基于数据挖掘与改进安时积分的SOC估算

提出改进安时积分法对电池荷电状态(SOC)进行在线估算。采用数据挖掘的方法对电池参数进行数据处理,以保证电池采样参数的精确性与标准化。鉴于电池电流倍率、工作温度及老化因子对电池容量的影响,对传统安时积分法进行修正,保证算法的精确性与可行性。基于数据挖掘丝技术的SOC估算算法提高了估算的精度,不同温度下基于联邦城市运行工况(FUDS)循环的SOC估算值与实测值的最大误差仅1.6%。

SOC estimation on basis of data mining and improved ampere-hour counting method

The improved ampere-hour (Ah) counting method was introduced for on-line state of charge (SOC) estimation. Data mining for battery parameters process was proposed to ensure the precision and nonnalization of the sampled data. Based on charge/ discharge current, battery temperature and state of health (SOH) of battery, the traditional Ah integral method was modified to ensure the accuracy and feasibility of the algorithm. The accuracy of estimation was improved based on data mining of SOC estimation algorithm. The maximal error of estimated SOC and measured SOC on different temperatures and federal urban driving schedule (FUDS) cycles was just 1.6%.