Asymmetric repulsive force model: A new car-following model with psycho-physical characteristics

Abstract

Dynamic traffic phenomena, such as hysteresis, stop-and-go traffic, stability, and capacity drop, have significant implications for traffic safety and congestion. Previous studies have suggested that these phenomena are caused by human behavior. However, existing models have not been able to fully reproduce these dynamic traffic phenomena. The focus of our research is on psycho-physical properties and asymmetric behavior, among other human factors, which provide a theoretical explanation for dynamic traffic phenomena. We propose a new car-following model, Asymmetric Repulsive force Model (ARM), based on the common fundamental mechanisms of the two theories. ARM is evaluated from both microscopic and macroscopic perspectives. ARM exhibited similar driving behavior similar to NGSIM data in the time-distance, speed, acceleration, speed-spacing, and spacing-relative speed domains. Additionally, ARM outperforms comparable models in quantitative evaluations of describing trajectory and hysteresis. From the macroscopic perspective, platoon simulation was performed to analyze stop-and-go situations. The results confirmed that ARM can simulate the processes of speed drop, speed recovery and flow non-recovery, and speed and flow recovery. This finding is consistent with previous research on capacity drop using empirical data and suggests that ARM has the potential to simulate various macroscopic traffic phenomena.