摘要
为解决智能电表累积采集信息量大、故障信息种类繁多和突发性强的问题,构建一种基于随机森林的电表故障及寿命预估模型。依靠大数据分析理论,通过对海量电表的累积数据进行挖掘分析,建立智能电表的故障预测及寿命预测模型来对故障和寿命进行预测,并同其他模型进行比较。实验结果表明:该预测模型是有效的和准确的,可为数据挖掘在智能电表管控研究提供参考。
In order to deal with the smart electric meters’ mass information,varied fault types and sudden faults,the fault and life prediction model for smart electric meter based on random forest is established.Based on the theory of large data analysis,this paper excavates the accumulative data of large numbers of smart electric meters and forecast its fault and life,compare with other models.The experimental results demonstrate the effectiveness and correctness of the proposed model.It can provide reference for data mining in the research of smart electric meter.
作者
黄吉涛
樊博
周媛奉
胡婷婷
梁飞
曾晓东
Huang Jitao;Fan Bo;Zhou Yuanfeng;Hu Tingting;Liang Fei;Zeng Xiaodong(Electric Power Research Institute of State Grid Ningxia Power Co.,Ltd.,Yinchuan 750011,China;College of Electrical Engineering&Information Technology,Sichuan University,Chengdu 610065,China)
出处
《兵工自动化》
2019年第10期57-60,共4页
Ordnance Industry Automation
基金
国家电网公司科技项目(宁电发展[2018]54号)
关键词
智能电表
数据挖掘
随机森林
故障预估
smart electric meter
data mining
random forest
fault prediction