In order to improve the recognition ability of transformer grounding fault,a fast recognition method of inrush current is designed.In this method,the least square matrix algorithm is used to extract the current components accurately,and the current energy infor-mation entropy model is constructed to achieve the accurate classification effect of inrush current and short-circuit current.The re-sults show that when the inrush current is formed in the transformer,the H value obtained by the differential flow algorithm is larger,and the fundamental wave and aperiodic components are weakened.The short circuit causes the aperiodic component to in-crease significantly,which can be reduced to 0 at least.The waveform curve composed of obvious fault signals is obtained,which can realize the efficient identification of short-circuit faults.This scheme can obtain the fault information and disturbance varia-tion parameters under various working conditions through in-depth discussion of waveform frequency domain data.Compared with the second harmonic braking mode,it has greater advantages and can fully optimize the control performance of differential protection action.
万方数据
