accepted version (pdf 2mb)

Download Accepted Version (PDF 2MB)

Post on 11-Feb-2017




1 download

Embed Size (px)


  • This is the author version published as: This is the accepted version of this article. To be published as : This is the authors version published as: Catalogue from Homo Faber 2007

    QUT Digital Repository:




    Control of Aircraft for Inspection of LinearInfrastructure

    Troy S. Bruggemann, Jason J. Ford and Rodney A. Walker, Member, IEEE

    AbstractInspection aircraft equipped with cameras and othersensors are routinely used for asset location, inspection, mon-itoring and hazard identification of oil-gas pipelines, roads,bridges and power transmission grids. This paper is concernedwith automated flight of fixed-wing inspection aircraft to trackapproximately linear infrastructure. We propose a guidancelaw approach that seeks to maintain aircraft trajectories withdesirable position and orientation properties relative to theinfrastructure under inspection. Furthermore, this paper alsoproposes the use of an adaptive maneuver selection approach,in which maneuver primitives are adaptively selected to improvethe aircrafts attitude behaviour. We employ an integrated designmethodology particularly suited for an automated inspectionaircraft. Simulation studies using full nonlinear semi-coupled sixdegree-of-freedom equations of motion are used to illustrate theeffectiveness of the proposed guidance and adaptive maneuverselection approaches in realistic flight conditions. Experimentalflight test results are given to demonstrate the performance ofthe design.

    Index TermsAircraft control, adaptive maneuver selection,guidance, linear infrastructure, power line inspection.


    A IRCRAFT equipped with cameras and other sensors canbe used in a significant number of civilian applicationssuch as remote mapping [1], geolocation and feature tracking[2] and remote sensing [3]. Of particular importance amongstthese remote sensing applications, are the asset inspectiontasks which are costly, time-consuming and tedious (especiallywhen assets are extensive, sparse, or difficult to locate oraccess). One commonly appearing class of inspection tasksis those involving approximately piecewise linear assets suchas oil-gas pipelines, roads, bridges, power-lines, power gen-erate grids, rivers, coastlines, canals, highways and forest fireboundaries [4], [5], [6], [7], [8], [9], [10], [11], [12]. Flyingmanned or unmanned aircraft fitted with appropriate cameraand sensor payloads taking imagery of such infrastructureassets, could save countless man-hours and costs normallyassociated with asset management, thereby improving the ef-ficiency of the operations. Although low-altitude linear infras-tructure inspection applications such as power line surveillance

    T.S. Bruggemann is with the Cooperative Research Centre for SpatialInformation and the Australian Research Centre for Aerospace Automation,at the Queensland University of Technology, Brisbane, QLD, 4001 Australia(e-mail:

    J.J. Ford is with the Cooperative Research Centre for Spatial Informa-tion and the Australian Research Centre for Aerospace Automation, at theQueensland University of Technology, Brisbane, QLD, 4001 Australia (

    R.A. Walker is with the Cooperative Research Centre for Spatial Infor-mation and the Australian Research Centre for Aerospace Automation, at theQueensland University of Technology, Brisbane, QLD, 4001 Australia (

    have been studied since the mid-1990s, the need for furtherremote sensing automation continues to be highlighted bymany authors [3], [9], [10], [11], [12], [13], [14], [15].

    The purpose of this paper is to investigate flight automationissues for low-altitude fixed-wing inspection platforms track-ing linear infrastructure. One key issue for linear infrastructureinspection involves determining a suitable aircraft controlapproach that ensures aircraft flight with a fixed relativeposition and relative body attitude with respect to the linearinfrastructure under inspection. In typical operation, there isa requirement for downward-looking body-fixed cameras tocapture the objects on the ground within their fixed and limitedfield of view [3], [4], [5], [7], [8], [16], [17].

    The principle of operation is that after appropriate flightpath selection, the field of view seen by the camera will beflown down the length of the infrastructure under inspection.Unfortunately, any aircraft roll, pitch or yaw will translate theassets imaged position, and perhaps even move the asset outof the cameras field-of-view. For this reason, minimising roll-motion is an important issue faced by inspection aircraft withbody-fixed cameras. Even for inspection aircraft equipped withgimballed cameras, aircraft attitude motion is undesirable [17].

    The planning and control of fully autonomous platformsis typically separated into three sub-problems: the trajec-tory planning sub-problem [19], the guidance sub-problem oftracking the planned trajectory [7], [20], and the autopilotor maneuver sub-problem of following the issued guidancecommands [20], [21]. Generic versions of the first of these sub-problems, the trajectory planning problem, have been studiedby several researchers and many of the techniques developedcan be readily applied to infrastructure inspection. Theseinclude explicit planning techniques [19], [17], [22], motionprimitive planning techniques [18], and implicit techniquessuch as those involving virtual waypoints [23].

    Early investigations of the second sub-problem, the guid-ance problem, illustrated that simple PID based control loops(directly based on GPS derived tracking errors) leads to poorcross-track position and velocity performance [4], [24]. In[24], new heading error rate type lateral track controllerswere proposed for the Aerosonde UAV and these approacheswere illustrated to reject certainty types of wind disturbances.In [4], hardware-in-the-loop experiments illustrated the useof combining image-based information and heading errorrate controllers in the problem of tracking road infrastruc-ture. Others [7], [16] have proposed a biased proportionalnavigation (BPN) guidance approach that used image-basedmeasurements for road tracking. Alternatively, vector fieldbased [25] and Lyapunov approaches [26], have been proposed


    in a number of slightly modified tracking problems. However,few of these previous guidance approaches rigorously considerboth the relative position and orientation of the platform.Furthermore, most fail to consider the impact of maneuverchoice on trajectory tracking performance, nor the impact thatany tracking error has on the underlying inspection activity.

    The third sub-problem, the maneuver problem, has par-ticular importance in the inspection application because air-craft maneuver behaviour can have a significant impact onthe inspection task because heading corrections are typicallyindirectly actuated through aerodynamically efficient bank-to-turn (BTT) maneuvers. That is, in a standard fixed-wing air-craft autopilot configuration, heading adjustments are achievedthrough banking the aircrafts airframe [27]. Clearly, bankingmaneuvers are a problematic strategy for inspection platformsbecause the banking action risks changing the cameras field-of-view too much, causing the asset to no longer be underinspection [8]. One possible avenue for reducing the apparentconflict between heading requirements and banking behaviouremerges from the missile control community where systemsare often designed to exploit different maneuver behavioursduring different mission phases. Specifically, a missile mightgain-schedule between aerodynamically efficient BTT maneu-vers during early stages and skid-to-turn (STT) maneuversfor small corrections during terminal stages [28], [29]. Thisadaptive maneuver selection approach allows the differentcharacteristics of various maneuver regimes to be exploitedduring different flight stages.

    Each of these three sub-problems could be considered insequential manner. However, separate sub-system design doesnot often lead to acceptable performance because it cannotexploit any beneficial relationships between the sub-systems[20].

    This paper makes three contributions to the autonomousaircraft inspection problem. The first contribution is to proposethe use of a precision guidance law that commands towardsflight trajectories with desirable position and heading relativeto the infrastructure under inspection [30], [31]. This guidancelaw was originally designed for the purpose of controllinga missile to impact a target at a desired impact angle. Incontrast, we investigate the performance of this guidance lawfor inspection of linear assets where the aim is to interceptand track a desired line of inspection.

    The second contribution is to propose an adaptive maneuverapproach, inspired by hybrid autopilot designs used in missileautopilots, which aims to achieve guidance commands whilstmaintaining desirable attitude behaviour [28], [29]. The inclu-sion of adaptive maneuver behavior within the aircraft controlsystem allows a greater range of performance characteristicsto be exploited. We highlight that care must be taken toensure that flight stability remains a priority, but these low-level platform-specific issues are not explicitly considered inthis paper. However, full autopilot dynamics were used in oursimulation studies.

    The third contribution is our integrated system design forautonomous control of inspection aircraft