非接触式继电器故障检测系统设计  

Design of non⁃contact fault detection system for relay

作　　者：陈功 邵雷[1] 赵鹤鸣[1] CHEN Gong;SHAO Lei;ZHAO Heming(School of Electronic Information,Soochow University,Suzhou 215000,China)

机构地区：[1]苏州大学电子信息学院,江苏苏州215000

出　　处：《现代电子技术》2021年第18期48-52,共5页Modern Electronics Technique

基　　金：国家自然科学基金(61803278)。

摘　　要：工业生产过程中,继电器的故障检测需求越来越大。然而,目前对继电器的故障检测需将其从电路中拆除,采用直接接触的方式测量特征参数,通过对参数的进一步分析来判断继电器的损坏程度。这种方式经济成本较高,且无法掌握正在使用中的继电器状态。为解决这一弊端,文中研制一款专用天线来采集继电器电磁信号,基于FPGA通过以太网将数据传输到上位机。创建继电器故障信号数据集,构建神经网络模型,设计一个非接触式继电器故障检测系统,图形用户界面能直观地给出信号图像和检测结果。最后搭建测试环境,对整个系统识别率进行验证,综合准确率达到了91.2%。结果表明,该系统能以非接触的方式分析继电器电磁信号,实时可靠地检测继电器的状态,降低破坏电路的损失。At present,the fault detection of relay needs to remove the relay from the circuit to detect its characteristic parameters in direct contact way,and the damage degree of the relay is judged by further analysis of the parameters.This method has a high economic cost and cannot find the status of the relay in use.In order to solve this shortcoming,a special antenna was developed to collect the electromagnetic signals of the relay.The data is transmitted to the host computer via Ethernet on the basis of FPGA.A data set of relay fault signal is created,a neural network model is built,and then a non⁃contact detection system for relay fault is designed,where signal images and detected results are intuitively given in graphical user interface.A testing environment was constructed to verify the recognition rate of the whole system.The testing results show that the comprehensive accuracy of the system can reach 91.2%,the system can analyze the electromagnetic signal of the relay in a non⁃contact manner,and reliably detect the status of the relay in real time,which can avoid the circuit damage.

关 键 词：继电器 故障检测 非接触式检测 电磁信号采集 数据传输 神经网络模型 软件设计 

分 类 号：TN916.424-34[电子电信—通信与信息系统] TM58[电子电信—信息与通信工程]