The acceleration profile of leading vehicles at intersections is critical for emission estimation and microlevel queue simulation. Data obtained from experiments using a high-resolution driving simulator can deliver useful insights into microscale acceleration behaviors at signalized intersections. Acceleration data of the leading vehicles in queues are collected by the simulator. The observed accelerations are found to be stochastic. The acceleration characteristics are also significantly diversified among participants. Hence, a Markov chain is implemented to simulate the acceleration behaviors. The acceleration data are classified into varied operation states. And the Markov chain reconstructs the acceleration profiles of leading vehicles and reproduces the randomness of acceleration behaviors. Among numerous candidate profiles, a speed profile is selected by a proposed criterion that represents the typical acceleration behaviors at signalized intersections.
This research was partially supported by the Natural Science Foundation of China (NSFC) # 51678045, the Henan Department of Transportation project # 2018G3, and the research grant of the Kongju National University in 2020.
