Multi-perspective system-wide analyses of adaptive traffic signal control systems using microsimulation and contemporary data sources
The primary function of traffic signals is to assign the right of way to vehicular and pedestrian traffic at intersections. Effective traffic signal system reduces congestion, increases intersection capacity, and improves other traffic related performance measures such as safety and mobility. To ensure these goals are met, traffic signals require updated timings to maintain proper operation. These updated signal timings impact not only traffic performance, but overall transportation system efficiency. Because traditional signal timing plans may not accommodate variable and unpredictable traffic demands, a more proactive approach is necessary to ensure properly timed and maintained traffic signals. Adaptive traffic control systems (ATCS) continually collect data and optimize signal timing on a real time basis thereby reducing the aforementioned drawbacks of traditional signal retiming. Understanding and characterizing how these systems are working is important to transportation engineers, and evaluating these systems can provide useful insights. The objective of this dissertation is to develop evaluation methodologies (both operational and economical) for adaptive traffic signal control that go beyond the traditional assessments that use traffic measures of effectiveness (MOEs). Case studies are conducted for Sydney Coordinated Adaptive Traffic System (SCATS) implementations in Alabama, which are useful in objective evaluations of ATCS (in general) for both their current and future operational environments by using microsimulation techniques and/or field data from contemporary data sources. The study contains detailed comparative analyses of traffic operations of the study corridors for existing peak hour traffic conditions under the previous time-of-day (TOD) plan and similar peak hour conditions after SCATS implementation. Although simulation analysis using VISSIM traffic microsimulation software is the primary methodological technique used for evaluating comparative performances, arterial data from other sources (Bluetooth MAC Address Matching and crowdsourced travel data) are also used to perform the evaluations, which is a novel application for this context. While past studies have considered either the arterial or its side-streets performances in their evaluations, this work explored a system-wide approach looking at the composite performance of both dimensions together. Finally, for transportation agencies which operate within budget constraints, it is important to know the real worth of attaining the benefits from ATCS implementations. The last chapter of this dissertation extends the evaluation methodology to include benefit-cost analysis (BCA) by evaluating the ATCS performance for both current and future traffic conditions. This information will be helpful for transportation agencies, planners, and practitioners to understand and justify their ATCS investment and also serve as a guideline for their future ITS projects.