Font Size: a A A

Research On Detection And Pattern Recognition Of Partial Discharge In XLPE Cable Joint

Posted on:2013-10-19Degree:MasterType:Thesis
Country:ChinaCandidate:X Y JiFull Text:PDF
GTID:2232330395476131Subject:Power system and its automation
Abstract/Summary:PDF Full Text Request
XLPE cable are widely used in grid of various voltage levels.Partial discharge are the main threats to the level of XLPE cable insulation, and cable joint is the weak link part of the cable insulation, so its on-line monitoring become the key content to the protect of XLPE cable.According to the needs of Xuejiawan Power Supply Bureau, we independently design and development the wideband current sensor.Through the experimental comparison,we determine the optimal experiment materials and parameters, and test the performance of the sensor by simulation experiments in the laboratory.We further build a partial discharge detection system, composed by the sensor,the signal conditioning circuit, data acquisition card and PC, to achieve on-site detection of partial discharge signals.We focuse on four typical discharge models in laboratory simulation experiments, the PD signals of four models are denoised by the application of the multi-wavelet theory, in order to avoid the interference of the noise. The PD signals After processing can be used as the original datas to pattern recognition. Through the build of φ-q-n three-dimensional map,we determine the six kinds of characteristic quantities.Finally, we apply the support vector machine theory to complete pattern recognition of partial discharge signals, and the results of recognition achieve the desired requirements.What we have done will provides an important basis for the further analysis.
Keywords/Search Tags:XLPE cable, partial discharge, multiwavelet, pattern recognition, Φ-q-nmap, support vector machine
PDF Full Text Request
Related items