基于PID⁃BPNN的矿用铅酸蓄电池SOC在线估计

摘要

针对矿山避难硐室安全供电系统中铅酸蓄电池内化成过程中检测是否已经达到满电荷量,且在组装铅酸蓄电池时需要考虑电池均衡问题都需要进行准确估算SOC的问题,提出基于BP神经网络的PID控制通过修正反馈误差来实现铅酸蓄电池SOC在线估计.采用实验的方法获取数据,选取与电池SOC相关的因子作为BP神经网络的输入参数,最终准确在线预测蓄电池SOC值.仿真结果表明,基于BP神经网络的PID控制的铅酸蓄电池SOC估计的精度大大提高,同时为电池管理系统提供一个新的估计方法.%Since accurate SOC estimation is needed when detecting whether the lead-acid battery has reached the full charge during its internal formation and when considering the battery balance problem during the assembling of the lead-acid bat-tery in the safe power supply system of the mine refuge chamber,SOC online estimation of lead-acid battery is achieved based on PID control of BPNN by means of feedback error modification. The experimental method is adopted to obtain data,and the factors related to battery SOC are selected as the input parameters of BP neural network to perform accurate online prediction of the battery′s SOC values. The simulation results show that the lead-acid battery SOC estimation based on PID control of BPNN has improved a lot in its precision,and meanwhile provides a new estimation method for the battery management system.