Using Artificial Potential Field Theory for a Cooperative Control Model in a Connected and Automated Vehicles Environment

Abstract

Artificial potential field (APF) theory has been extensively applied in traffic path planning as an efficient method to avoid collision. However, studies in collision avoidance based on APF theory only considered the movement of single vehicle. In this paper, a vehicle cooperative control model for avoiding collision in the connected and autonomous vehicles (CAVs) environment is presented, using APF theory. The proposed model not merely guarantees the travel safety of vehicles in avoiding collision, but also promotes driving comfort and improves traffic efficiency. To verify the cooperative control model, simulations of four scenarios are designed and compared with the human driving environment. Five indicators are selected to evaluate the results, that is, time–space diagram, time mean speed (TMS), the rate of large deceleration time (large deceleration is that deceleration larger than –2 m/s²), the inverse time-to-collision ( ${\mathrm{TTC}}^{-1}$ ), and lane-changing times. According to the simulation results, the cooperative control model could alleviate the capacity drop and increase the TMS to improve traffic efficiency, reduce the rate of the large deceleration time to promote driving comfort, and decrease ${\mathrm{TTC}}^{-1}$ to promote safety in small and large input flow rates. The results reveal the proposed model is significantly superior to the human driving environment whether in free or congested situations, except for the lane-change times, which are slightly larger.