A Current Signal Compression Acquisition and Reconstruction Method Considering Fault Arc Identification

Aiming at the challenge of high frequency arc fault signal acquisition, a current signal compression acquisition and reconstruction approach taking into account arc-fault identification is proposed for the low-body-volume acquisition of arc-fault signals, which effectively preserves the high-frequency arc-fault characteristics in the current data. Firstly, the current signals characterizing the fault arcs are collected using a flexible Blackman window for sliding acquisition. Then, it employs an improved partially orthogonal Gaussian measurement matrix to achieve compressed acquisition of the current signals through Compressed Sensing (CS) technology. Subsequently, the collected current signals undergo sparse transformation using Discrete Wavelet Transform (DWT). Finally, the method utilizes the Compressive Sampling Matching Pursuit based on Optimal Sparsity Estimation (CSMP-OSE) to achieve accurate iterative reconstruction of the compressed high-frequency arc low-dimensional data. Experimental validation based on the substation arc fault dataset demonstrates that this method offers robust arc fault characterization capability, high reconstruction accuracy, and efficiency.