Research On Partial Discharge Location Method Of Transformer Based On Time Reversal

The deterioration degree of insulation performance of transformer can be shown by partial discharge.At the same time,partial discharge will aggravate the insulation deterioration and reduce the service life of transformer.Therefore,in order to improve the reliability of transformer operation,it is necessary to detect the partial discharge of transformer.It is of great significance to detect partial discharge in time and locate the discharge source accurately.Ultra-high frequency detection is an effective method to detect partial discharge.At present,the most commonly used ultra-high frequency positioning method is time difference positioning method.In this thesis,two key technologies of time difference location method,time delay estimation and location algorithm,are studied:(1)In the process of time delay extraction,the background noise affects the time delay estimation,which makes the time delay estimation error larger.In order to obtain a clean partial discharge waveform,empirical wavelet transform is used to suppress the common white noise interference and periodic narrow-band interference in substation.It lay a solid foundation for improving the accuracy of time delay.(2)Aiming at the problem that the feature points are easy to be annihilated in the background noise,which leads to the increase of delay error.The signal waveform characteristic point method was improved.An improved time difference extraction method based on slope correction was proposed.The time delay was extracted by the rising edge which is not easily affected by the interference instead of the feature points,which improved the accuracy of time delay extraction and provided more accurate delay for the final positioning.(3)Aiming at the inevitable problem of time delay difference in time difference algorithm.The time reversal technique was used to localize the ultra-high frequency partial discharge.A method of partial discharge ultra-high frequency location based on time reversal focusing imaging was proposed.This method can ignore the influence of waveform mode,without direct time delay estimation.It provided a good foundation for the combination of time reversal method and location optimization algorithm in Chapter 5.(4)Due to the large delay error in the process of location algorithm solving,the optimization algorithm falls into local optimum or even has no solution.Combined with time reversal positioning results,the fitness function,initial population and weight strategy were improved.An improved whale optimization algorithm suitable for partial discharge localization was proposed.Simulation results and laboratory results verified the effectiveness of the proposed algorithm.To sum up,points 1 and 2 aim to improve the accuracy of time delay extraction.Points 3 and 4 aim to improve the accuracy of the positioning algorithm.These four points work together to improve the positioning accuracy of partial discharge.