Modeling Drivers' Dynamic Decision-Making Behavior During the Phase Transition Period: An Analytical Approach Based on Hidden Markov Model Theory

摘要

A flashing green indication of 3 s followed by a yellow indication of 3s is commonly applied to end a green phase at signalized intersections in many Chinese cities. This paper proposes an analytical approach based on the hidden Markov model theory to interpret the dynamic decision-making process of drivers during the phase transition period at high-speed signalized intersections. In the proposed model, the hidden states are the unobservable time-dependent decisions of drivers concerning whether to stop or pass, and the observable states are the instantaneous vehicle speeds and acceleration/deceleration rates that are obtained from the high-resolution vehicle trajectory data. The data were collected by videotaping four typical high-speed intersection approaches with a speed limit of 80 km/h in Shanghai. Eventually, 698 vehicle trajectories including 179 trucks and 519 passenger cars were extracted from the videos and used for model estimation and validation. It was found that the proposed model could predict stop–pass decisions with very high accuracy and revealed that approximately 50% of drivers used a two-step decision-making process. In addition, a large percentage of decision changes occurred 0–1.2 s after the onset of yellow, which is based on a driver's perceived environment. The important implications of the proposed model and the findings are also discussed in this paper.