摘要
为提高开关柜故障诊断的准确性,提出一种基于RFID传感器和深度学习的开关柜故障诊断算法。首先,设计用于采集开关柜电流信号和温度射频识别(radio frequency identification,RFID)的传感标签;其次,采集的信号通过深度信念网络(deep belief networks,DBN)进行深层次特征提取,并将稀疏编码(sparse code,SC)融合到DBN网络中,提高其检测精度;最后,为提高检测速度,采用极限学习机(extreme learning machine,ELM)对特征提取的信号进行分类识别。研究结果表明,相比于其他算法,本文提出的SDBN-ELM故障诊断模型检测精度更高,识别速度更快,其准确率可达99.63%。
In order to improve the accuracy of switchgear fault detection,this paper proposes a fault detection algorithm for switchgear based on RFID sensors and deep learning.Firstly,RFID sensing tags are designed to collect the current signals and temperature of the switchgear.Secondly,the collected signals are subjected to deep-level feature extraction through a deep belief network(DBN),and sparse coding(SC)is integrated into the DBN to improve its detection accuracy.Finally,in order to improve the detection speed,an extreme learning machine(ELM)is used to classify and recognize the signals extracted from the features.The experimental results show that compared to other algorithms,the sparse DBN-ELM(SDBN–ELM)fault detection model proposed in the paper offers higher detection accuracy,faster recognition speed,and an accuracy rate of 99.63%.
作者
王真
刘子全
路永玲
李玉杰
WANG Zhen;LIU Ziquan;LU Yonging;LI Yujie(Electric Power Research Institute,State Grid Jiangsu Electric Power Co.,Ltd.,Nanjing 211103,China)
出处
《电力科学与技术学报》
北大核心
2025年第2期179-185,共7页
Journal of Electric Power Science And Technology
基金
国家电网有限公司科技项目(J2023091)。
