Optimization method for coordinated control strategy of IEEB in electric vehicles based on digital twin

The Intelligent braking system with energy-regenerative and electro-mechanical braking (IEEB) system offers significant advantages for electric vehicles, including energy recovery capability and rapid response, making it a promising solution for alleviating range anxiety and improving braking energy efficiency. To fully leverage the potential of the IEEB system in electric vehicles, this paper proposes an optimization method for IEEB coordinated control based on a digital twin (DT) framework. First, a DT of a vehicle equipped with a distributed IEEB system was constructed, integrating both physics-based and data-driven models. This DT serves as a dynamic simulation environment for training control strategies, replacing the traditional static physics-based model (PM) and effectively bridging the gap between simulation and real-world conditions. Next, a twin delayed deep deterministic policy gradient algorithm with hybrid prioritized experience replay (TD3-HPER) is introduced. This method incorporates temporal difference error, state visitation frequency, and experience age into the TD3 framework, leading to faster convergence and enhanced stability. Using the vehicle DT environment, the TD3-HPER algorithm was applied to optimize the speed threshold and regenerative factor within the coordinated control strategy. Simulation results show that the proposed method (TD3 with DT and HPER) achieves 107 and 152 higher reward values with faster convergence, compared to TD3 with PM and PER, and TD3 with DT and PER. Hardware-in-the-loop (HIL) tests under various conditions confirm the DT's role in reducing the simulation-reality gap and the performance improvements attributable to HPER. Real vehicle testing confirms the strategy's effectiveness, achieving smooth braking transitions (maximum jerk: 10.19 m/s3) and high energy efficiency (SOC decrease: only 0.89 %).