Simulation of turning vehicles’ behaviors at mixed-flow intersections based on potential field theory

The classical lane-based one-dimensional simulation models cannot describe the complex features of turning vehicles including variation of trajectories and shared-priority at mixed-flow intersection (MFI). This paper proposes a quasi-two-dimensional model to simulate turning vehicles’ behaviors at MFI. The proposed model consists of three layers: decision layer, plan layer and action layer. The decision layer firstly employs discrete choice model to decide whether to turn or not. The plan layer plans the desired path to destination considering interactive objects by using distance potential field. The action layer implements movement of turning vehicles under the impact of the artificial potential field. The proposed model is tested by reproducing trajectories of left-turn vehicles at a two-phase MFI in Shanghai. As a result, the simulation reproduces the variation of trajectories, while the overlapping ratio of trajectory is 75.02%. Meanwhile, the travel time of simulation and empirical data do not show statistically significant difference.