基于EEMD样本熵和支持向量机的高压断路器故障诊断

Fault Diagnosis of High Voltage Circuit Breaker Based on EEMD Sample Entropy and Support Vector Machine

首先提出了一种以集合经验模态分解(ensemble empirical mode decomposition,EEMD)样本熵和支持向量机(support vector machine,SVM)相结合的高压断路器振动信号的特征向量提取和故障诊断的分析方法。然后提出并分析了分闸弹簧力减小、缓冲弹簧无效两种新故障对高压断路器的性能影响,同时针对输电线路发生永久故障时,断路器按照跳闸、重合闸然后二次跳闸指令动作的实际情况,首次探索性选择振动能量相对较小的分闸振动信号作为诊断对象。最后实测了某35 k V高压SF6断路器正常运行、分闸弹簧力减小、缓冲弹簧无效、基座螺丝松动工况下断路器分闸振动数据进行识别,结果表明该方法检测高压断路器故障快速、准确,为断路器的故障诊断提供了新的思路。

This paper firstly proposed a new method for fault diagnosis of high voltage circuit breakers based on ensemble empirical mode decomposition(EEMD)sample entropy and support vector machine(SVM).Secondly,the paper analyzed the influence of two kinds of new faults on the high voltage circuit breaker,including opening spring force decrease and buffer spring invalid.When permanent fault occurred in transmission circuit and the circuit breaker was operated by the instruction of tripping,reclosing and re-tripping,the researchers tried to choose the opening vibration signal whose vibration energy was smaller than close vibration signal as the diagnostic object for the first time.Finally,through experiment on a 35 kV high voltage SF6 circuit breaker,the circuit breaker opening vibration data were measured under the conditions of normal operation,opening spring force decrease,buffer spring invalid and base screw loose.The results indicated that the method can diagnose breaker faults fastly and accurately thus offeringa new method for fault diagnosis of high voltage circuit breakers.