Automatic Detection Method for Substation Line Fault Based on Wavelet Neural Network
The existing methods are difficult to capture subtle changes in substation line fault signals,resulting in significant deviations from actual results.Therefore,a method for automatic detection of substation line faults based on wavelet neural network is proposed.Firstly,based on wavelet packet decomposition,information such as energy distribution and mutation characteristics of substation lines in different frequency bands are obtained.Secondly,a wavelet neural network fault detection model is constructed,and the feature information obtained from wavelet packet decomposition is used as the input of the model to establish the mapping relationship between fault signals and fault types or degrees.Finally,repeatedly train the model to detect fault signals in substation lines.The results indicate that the detection results of this design method are in line with actual results,improving the accuracy of fault detection.
万方数据
维普
