摘要
越来越多的动力机械采用锂离子电池作为其动力源。准确估算锂离子电池的SOH能够给动力机械安全可靠地运行提供保障。在不同的实验条件下对18650型锂离子电池做充放电循环实验,由实验结果得到锂离子电池的充放电循环特性,根据其循环特性采用人工神经网络寻找电池端电压与SOH之间的非线性关系,进而估算锂离子电池的SOH。估算结果表明,采用BP神经网络能够准确地估算锂离子电池的SOH,估算误差基本控制在3%以内。
More and more power machines choose lithium-ion batteries as their power source. The accurate estimation of SOH for the lithium-ion batteries can ensure the power machine safely and reliably operate. Under different experimental conditions, the charge-discharge cycle test was conducted for 18650-type lithium-ion batteries, and the batteries＇ charge-discharge cycle characteristics were obtained by the experimental results. According to the cycle characteristics, the artificial neural network was chosen to find the nonlinear relationship between the battery terminal voltage and SOH, and then the SOH of the lithium-ion batteries was estimated. The estimation results show that the BP neural network can accurately estimate the SOH of lithium-ion batteries, and the estimation error is basically less than 3%,
出处
《电源技术》
CAS
CSCD
北大核心
2017年第5期708-710,共3页
Chinese Journal of Power Sources
关键词
锂离子电池
SOH
动力电池
神经网络
lithium-ion batteries
SOH
power batteries
neural network
