三相串联故障电弧中故障相数的识别

Identification of number of fault phases in three-phase series fault arc

三相异步电动机在现代工业和煤炭企业中地位日益突出。随着工艺的发展,变频器与三相异步电动机配合使用成为了主流。在一些情况下,三相异步电机配合变频器发生串联故障电弧,造成了重大的电气安全隐患。为了减少发生串联故障电弧的危害,对故障电弧进行了分析与研究。在实验室的条件下,对于负载为带有变频器的三相异步电动机进行了实验测试,首次提出采用相空间重构理论提取某一相负载电流的轨迹,可以明显的识别出三相串联故障电弧中故障电弧发生相的相数,然后使用带有核函数的MeanShift聚类算法对提取的轨迹进行聚类,使得识别效果更加突出。阐明了该算法对于三相串联故障电弧中的故障电弧相数识别的有效性。

Three phase asynchronous motor plays an increasingly prominent role in modern industry and coal enterprises. With the development of technology,the use of three-phase asynchronous motor equipped with frequency converter has become the mainstream. In some cases,three-phase asynchronous motor equipped with frequency converter produces series fault arc,which causes serious potential electrical safety hazards. In order to reduce the harm of series arc fault,fault arcs are analyzed and studied. Under the condition of laboratory,the three-phase asynchronous motor equipped with frequency converter is tested experimentally. For the first time,the phase space reconstruction theory is used to extract the load current trajectory of a phase. The phase number of the fault arc in the three-phase series fault arc can be clearly identified. Then Meanshift clustering algorithm with kernel function is used to cluster the extraction of trajectories,making the recognition effect more prominent. The effectiveness of the algorithm in identifying the number of arc phases in three-phase series fault arcs is illustrated.