Study On Prediction Of Insulator Pollution Flashover Based On Leakage Current Characteristics

With rapid industrial and agricultural development, atmospheric pollution has become more and more severe. With voltage level of electric transmission lines has constantly increasing, the distribution range of power network is becoming wider day by day, and as a result, the possibility of pollution flashover in the power network is becoming higher and higher. According to statistics, the trip-out times caused by pollution flashover is only less than by lightning strike, namely has the second highest trip-out times, but economical loss caused by the former is 10 times by the latter. So, how to monitor the contamination level of insulator surface in real time, and then to scientificlly predict and evaluate the potential threat of pollution flashover has been one of the tough technical problems under exploration and study in power network. In this paper, characteristics of leakage current that flows over the surface of the contaminated insulators are studied through trial, with relationship between the said leakage current and its impact factors analyzed, and the evaluation method on prediction of pollution flashover of insulators is established based on leakage current characteristics values, which has certain reference significance in anti-contamination of transmission lines.In this paper, according to the trial results of porcelain insulator XP-70, the effects of salt deposit density(SDD) and the relative humidity(RH) on the amplitude of leakage current(Ih) are analyzed, as well as the variation feature of phase difference(θ) between the leakage current and operation voltage on the polluted insulator. The experiments result show thatθdecreases with increasing of RH at a certain SDD, while at a certain RH, it decreases with increasing of SDD. Andθwill become negative when local arc appears on the insulator surface. Therefore,θcan be taken as one of characteristic values to predict the risk of pollution flashover. Besides, the experiment results also indicate that amplitude of leakage current is a power function of SDD, while an exponential function of RH at a certain temperature.In this paper, the characters of leakage current in frequency domain are analyzed and the results show that harmonics do exist in the leakage current, and the greater is the RH and SDD, the higher is the maximum of harmonics such as the 1st,3rd and 5th components. Meanwhile, the amplitude ratio (δ) of the 1st to the 3rd harmonic will become greater with the increase of RH first and then tends to get saturated gradually.Through the analysis in this paper, three leakage current characteristics values, namely Ih (the magnitude of leakage current),δ(amplitude ratio of the 1st to the 3rd harmonic)andθ(phrase difference between leakage current and operation voltage) can be taken as the major characteristic values for predicting the risk of pollution flashover of insulators. At last, a model for estimating contamination extent of insulator surface and predicting the risk of pollution flashover is built based on RBF neural network.