Research On Cable Fault Diagnosis Based On Wavelet Transform And Weibull Function

XLPE cable with its high reliability, the advantages is not affected by the effect of building wiring, and in the city power supply system has been widely used. With the increased use of power cables, such as ground fault frequency is gradually increasing, and these faults, mostly by cable insulation in partial discharge(PD) developed. The inevitable process of power cable in the manufacturing and processing of the introduction of impurities and voids in the insulation layer, the weak link of these voids and impurities are located is in the process of cable operation. Partial discharge generated layer voids and cavities in the insulation condition of down insulated cables, and accelerate the aging of the insulation materials, resulting in scrap and cable insulation breakdown. Therefore, the detection and identification of PD and PD positioning has become the key point of insulation condition monitoring system.In order to solve the problem that traditional method is poor ability in partial discharge signal denoising, low accuracy of location the partial discharge point positioning and the cable fault diagnosis is not accurate. This paper research and analysis the partial discharge signal denoising based on the wavelet method, partial discharge point location, fault type identification and diagnosis technology, and involved in the development of the partial discharge measurement and diagnostic equipment and corresponding software. The main work and innovation of this paper are listed as below:First of all, for The problem that traditional wavelet analysis method is poor in signal denoising, improve the second generation wavelet transform(SGWT) algorithm, use compromise that between soft and hard threshold function to replace the traditional hard threshold function to deal with the coefficients after wavelet decomposition, to overcome the disadvantage of hard threshold algorithm, enhanced the effect of second generation wavelet denoising.In the second, On the basis of the existing correlation analysis algorithm, introduce the algorithm of time delay estimation, and combining the two algorithms to obtain the partial discharge location algorithm based on correlation time delay estimation. Application of correlation analysis to achieve the two times reduction of the partial discharge signal based on second generation wavelet transform. By comparing with the results of direct correlation analysis, the advantages of this method are verified in the two times denoising and improving the rapidity and accuracy of the partial discharge location algorithm.And then, introduce the weibull function that used in life assessment of physical material to fault diagnosis and predict the life of power cable, realized The diagnosis of partial discharge type and discharge position of power cable.In addition, use MATLAB and Simulink simulation tools to simulate and studied the theory and improved method that proposed in this dissertation. Through the analysis and comparison of simulation results, validate the proposed theoretical feasibility and the improved method is superior to the traditional one.At last, some of the conclusions from the study of popularization and application, cooperation with the relevant power companies and developed a set of power cable partial discharge detection equipment and the corresponding software, now it has been put into practical application and achieved good results.