摘要
随着电动车充电站的迅速普及,使用中的充电站出现的问题种类和数量持续上升,相应产生的数据量也在不断增长。传统的手动数据管理方法已显得效率不高,不利于充电桩系统的正常运行。为此,文中提出了一种反向学习状态空间模型演化算法,应用于充电桩的故障诊断与预测,可进行反向学习自适应诊断,预测结果准确,处理速度快,有效提高了充电桩大数据的处理效率;同时将训练结果与实际故障数进行对比,验证了算法的准确性和有效性。研究成果为今后充电桩的安全运行提供了有效的技术支撑,具有重要意义。
With the rapid proliferation of electric vehicle charging stations,the types and quantities of problems encountered in operational charging stations continue to rise,accompanied by a corresponding increase in the volume of generated data.Traditional manual data management methods have proven inefficient and are detrimental to the operation of charging station systems.Therefore,a reverse learning state space model evolutionary algorithm is proposed for fault diagnosis and prediction of charging piles,enabling adaptive diagnosis through reverse learning,accurate prediction results,and fast processing speeds,thereby effectively enhancing the efficiency of big data processing for charging piles.Furthermore,the comparison between the training results and actual fault numbers validates the accuracy and effectiveness of the algorithm.The research findings provide effective technical support for the safe operation of charging piles in the future,holding significant importance.
作者
李翟严
胡耀杰
徐礼富
戴海兵
张流涛
LI Zhaiyan;HU Yaojie;XU Lifu;DAI Haibing;ZHANG Liutao(Huzhou Xinlun Comprehensive Energy Services Co.,Ltd.,Huzhou 313000;Key Laboratory of Carbon Electricity Digit of Huzhou,Huzhou 313000)
出处
《机械设计》
CSCD
北大核心
2024年第S01期192-195,共4页
Journal of Machine Design
关键词
充电桩
故障诊断与预测
反向学习自适应
状态空间模型
charging pile
fault diagnosis and prediction
reverse learning adaptive
state space model
