摘要
为了快速、准确地对高压断路器发生的故障进行分析和诊断,确定故障的性质、类别和部位,提出了一种高压断路器故障诊断的新方法。首先对高压断路器分合闸线圈电流进行分析,提取电流和时间特征量形成特征向量,然后用遗传算法对最小二乘支持向量机(least square support vector machine,LS-SVM)参数进行优化,最后,将特征向量输入到优化后的最小二乘支持向量机中进行故障识别、分类。试验表明,该方法可以准确地识别断路器的多种故障类型,为断路器故障定位和状态检修提供了依据。与广泛使用的神经网络方法相比,该方法在样本较少时仍能获得较好的诊断效果,更适用于高压断路器等小样本设备的故障诊断。
A new fault diagnosis method is proposed to quickly and accurately diagnose faults in high voltage circuit breakers, and to determine fault＇s character, type and location. Firstly, the opening/closing coil current of a high voltage circuit breaker is analyzed, and the characteristic current and characteristic time of the coil current are extracted to form eigenvectors. Secondly, genetic algorithm is used to optimize the parameters of least square support vector machine（LS-SVM）. Finally, the eigenvectors are input into the optimized LS-SVM to identify and classify faults. Experimentation shows that this method can accurately identify a variety of fault types and provide a basis for fault location and condition-based maintenance of high voltage circuit breakers. Compared with the widely used neural network method, the proposed method performs better than the neural network method even with smaller sample set and is more applicable to fault recognition of high voltage circuit breakers.
出处
《高压电器》
CAS
CSCD
北大核心
2015年第12期79-83,共5页
High Voltage Apparatus
基金
国家电网公司大电网重大专项资助项目课题(SGCC-MPLG028-2012)~~
