基于传感技术的电池全周期管控系统优化与检测

Optimization and detection of battery full cycle control system based on sensor technology

为提高锂钴充电蓄电池使用的安全性,提出基于深度学习的蓄电池检测系统,并建立蓄电池异常故障检测与分类模型。试验结果表明,窗口尺寸越大,输入数据包含的故障信息越多,蓄电池故障检测效率越高,且延迟故障检测效果最好,精度及准确率等指标均达到100%,其次为蓄电池重放故障。且基于深度学习模型可以检测到电池之间发生连接故障,并可预测电池的故障发生率会进一步增加。且随着检测时间,测量值和检测值之间的误差逐渐增大,进一步说明所建立的深度学习模型,可以检测到电池退化特征,从而为电池健康状况检测提供参考。

In order to improve the safety of lithium-cobalt rechargeable batteries,a battery detection system based on deep learning was proposed,and a battery abnormal fault detection and classification model was established.The experimental results showed that the larger the window size,the more fault information contained in the input data,the higher the efficiency of battery fault detection,and the best delay fault detection effect was achieved,with accuracy and other indicators reaching 100%,followed by battery replay fault.In addition,based on the deep learning model,the connection failure between the batteries could be detected,and the failure rate of the battery could be predicted to further increase.With gradually increases among the detection time,the error between the measured value and the detected value,which further indicates that the established deep learning model can detect the degradation characteristics of the battery,so as to provide a reference for the detection of battery health status.