The University of Massachusetts Amherst
University of Massachusetts Amherst

Search Google Appliance


Person-Based Signal Timing Optimization to Account for Flexible Cycle Lengths and Uncertain Transit Vehicle Arrivals

TitlePerson-Based Signal Timing Optimization to Account for Flexible Cycle Lengths and Uncertain Transit Vehicle Arrivals
Publication TypeConference Proceedings
Year of Publication2017
AuthorsYu Z, Gayah VV, Christofa E
Conference NameTransportation Research Board 96th Annual Meeting
Date Published01/2017
Accession Number01622846
KeywordsAlgorithms, Arrivals and departures, Bus transit, Green interval (Traffic signal cycle), Programming (Mathematics), Traffic delays, Traffic signal cycle, Traffic signal priority

Recent studies have proposed using person-based traffic signal timing optimization frameworks to minimize total passenger delay experienced by passenger cars and transit vehicles at signalized intersections. However, the efficiency and the practical application of existing efforts is limited by the assumption of fixed cycle lengths and deterministic bus arrival times. This paper extends the algorithms developed in previous work for isolated intersections to accommodate flexible cycle lengths and uncertain bus arrivals. Flexible cycle lengths are accommodated by minimizing total passenger delay within some fixed planning horizon that allows cycle lengths to vary within a feasible range. Two methods are proposed to accommodate uncertain bus arrival times: 1) a robust optimization approach that seeks to achieve a “best” worst-case scenario; and, 2) a rule-based strategy that applies green extension to signal timings obtained from deterministic optimization. The proposed strategies are tested using numerical simulations of an intersection in State College, PA. The results reveal that the flexible cycle length algorithm can significantly reduce bus passenger and total passenger delay with negligible increases in car passenger delay. These results are robust to both the bus and car flow expected at the intersection. The robust optimization strategy appears to reduce the additional passenger delay generated by bus arrival uncertainty for low uncertainty levels, while the rule-based strategy performs better for larger uncertainty levels when intersection flow ratios are low. The anticipated benefits decrease with the intersection flow ratio due to the inflexibility of signal timings at the intersection.