| Nowadays, the atmospheric pollution is becoming deteriorative year by year with the development of the economy. Especially in our country, the contamination flashover of high voltage insulators has been a serious issue for the safe and reliable operation of power transmission lines. Perhaps no single topic of insulator behavior has engaged researchers more so than predicting flashovers under contaminated conditions. Therefore, how to prevent the contamination flashover in advance has become the focus for the safe operation of transmission and distribution systems. There is a large body of literature published to explore the features of the contamination flashover. Yet there are still some critical gaps in the solution of this complex problem. Contamination flashover is the ultimate result of the creeping discharge of polluted insulators. Since there are close ties between the contamination flashover and the leakage currents, it is critical to research the leakage current characteristics during the entire contamination discharge process.As the leakage currents of the contamination insulators comprised a large number of noises before the occurance of partial discharge, it was very difficult to obtain the valid characteristics of signals. Here four wavelet threshold de-noised methods were used respectively to filter the noises from the leakage currents with different signal-to-noise ratios (SNR) measured in our laboratory.Three characteristics, namely the waveforms, the RMS value and the amplitude ratio of third harmonic to first harmonic of leakage currents before and after separating noises were extracted with good effects. The optimal wavelet threshold de-noised method for the leakage currents is selected out. Finally, it is showed that when SNR is greater than 1.0, the best de-noised method is self-adaptive threshold method for the two characteristics at the security stage, namely waveforms and RMS value. However, this kind of approach is almost equally between real value and de-noised value for another characteristic, i.e. the amplitude ratio of third harmonic to first harmonic of leakage currents. Generally, it is well known that the self-adaptive threshold method is optimal to extract the characteristics of leakage currents.The leakage currents of porcelain and glass insulators are monitored and analyzed through a number of laboratory tests in various polluted cases. The goal is to find the characteristics that are useful for the pre-warning of the contamination flashover. The emphasis of this research is on the leakage current RMS (root-mean-square) values and waveforms, and power spectrum estimation of the leakage currents. Results of the experiments show that the progress of the contamination discharge process can be classified into three stages, i.e., security stage, forecast stage and danger stage. In addition, the boundaries of the three stages in both the time domain and the power spectrum domain are given based on the analysis of the test results of various insulators. That is very helpful for the stage pre-warning of the contamination flashover. Also, a new characteristic index is introduced. It is the rate of rise of the maximum peak value of the power spectrum. It can provide a more comprehensive theoretical guidance for the contamination flashover stage pre-warning together with the leakage current and waveforms. The three-stage classification for the entire flashover process is also meaningful in order to perfect the prediction of the contamination flashover. Finally, other characteristics of the power spectrum estimation are also discussed based on the three-stage classification.The leakage currents are the most important information for the monitoring of contamination insulator, especially the forecast of contamination flashover. But the contamination severity and the ambient humidity can both result in the increase of the leakage currents, so it is crucial to distinguish the exact reason for the changes of leakage currents. Lots of repeated experiments have shown that the normal range of the leakage currents is less than 50 mA under the operating voltage, which is called as the security stage for far away from the flashover. The characteristics of the leakage currents at the security stage are researched when the ambient humidity or the contamination severity changes respectively in the artificial fog chamber. A joint approach of FFT (Fast Fourier Transform) and power spectrum estimation is used to analyze the leakage currents in combination with the leakage current waveform. The main purpose is to improve the pre-warning of contamination flashover at the security stage and to win the more time for the cleaning or replacement of the heavy polluted insulators. Test results show the two spectrum frequency characteristics, namely, the amplitude ratio of third harmonic to first harmonic and the ratio of high frequency energy to total energy, can play an important role in the state evaluation of the insulator surface aimed at the relative humidity and the contamination severity. Furthermore, the pre-warning thresholds of the two characteristics for the heavy contamination severity are existent under the condition of the high humidity, which is very helpful for the pre-warning before the flashover. In order to assess how severe the contamination level of the surface of power line insulators and to prevent unpredictable contamination flashovers, it is important to seek optimal prediction characteristics. That leads to the increase of the warning time and to the improvement of the reliability of the pre-warning system. Nearly 30 insulator strings at five pollution levels were tested in an artificial fog chamber, where their leakage currents were continuously recorded at the same operation conditions. The three characteristics of the leakage current, namely the mean value, maximum value, and the standard deviation of the RMS value of the leakage current, have been extracted. They describe jointly the current contamination levels of an insulator surface. In addition, regression equations between the three characteristics and various contamination levels have been established. The same three characteristics have been selected and used as the inputs of a neural network model together with two more parameters, the relative humidity and operating voltage. Also, the influence of each characteristic on the contamination prediction results has been investigated. The comparison of the simulated and actual (measured) results demonstrates that the ESDD prediction model has a very low relative error, less than 3.6%, if the training data and the testing data both come from the security stage. The application of this research results in (1) optimal ESDD prediction inputs and (2) sufficient pre-warning time before the ultimate contamination flashover.The findings of this paper have enriched the current understanding of leakage current characteristics of polluted insulators, improved the description of flashover mechanism according to leakage current as a basis, and established the relationship between leakage current and the stage characteristics of contamination discharge on the surface of polluted insulators. These provide theoretical basis not only to optimize the input characteristics of contamination flashover pre-warning system, but also to guide the choice of heavy polluted transmission lines.
|
PDF Full Text Request