Use of Connected Vehicle Data to Identify Signal Timing Plans on Signalized Arterial Corridors

This study demonstrates an approach for applying connected vehicle (CV) trajectory datasets to identify the cycle length, time-of-day (TOD) plan, and distributions of green times of coordinated phases in signalized arterial corridors. Current methods of measuring performance and controlling traffic signal systems require intersections to be equipped with vehicle detectors and communication devices, which require significant resources for both implementation and maintenance. Several commercial providers have recently started marketing high-fidelity CV trajectory data sourced from auto manufacturers. This paper presents a methodology that can identify key parameters of signal timing plan using CV data. A virtual detection technique for identifying cyclic patterns of when traffic is flowing is the core of the method. Change point detection is used to identify the TOD plan when the flow patterns change. The proposed approach is implemented in a real-world signalized corridor in Dubuque, Iowa, yielding satisfactory outcomes in parameter estimation. Such information can be employed scalably to identify signal timing plans, which is a starting point for developing automated processes that can ultimately improve signal timing.