Car-Following Control Using Recurrent Cerebellar Model Articulation Controller

Car following for traffic safety has been an active area of research. However, when a human is driving, it involves reaction time, delay, and human error that adversely affects safe driving. One way to eliminate human error and delay in vehicle following is to integrate an automated car-following control system into the driving system. This paper presents a recurrent-cerebellar-model-articulation-controller (RCMAC)-based control system for the car-following control of a platoon of automated vehicles. In this control system, an RCMAC is the main controller that is used to mimic a perfect control law, and a compensation controller is designed to compensate for the difference between the perfect control law and the RCMAC. The online adaptive laws of the control system are derived from the Lyapunov stability theorem so that the stability of the system can be guaranteed. Finally, a comparison of a sliding-mode control (SMC), an intelligent adaptive SMC, and the proposed RCMAC-based control for the car-following control is made. Simulation results demonstrate that the proposed RCMAC-based control system can achieve better control performance than other control systems.