Research on coordinated control of electro-hydraulic composite braking for an electric vehicle based on the Fuzzy-TD3 deep reinforcement learning algorithm

In order to improve the energy utilization efficiency of electric vehicles, alleviate range anxiety, and ensure braking stability and comfort, a coordinated control strategy of electro-hydraulic composite braking (EHB) is proposed based on the fuzzy twin delayed deep deterministic policy gradient (Fuzzy-TD3) algorithm. A mathematical EHB system model is established. A particle swarm backpropagation (PSO-BP) neural network is used to determine the braking intensity of driver. Combined with the Fuzzy-TD3 algorithm to optimize the distribution of regenerative braking torque and hydraulic braking torque under normal braking, efficient recovery of braking energy is achieved to ensure braking stability and comfort. For emergency braking, the coordinated control of the anti-lock braking system (ABS) and the regenerative braking system (RBS) is realized by combining the sliding mode control (SMC) and the Fuzzy-TD3 algorithm. This effectively lowers the risk of wheel slip during emergency braking and enhances safety and ride comfort. The results show that compared to the conventional PID control method, the Fuzzy-TD3 control strategy lowers braking time by 11.5 % and 9.5 % under normal and emergency braking conditions, respectively. Additionally, the state of charge (SOC) of the battery increases by 0.487 % and 0.266 %, respectively. These findings are consistent with experimental data and validate the effectiveness of this strategy in improving braking performance and energy recovery efficiency.