Signal Offset Optimization Using Vehicle Trajectory Data with Different Sampling Frequencies

With the development of information technology and connected vehicles technology, vehicle trajectory crowdsourcing is becoming more prevalent. Detailed vehicle trajectory data provides a new means of signal offset optimization on urban arterials. One critical factor is the sampling frequency (SF) of probe vehicles, which is can significantly impact communication cost and signal offset optimization. Thus, in this study, these issues were examined using field DiDi vehicle trajectory data collected in one arterial of Shenzhen, China. Given an average penetration rate of DiDi vehicles of 7.4%, four SFs, i.e., 1/3, 1/6, 1/9, and 1/12 Hz are set. Signal offset was optimized by examining the average vehicle delay at each intersection. It was found that the performance of offset optimization is enhanced with increased SF but becomes less insensitive when SF is below 1/6 Hz.