Research on GIS partial discharge pattern recognition based on EFPI sensor
In order to accurately capture weak signals generated by partial discharge in complex electromagnetic environment,gas-insulated switchgear(GIS)partial discharge pattern recognition method based on EFPI sensor is designed. The cavity length of EFPI sensor is coarse demodulated by differential dual-wavelength light intensity ratio demodulation algorithm. Using complex domain correlation algorithm,combined with coarse demodulation results and precise demodulation cavity length,the final EFPI sensor cavity length is determined,which is used to detect ultrasonic signals of GIS equipment. Hilbert marginal spectrum is obtained by transforming ultrasonic signal of GIS equipment by Hilbert. The features of Hilbert marginal spectrum are extracted by using 2D-1D depth residual network,and the results of partial discharge pattern recognition in GIS are output. Experimental analysis result shows that:this method can effectively determine the cavity length of EFPI sensor and collect ultrasonic wave signal of GIS equipment;obtain the Hilbert marginal spectrum of ultrasonic signals of GIS equipment. Under complex noise environment,this method can accurately identify GIS partial discharge mode.
万方数据
维普
