An Adaptive Semi-Active Suspension System for Design of Vehicle Stability Control System

In this paper, an adaptive semi-active suspension system by adjusting air springs is proposed to design of vehicle stability control system, which can achieve handling stability. In order to develop the adaptive control algorithm, the CarSim dynamic model is acquired, and a reinforcement learning scheme using the deep deterministic policy gradient (DDPG) control strategy is used to decide the best amount of air spring pressure. This method is able to adjust the yaw activity to avoid understeer and oversteer issues as well as increasing the vehicle stability. In summary, the simulation results show that the handling stability is improved by 25.78
^%
under high road curvatures.