Simulation of Partial Discharge Monitoring for High Voltage Switchgear Based on Wavelet Analysis
The traditional partial discharge signal detection algorithm for switchgear is susceptible to external elec-tromagnetic interference and has certain limitations.In order to solve the problem of low accuracy and poor timeliness of partial discharge signal detection caused by noise interference,a discharge signal optimization algorithm based on wavelet analysis and feature analysis is proposed.Firstly,the wavelet transform algorithm was used to decompose and reconstruct the partial discharge signal to solve the problem of low frequency resolution in STFT transform.Then,the soft threshold processing algorithm was used to reduce the noise of PD model,improve the smoothness of waveform and remove the external electromagnetic interference.Then,the Pearson correlation analysis method was used to opti-mize the high-latitude characteristic data to improve the efficiency of model construction;The system parameters of SVM were optimized by the best ten fold cross validation method.The classification model of SVM-PD discharge sig-nal was constructed.The results of ablation experiments show that the introduction of three groups of optimization modules can significantly improve the comprehensive performance of the classification model and reduce the construc-tion time of the model;The comparative experimental results show that compared with the baseline algorithm,the R of the proposed model increases by 2.41%,the F1 increases by 1.32%on average,and the migration increases by 1.43%.Compared with literature[9],the accuracy of the SVM-PD model increases by2.36%.Therefore,the SVM-PD discharge signal classification model based on wavelet analysis method in this paper not only solves the external electromagnetic interference,but also improves the accuracy and timeliness of signal recognition.
NETL
NSTL
万方数据
