The Research And Development Of Partial Discharge On-line Monitoring System Based On The UHF

With the development of the national grid, the Gas Isolated Switcher(GIS) is widely used in high power transmission and distribution system, the advantages of GIS is less influenced by outside factors, stable operation, long maintenance cycle, long-term survey of Operation fault on GIS shows that insulation damage is a major cause of GIS fault. Partial discharge is a sign of insulation deterioration, and also a major cause of accelerating insulation degradation. The insulation state of GIS can be effectively known through analyzing partial discharge. The on-line detection of partial discharge was recommended as the only effective insulation tests, so there is great significance to carry out the on-line monitoring of partial discharge in GIS.Compared with the common methods partial discharge detection, the UHF detection method is used in this paper. There is an important position for UHF detection among the partial discharge monitoring of GIS. The reasons of partial discharge was analysed from the nature and the main external factors of causing that phenomenon were listed. The electromagnetic wave propagation path in the GIS was researched, the location principle based on the minimum path difference was presented.the characteristic parameter of partial discharge was introduced.Based on the disadvantages of centralized monitoring system, committed to the development of a partial discharge detection system with distributed structure. Data were collected and processed in situ processing unit, The judgment of fault and discharge location were realized. Each local processing unit has independent function, damage and exit of the individual unit and will not affect the normal work PC, the stability and flexibility of the detection system were improved. Due to the envelope curve caused by different fault discharge is different, UHF signal frequency was reduced by using the envelope detection method in situ data processing unit to solve the contradiction between UHF signals and the acquisition system. High-speed DSP as the core control circuit to complete the signal amplification, filtering, collecting, the data processed with corresponding algorithm was uploaded to monitoring and diagnosis center.4kinds of GIS common fault model were designed to test to collecte envelope waveform of discharge signal. Based on characteristics of envelope signal, pattern recognition and fault diagnosis were developed.The characteristic information of envelope curve was extracted with wavelet packet decomposition, the kernel principal component analysis method to verify the effectiveness of the proposed method. System software development is divided into upper and lower two module, Signal processing algorithms and control program of DSP system was developed with C language in the CCS. the diagnostic analysis software was developed by object oriented programming tool VC++.