BELM was proposed based on approach to predict the battery's internal states of HEVs.The basic principles of BELM were described in detail,and the performance parameters of bat-tery were collected under advanced vehicle simulator(ADVISOR)including voltages,currents,tem-peratures and so on.Then the BELM was used in the predictions of SOC and SOH,at the same time, the influences of aging battery were taken into account.The results of BELM indicate that the predic-tion model possesses higher prediction accuracy,and achieves real-time and accurate SOC and SOH predictions with high effectiveness.%针对混合动力汽车(HEV)电池内部状态预测问题，引入贝叶斯极限学习机(BELM)方法.对BELM的基本原理进行了详细介绍，在高级车辆仿真软件 ADVISOR 中采集 HEV电池的各项性能参数，包括电压、电流、温度和内阻等.基于此，将 BELM应用于电池的荷电状态(SOC)和健康状态(SOH)的预测，同时考虑电池老化对内部状态预测效果的影响.BELM预测结果表明：所设计的预测模型具有较高的精度，能够实时准确地预测出电池的SOC和SOH 值.
展开▼
