Yogi-mLSTM Cable Partial Discharge Identification Method Based on Wavelet Threshold Denoising

High voltage cables are important power equipment in the power system, and their operating status affects the safety and reliability of power supply in the power grid. At present, there is no comprehensive technical means for identifying cable partial discharge defects, and there are also significant difficulties in collecting on-site data. Moreover, the existing cable partial discharge recognition methods have low accuracy, large workload, and incomplete systems. This paper proposes a Yogi-mLSTM cable partial discharge recognition method based on wavelet threshold denoising, which includes: collecting and obtaining cable partial discharge defect signals; Using wavelet threshold method to denoise the partial discharge signal and obtain the raw data of cable partial discharge; Normalize the raw data of cable partial discharge, then extract the main feature values as sample data, and split them into training and testing samples; Build a Yogi-mLSTM network model; Transfer the sample data to the Yogi-mLSTM network model for training, while determining the termination condition (number of iterations); Determine whether the model training meets the termination condition. If it does, input the data to be recognized into the trained diagnostic model, and then output the recognition results. If it does not, continue training the model. This method provides a feasible and reliable method for cable fault diagnosis, improves the intelligence level of the power station, and effectively ensures the safety and stability of cable operation.