摘要
针对传统的电动机保护装置无法实现早期故障诊断、不具备联网功能的问题,提出了一种基于物联网和支持向量机算法的分布式电动机故障诊断与保护系统的设计方案。该系统的下位机利用对称分量法将采集到的电动机定子电流进行分解,根据电流分量值判断故障类型来实现电动机的现场保护,并将定子电流数据通过ZigBee技术发送至嵌入式网关,通过GPRS网络实时上传给上位机;上位机通过小波包分解提取故障特征向量,采用支持向量机对电动机故障进行分类,实现故障早期诊断和预测。实际运行结果表明,该系统能准确诊断电动机故障并实施有效的综合保护。
In view of problems that traditional motor protection device cannot realize early fault diagnosis and does not have networking function, the paper proposed a design scheme of distributed fault diagnosis and protection system of motor based on Internet of Things and support vector machine algorithm. Lower computer of the system uses symmetrical component method to decompose stator current of motor, judges fault type according to current component value to realize site protection of motor, and sends current data to embedded gateway by ZigBee technology, uploads it to upper computer through GPRS network. The upper computer extracts fault feature vector by wavelet packet decomposition, uses support vector machine to classify motor fault, so as to realize early fault diagnosis and prediction. The actual running results show that the system can diagnose motor fault accurately and implement comprehensive protection effectively.
出处
《工矿自动化》
北大核心
2013年第6期17-21,共5页
Journal Of Mine Automation
基金
河南省教育厅科学技术研究重点项目(13B120889)
安阳师范学院博士启动基金项目(308765)
关键词
电动机
故障诊断
综合保护
物联网
支持向量机
motor
fault diagnosis
comprehensive protection
Internet of Things
support vectormachine