This research project included work in several areas related to the operation of traffic signal systems in oversaturated traffic conditions. Quantitative measures were developed that characterize the intensity of oversaturated conditions and can be calculated based on high-resolution data from advance detectors and second-by-second phase timing data. The technique measures the length of the overflow queue at the beginning of the red light and estimates the amount of green time that is wasted either by dispersing this overflow queue or by vehicles that cannot proceed due to a downstream blockage. A heuristic procedure was developed to process the measurements of oversaturation and compute green time modifications on an oversaturated route. A methodology was developed to design timing plans that can improve system throughput and manage queues in oversaturated conditions. The methodology considers three regimes of operation of an oversaturated scenario: (1) loading, (2) processing, and (3) recovery. Individual timing plans are designed for each regime and the switching time between the plans is also determined by the methodology. It was found that application of strategies that maximize throughput in the loading and recovery phases of operation can result in performance improvements over timing plans designed for under-saturated operation. Using timing plans that are designed to minimize the degree of saturation on critical routes (i.e. manage queues) are most effective during the processing regime. An online logic tool was also developed that can directly use the quantitative oversaturation intensity measures for real-time, traffic-responsive application of pre-configured mitigation strategies. Finally, practitioner guidance was developed to assist in the process of matching mitigation strategies with specific oversaturated conditions scenarios using a systems engineering approach. Six simulation test cases were executed to evaluate and document the benefits of the techniques. In general the research found that there are tangible performance improvements that are achievable in certain types of oversaturated scenarios by applying mitigation strategies that consider overflow queues. (Author/publisher)
Samenvatting