Research on Longitudinal–Vertical‐Coordinated Control Strategy for Pure Electric Vehicles Based on the TD3 Algorithm

To alleviate range anxiety and improve braking stability and ride comfort, this article proposes a longitudinal–vertical‐coordinated control strategy based on the twin delayed deep deterministic policy gradient (TD3) algorithm for a dual‐axle‐driven pure electric vehicle. Considering the coupling effects between the electro‐hydraulic composite braking system and the suspension system, a longitudinal–vertical‐coupled dynamics model is established. An adaptive control scheme is designed to address varying control objectives under general and emergency braking conditions, achieving global optimal control for regenerative braking energy recovery, braking stability, and ride comfort. A comparison of the model predictive control‐proportional integral derivative (MPC‐PID) and deep deterministic policy gradient (DDPG) control strategies shows the performance advantages of the TD3‐based control strategy. The simulation results show that the TD3‐based strategy achieves energy recovery efficiencies of 34.16 and 34.78% under general and emergency braking, respectively, while reducing braking time by 13.4 and 9.8% compared to PID control. In suspension control, the TD3 algorithm effectively suppresses body and pitch angular acceleration, outperforming MPC and DDPG strategies. The experiment confirms the effectiveness of the proposed control strategy in improving energy recovery and ride comfort.