| Partial discharge (PD) damages the insulation of equipment and indicates theattenuation of complete insulation. Since partial discharge is a paroxysmal phenomenon,it can hardly be found in time based maintenance. So, on-line monitoring, which canreport delitescent insulation fault timely, is a more effective method to avoid insulationfault. High-voltage switchgears, widely used in the power grid, affect the power supplydirectly. It will cause a great loss to our economy and society if power outage occurs byswitchgears fault. In a word, the on-line monitoring of operating high-voltage switchgearscan provide important information for theirs own condition based maintenance and thenavoid accident effectively. So, this research is valuable for both science and society.Ultra-high-frequency (UHF) method, as one of the most important detectingmethods for PD, is widely used because of its high sensibility and good interferencerejection. In this paper, the application of UHF method in PD on-line monitoring ofswitchgear is analyzed. According to the structural features of the switchgear and therequirements of the PD detecting, a UHF fractal antenna based on Koch curve is designdwith a size of200mm×200mm×25mm. By actual measurement, it had a working bandof565MHz-850MHz, an average gain of1.2dB and good directivity. Meanwhile, aftercomparing the amplitudes of PD signals of this antenna with that of othe sensors, it can befound that this UHF sensor performs better in actual detecting——It can detect differenttypes of PDs effectively. In this research, four typical defect models are developed togenerate different kinds of PDs in testing switchgear. On this basis, the regularity ofvarious PDs is analyzed, and the three-dimensional spectrums and scatter plots areconstrusted by a large numbers of PD patterns. The foundation of PD pattern recognitionhas been laid by the above research. In order to find the optimal location of the UHFsensor in the switchgear, the propagation characteristics of UHF signals in switchgears islucubrated by FDTD method in this thesis. For this purpose, the influence of PDwaveform parameters, detecting distance and each component of switchgear on thepropagation characteristics of UHF signals has been definituded by simulation. Byconsidering both the overall detection effect and volatility of each detection point, theoptimal position of the inner UHF sensor was determined the right side of bottom plate inthe front cabinet. In addition, the simulation result was verified by experiment.On the basis of the above research, based on the hardware equipments, which consist of UHF fractal antenna, filter amplifier, high-speed data acquisition card,power-frequency voltage acquisition module, industrial personal computer (IPC) and soon, a monitoring system for PD on-line monitoring of switchgears is developed.Furthermore, a corresponding system software platform is also developed. Themonitoring system has been installed in an110kV substation to monitor operatinghigh-voltage switchgears for a period of time. In practical use, this system has beenshown to detect paroxysmal discharge and works well also. |