

gQ.

■DCD

C/)C/)

8

ci'

Table 11 Traffic Delay Due To Non Recurrent Congestion (in seconds)

33"CD

CD■DOQ.CaO3"OO

CDQ.

■DCD

C/)C/)

400

Q

^ 1

' s 'u

fc/3

?Dh

C/) fc/3

3

l '

?C/)

00

i 'CL

c/3

o

fC/5

i '

fc/3

"^1

l '

Î<D

ÎÎ'Î

CO 00

i 'CL

i

i '

i

5

i '

Î2130213121342136 4.562143 6.132145214621482153 2.31215721592173 2.5621812183 1.78218421862194220822132215

*Note: Blank cel indicates no delay

Interface

The interface loads on clicking the Route finder button customized in the Arc Info

9 GIS suite. The interface has options for selecting the origin and the destination

locations. The application is run with a test origin and destination street intersection, the

results with the available options are shown next. The user has to also select the time of

the day for the travel with the day of the week. To select the start location, the user

selects the first cross street in the first pull down box in the “start” location (Figure 9).

Upon selection of the first street, all the crossing street for this street is populated in the

second pull down box (Figure 10). Selecting the second cross street from the pull down

box would identify the start location for the route on the map as a blue selection. Next

the user identifies the time of travel from a pull down menu (Figure 11) and the day of

the week (Figure 12).

49


Route Finder

Intersection Address

# $ # # # * * # * #

BUFFALO DRAINAGE CHL ^BURKHOLDER H.VDBURNHAM AV BURNINGWOCD LA BURNS RD BUSINESS LA DUTTONVOOD LA CSTCABANA DR CACTUS AV

anà

and

3

111

3 Da y 3

Free flow Non Recurrent

Recurrent delay Recurrent + Non Recurrent

Shotest Route

Clear Selection

Exit

FIGURE 8 Selecting Start Major Street

50


Route Finder

In tersection Address

g S u îg jÇ ln ^

BURKHOLDER BLVD a»d

3 a«d

PALO VERDE DRCAKtlLLO DRMAJOR AV CENTER ST LAKE MEAD PKWY

Tim e o f Travel

Free flow

Recurrent delay

3 D a y I 3

Non Recurrent shotest Route

Recurrent + Non Recurrent

Exit

Clear Selection

» $ » # » # # # # # # # * # * # $ » # * # # # # # # # #

FIGURE 9 Selecting Cross Street For Start Location

The same way the destination location for the route can be selected using the pull

down boxes below the End Location label. Upon selecting both the cross street the

destination intersection would be selected on the map as a blue selection.

The time of travel and the day of the travel have to be also selected for the application

to calculate the routes.

51


U IÊ Ê Ê Ê Ê Ê

1 Intersection, j Address j1

B s y î a Ç i f f i â s y ..............................................................

V 1 and ? 1

Dl rx 1 yr,.jr.;rA]n/,.-.TA':.f ''j

Time o f Travel 1 - D a y 1 3

Free flow jlam - 7am 7am - 9am int 1 Shotest Route j ,

i r r 'i ' |w f i ' ' 'i iw i iM M ii Exit jRecurrent delay | 4pm - 6pm

6pm - 8pm 8pm - Midnight

ecurrent j Clear Selection j

FIGURE 10 Selecting Time Period

52


Route Finder

In tersection j Address

BiKlCoss nand

and

Tim e o f Travel

Free flow {

Recurrent delay

In-, 2 ) D ay

Non Recurrent

Recurrent + Non Recurrent

1 2 .SundayMondayTuesday

ThursdayFridaySaturday

Exit

FIGURE 11 Selecting Day Of The Week

The User can find the following routes from a start location to destination location in

Las Vegas valley

• Free Flow Route

• Route based on recurrent delay

• Route based on non recurrent delay

• Route based on recurrent plus non recurrent delay (reliable route)

• Shortest route

53


The results of the analysis using these criteria are shown in Figure 13 to 17.If any

correction is desired or if the user desires to move back a step, the button clear selection

has to be used and the whole process has to be redone. An Exit button is provided to

unload and stop the application.

The processing time for route selection in the application varies with different start

and end location selected. As the distance increases the computation time also increases.

On a Pentium 4 personal computer with 2.4 GHz speed and 512 MB RAM the processing

was found to vary from 2 seconds to 3.5 seconds for Las Vegas Valley.

The output obtained from the system based on the selected options is shown in

Figures 12 to 16 and summarized in Tables 12 to 16.

54


LWEBÉ

MOKE

VEGf

HINGTOI*

G#GSWEST!

4LTA

CtfARLEST

MIN GO

HAR'lOAf A

^ ŒNC

HACIENDA

FIGURE 12 Free Flow Route Result

Table 12 Summary Of Free Flow Route ResultFree Flow

Origin DestinationTime of Travel

Day of Week

Distance(miles)

Time(nun)

RainbowBlvd/Russell

Decatur Blvd / Rancho 4pm - 6pm Wednesday 10.77 16.54

55


WEBd

lOKE RANCH CAREY

KE 1 EAD

V E G A . 4aHINGTOI*

G RAGSBIB

WEST< LI FF (AN K

ALTA

fCKARL

OAKEY

MINGO

HAW I

HACIENDA

R U S S

FIGURE 13 Route with Recurrent Delay

Table 13 Summary Of Recurrent Delay RouteRecurrent Delay


Day of Week

Distance(miles) Timec(min)

RainbowBlvd/Russell


56


LWEBd

lOKE RAM CH

VEGA•ta

HINGTOP

We'^ Î G RAGSÎAN k

Mi

TCHARLEST

MINQO

HAK ION

]|{ Ar

HACIENDA

FIGURE 14 Route With Non Recurrent Delay

Table 14 Summary Of Non Recurrent Delay RouteMon Recurrent Delay


Day of Week

Distance(miles)

Time(nun)

RainbowBlvd/Russell


57


LWFBri

lÜKE RANCH

HINGTOP

G M G S(AN h «

fCKARL EST

MINQO

MAR' I

H A C IEN D A

FIGURE 15 Route With Recurrent Plus Non Recurrent Delay

Table 15 Summary Of Recurrent + Non Recurrent Delay RouteRecurrent + Non recurrent Delay


Day of Week

Distance(miles)

Time(min)

RainbowBlvd/Russell


58


WEBd

M OKE RANCH

A KE U EAD

V E G f

HINGTOf

(AN h

(kLTA

fC H A R L E S T

M N Q O

HAR'

TR O ’ l<! A^ A

iACIgNpJ 7HACIENDA

FIGURE 16 Shortest Route

Table 16 Summary Of Shortest Distance RouteShortest Distance


Day of Week

Distance(miles)

Time(min)

RainbowBlvd/Russell


59


CHAPTER 5

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

A summary of the research undertaken and its findings are presented in this chapter.

A brief conclusion is presented along with recommendations for further research on the

topic and modifications that have to be incorporated.

A methodology for finding optimal routes from an origin to destination based on

historical data of recurrent and non recurrent congestion was developed in this research

using V/C ratio for recurrent delay analysis and using historical crash data for Non

recurrent delay analysis. An application was also developed based on this methodology.

It was tested using the Las Vegas valley street network data. The GIS application to

identify optimal routes based on travel time considered temporal variations such as time

of the day and day of the week. Travel time was estimated based on degree of congestion,

incident frequency and severity and non recurrent congestion. Djikstra’s algorithm was

applied in developing the algorithm.

The methodology developed has some limitations relating to the delays at nodes and

addressing the direction of travel constraint on the network. The methodology does not

take into consideration delays at intersections. Also, another limitation is in

60


addressing the direction constraints of the streets in the network. The network has to be

coded with street movement restriction to address this limitation.

In conclusion the research addressed the need to develop a methodology to identify

optimal routes based on estimated travel time. It also resulted in the development of a

GIS based tool for this purpose. The use of information and software technology was

successfully tested using a case study of the Las Vegas metropolitan area.

It is recommended that the application be fine-tuned with more real times and up to

date data for the street network. Use of more robust data analysis routines is

recommended to work on the large quantities of data volume. The traffic volumes during

different time periods have to be collected in near real time and used for analysis. Also,

the extent of disruption in traffic flow due to incidents on the streets has to be validated

and analysis has to be done to obtain more realistic reduction in capacity, resulting delays

and duration of delays. The data have to be also periodically updated to account for the

changes in traffic behavior with time. Only Congestion due to crashes was accounted for

in this research, other factors such as work zone, weather have to be also incorporated to

get more realistic results.

61


REFERENCES

Bell, lida, Y. (1997) Network Reliability, In: Transportation Network Analysis,eds. M.G.H. Bell and Y lida, pp.179-192, John Wiley & Sons, Chichester, West Sussex

Berdica, K. (2002) An introduction to road vulnerability: what has been done, is done and should be done. Transport Policy, vol. 9, Number 2, April 2002, pp. 117-127(11). Publisher: Elsevier Science, Pergamon Press, UK.

Burrough, P.A. and McDonell, R.A. (1998) Principles o f Geographic Information Systems, Ch.7, pp. 180-181, Publisher: Oxford University Press.

C. Robin Lindsey and Erik T. Verhoef (2001). Traffic Congestion and Congestion Pricing. Handbook of Transport Systems and Traffic Control, K.J. Button and D.A. Hensher (eds.), Oxford: Elsevier Science, 2001, 77-105.

Cherkassky, B., Goldberg, A. and Radzik, T. (1993) Shortest paths algorithms: theory and experimental evaluation. Technical Report 93-1480, Department of Computer Science, Stanford University, 1993.

Chou, Y.H. (1997) Exploring Spatial Analysis in Geographic Information Systems, Ch.7, pp. 215-264, OnWord Press, 1997.

Deo, N. and Pang, C.-Y. (1984) Shortest path algorithms: taxonomy and annotation. Networks, vol. 14, pp. 275-323

Dijkstra, E. W. (1959), A note on two problems in Connection with graphs, Numerische Mathematik, vol. 1, 1959, pp. 269-271

Douglas, D.H. (1994) Least-cost Path in GIS Using an Accumulated Cost Surface and Slopelines, Cartographica, vol.31, no. 3, pp. 37-51

Dowling, R., Kittelson, W., Zegeer, J., and Skabardonis, A. (1997). Planning Techniques to Estimate Speeds and Service Volumes for Planning Applications. NCHRP Report 387. Washington D C.: National Academy Press.

Dowling, R. and Skabardonis, A., Improving the Average Travel Speeds Estimated By Planning Models, presented at 71st Annual Meeting of the Transportation Research Board, January 1992.

62


Guide 93, Programming in ArcGIS using ArcObjects and AML : May 2003 University of Durham Information Technology Service, Durham.

Husdal, J. 2004. Why reliability and vulnerability should be an issue in road development projects. Molde University College, 2004.

Husdal, J. 2000. Fastest path problems in dynamic transportation networks. University of Leicester, UK, 2000.

Husdal, J. (2004) Reliability and vulnerability - a non issue in cost-benefit analyses?. Samferdsel 3/2004, pp 28-30. Norwegian Institute of Transport Economics, Oslo, Norway.

lida, Y. (1999). Basic Concepts and Future Directions o f Road Network Reliability Analysis. Journal of Advanced Transportation, Vol. 33, No. 2, Pg. 125-134, Institute of Transportation, Alberta, Canada.

Jacob, R., Marathe, M. V. and Nagel, K. (1998) A computational study of routing algorithms for realistic transportation networks, 2nd Workshop on Algorithmic Engineering (WAE’98), Saarbriicken, Germanz, August 19-21 1998

Lindley, J. (1987) Urban Freeway Congestion: Quantification o f the Problem and Effectiveness o f Potential Solutions. ITE Journal, No. 57, Pg. 27-32. Institute of Transportation Engineers, Washington, D.C.

Lindley, J., A Methodology fo r Quantifying Urban Freeway Congestion, Transportation Research Record 1132, Transportation Research Board, Washington, D C., 1987.

LLeras-Echeverri, G. and Sanchez-Silva, M. (2001) Vulnerability analysis of highway networks, methodology and case study. Transport 147 (4), pp.223- 230. November 2001

Lombard, K. and Church, R.L. (1993) The Gateway Shortest Path Problem: Generating Alternative Routes for a Corridor Routing Problem, Geographical Systems, vol. 1, pp. 25-45

Lupien, A.E., Moreland, W.H. and Dangermond, J. 1987. Network analysis in geographical information systems. Photogrammetric Engineering and Remote and Sensing 53:1417-1421

Mainguenaud, M. (19951 Modeling the network component o f geographical information systems, International Journal of Geographical Information Systems, vol. 9, no. 6, pp. 575-593

NDOT (2003). Nevada Traffic Crashes (1998 - 2002). Published by the Nevada Department of Transportation Safety Engineering Division in cooperation with the

63


Nevada department of Motor Vehicles and Public Safety, and State and Local Law Enforcement Agencies.

Nicholson, A.; Schmocker, J.-D.; Bell, M.G.H. and lida, Y (2003) Assessing Transport Reliability: Malevolence and User Knowledge. In: The Network Reliability of Transport, eds. M.G.H. Bell and Y. lida. Proceedings of the L* International Symposium on Transportation Network Reliability, pp.I-22. Elsevier Science, Kidlington.

ORA, KHA, UNLV-TRC (2003). Development o f the Las Vegas Valley Congestion Management System. Draft Report, Prepared for the Regional Transportation Commission of Southern Nevada. Orth Rodgers & Associates, Kimley Horn & Associates and Transportation Research Center, University of Nevada Las Vegas.

Parimi, V. (2004) Determination of Reliability Indices for Congestion on Urban Arterial Streets Masters Thesis, Department of Civil & Environmental Engineering, University of Nevada Las Vegas.

Presley, W.M., Wyrosdick K.G. (1999). Calculating Benefits for NAVIGATOR: Georgia’s Intelligent Transportation System. 6th World Congress on Intelligent Transport Systems, Toronto, Canada, Intelligent Transportation Systems Society of America.

Pulugurtha, S. S., Nambisan S. S. (2001). Development of Urban Transportation Congestion Management and Mitigation Measures. Report: UNLV/TRC/RK-01/02, Transportation Research Center, University of Nevada Las Vegas.

Pulugurtha, S. S. and S. S. Nambisan. Estimating Reliability and Extent for the Las Vegas Area Congestion Management System. Proceedings of the Eighth International conference of Applications of Advanced Technologies in Transportation, Beijing, China, May 26-28, 2004. Published by the American Society of Civil Engineering, Reston, Virginia, pp. 23-27.

Ravindra K. Ahuja, Thomas L. Magnanti, and James B. Orlin. Network Flows: Theory, Algorithms, and Applications. Prentice Hall, 1993

Schrank D, and T. Lomax (2004). The 2004 Aimual Mobility Report Information For Urban America,Texas Transportation Institute, Texas A&M University System, College Station, Texas httn.V/mobilitv.tamu.edu.

Sullivan, E.C. (1997). New Model for Predicting Freeway Incidents and Incident Delays. Journal of Transportation Engineering, Vol. 123, Issue. 4, Pg. 267-375, American Society of Civil Engineers, Reston, Virginia.

Taylor, M. A. P. and D ’Este, G.M.D. (2003a) Network Vulnerability: An Approach to Reliability Analysis at the Level of National Strategic Transport Networks. In: The Network Reliability of Transport, eds. M.G.H. Bell and Y. lida. Proceedings of the 1st

64


International Symposium on Transportation Network Reliability, pp.23-44. Elsevier Science, Kidlington.

Taylor, M. A. P. and D ’Este, G.M.D. (2003b) Concepts of network vulnerability and applications to the identification of critical elements of transport infrastructure. Paper presented at the 26th Australasian Transport Research Forum, Wellington, New Zealand, 1-3 October

Thill, J.F., Rogova, G.L., Yan, J. (2001). Delay Modeling fo r Corridor-Level Benefits Evaluation o f VMS Deployments. 80'** Annual Meeting of the Transportation Research Board, National Research Council (U.S.), Washington, D.C.

Van Eck, J.R. and De Jong, T. (1990), Adapting data structures and algorithms fo r faster transport network computations, Proceedings of the 4th int. symposium on spatial data handling, vol.l, pp. 295-304

Yi, Q. (2002). Incident Delay Analysis and Estimation. Ph.D. Dissertation, Department of Civil Engineering, Polytechnic University of New York.

Zhan, F. B. (1997) Three fastest shortest path algorithms on real road networks: Data structures and procedures. Journal of Geographic Information and Decision Analysis, vol.l, no.l, pp. 69-82 http://publish.uwo.ca/~jmalczew/gida_l/Zhan/Zhan.htm

Zhan, F. B. and Noon, C. E. (1996) Shortest Path Algorithms: An Evaluation using Real Road Transportation Science vol. 32, no. 1, pp. 65-73

Zhou Q., Yang X. and Melville M. D. (1996j A GIS Network Model for Sugarcane Field Drainage Simulation, in Proceedings of 8th Australasian Remote Sensing Conference, 25-29 March, Canberra, Vol. 2, pp 366-372.

www.esri.com (accessed on January 15^\ 2005)

www.wikipedia.com (accessed on November 23 *, 2004)

65


http://publish.uwo.ca/~jmalczew/gida_l/Zhan/Zhan.htm

http://www.esri.com

http://www.wikipedia.com

VITA

Graduate College University of Nevada, Las Vegas

Rajeev Naroor Graduate Student

Home Address:2511 N Green Valley Pkwy 2511 Henderson, NV 89014

Degree:Kerala University, IndiaBachelor of Engineering (Civil Engineering), November 1998

Professional Affliation:Student Member, ITE UNLV Chapter. 2003 to present Secretary, ITE UNLV Chapter, August 2003-July 2004

Presentation:Naroor Rajeev and Shashi Nambisan - “Queuing at Gated Communities”, 13 Annual Fall Transportation Conference, Sponsored by ASCE Nevada Chapter, by ASCE Nevada Chapter,lTE Nevada Chapter, ITS America Nevada Chapter and the UNLV Transportation Research Center, Las Vegas, Nevada. Oct 20-21,2005

Thesis Title:A GIS Based Tool To Determine Optimal Routes

Thesis Examination Committee:Chairperson, Dr. Shashi S. Nambisan, Ph. D., P. E.Committee Member, Dr. Mohamed S. Kaseko, Ph. D.Committee Member, Dr. Srinivas S. Pulugurtha, Ph. D.Graduate Faculty Representative, Dr. Wolfgang Bein, Ph. D.

jth

66


a gis based tool to determine optimal routes

Documents