Abstract: To investigate the frequent switching issues in energy recovery modes and the challenges in simultaneously optimizing braking performance and energy recovery efficiency for rear-wheeldrive battery electric vehicles, this study conducts a comprehensive parametric compatibility analysis focusing on core powertrain components of a specific pure electric vehicle. An adaptive energy recovery strategy employing back propagation (BP) neural network algorithm is designed to accommodate diverse driving conditions, while fuzzy control theory is integrated to identify the optimal balance between braking performance and energy recovery efficiency. The proposed control strategy underwent rigorous validation through a co-simulation platform combining Simulink and AVL-Cruise. Simulation results conclusively demonstrate that the developed strategy enhances energy recovery efficiency by 1 percentage point during braking while maintaining vehicle driving smoothness, accompanied by a 0.1 percentage point reduction in state of charge (SOC) fluctuation, thereby achieving the optimal equilibrium between braking performance and energy recovery efficiency.

Key words: power matching; energy recovery; fuzzy control; neural network

摘要： 为了研究后驱式纯电动汽车在能量回收模式中存在的频繁切换问题，以及难以同时优 化制动性能与制动能量回收效率的挑战，研究聚焦于某款纯电动汽车，针对其核心动力系统 的关键组件，展开了详尽的参数适配性分析。采用反向传播（BP）神经网络算法设计了一种 自适应于不同驾驶工况的能量回收策略，同时结合模糊控制理论找出制动性能与能量回收效 率的最佳平衡。通过 Simulink 与 AVL-Cruise 联合仿真平台，对所设计的控制策略进行了详尽 的验证。仿真结果清晰地展示出，该研究提出的策略在确保车辆行驶平顺性的基础上，制动 过程中的能量回收效率提高了 1 个百分点，荷电状态（SOC）波动下降 0.1 个百分点，实现制 动性能与能量回收效率之间的最佳平衡。

关键词: 动力匹配；能量回收；模糊控制；神经网络