面向全生命周期的锂电池健康状态估计

为研究锂电池在动态工况下以及全生命周期内健康状态的准确估计问题,提出一种基于固定充电电压片段的方法.选取充电过程中某固定电压片段内所充电量作为电池容量估算的等效健康因子,利用遗传算法选择最优的充电电压片段,在两类锂电池老化实验数据的基础上,设计放电电流不同、健康状态区间不同的8个验证算例.实验结果表明:8个验证算例中,训练集电池和测试集电池健康状态估计的平均绝对误差与均方根误差均低于1.55%;所提出的基于等效健康因子的方法,在100%～60%的全寿命健康状态区间,对于不同的放电电流、不同材料的电池,均能进行健康状态的准确在线估计,具有较强的适用性.

A state of health estimation method for full lifetime of lithium-ion batteries

To solve the problem that the state of health(SOH) of lithium-ion batteries is difficult to be estimated accurately under dynamic working conditions and full life cycle, a method based on fixed charging voltage segment was proposed. Firstly, the charging capacity in a fixed voltage segment during the charging process was treated as the equivalent health factor of battery capacity estimation. Secondly, the optimal charging voltage segment was selected by using genetic algorithm. Finally, eight verification numerical examples based on the aging experiment data of two types of lithium battery were designed, which were different in discharging current and SOH interval. Experimental results show that the value of MAE and RMSE that comes from the estimated SOH of training set batteries and testing set batteries in eight numerical examples were less than 1.55%. The proposed method can accurately estimate the SOH of lithium batteries under full lifetime(SOH between 100% and 60%) for different discharging rates and materials, which means this method is good applicable in practice.