Study On The Forecasting Model And Method Of Flashover Based On Characteristics Of Leakage Current Of Polluted Insulators

Contamination flashover is one of important factor which affects the safe operation of outside insulators, thus causing power system fault and huge economic losses. Although great deals of the pollution flashover test and theoretical exploration have been done at home and abroad, and many valuable results have been made, but the pollution flashover are still frequently reported, so flashover mechanism and prediction of flashover on contamination insulators still need to be depth study. Experimental studies and operating experience shows that the development of arcs on insulator surface are closely related with the leakage current, while the leakage current is a combination effect of contamination and humidity on surface of insulators. Therefore, the systematic studying leakage current characteristics of polluted insulators, and establishing security status assessment method based on leakage current characteristics, which have important academic significance and engineering application value.The characteristics of leakage currents on AC pollution insulators were studied in the artificial climate chamber in this paper. Combining with pollution flashover mechanism, the affections of the salt deposit density and humidity on the leakage current were in-depth analysis. The relationship between the development of leakage current and discharge process of the polluted insulators was studied by the high-speed camera. Based on the above analysis, flashover criterion and forecasting methods were put forward. The following is the main conclusions obtained in this article.According to the test results of the single XP-70 insulator carried in the artificial fog chamber, it is analyzed that the effects of the applied voltage(U),the equivalent salt deposit density (ESDD) and the relative humidity density(RH) on the leakage current of the insulator based on the least squares algorithm. The result shows that the peak maxim of leakage current increases with the increase of ESDD by a power function, while increases with the increase of both U and RH by an exponential function. The result could be used as to estaminate the leakage current of contamination insultors.The phase (? ) between the leakage current and operated voltage is studied under different RH. It was found that, ? decreases with increasing of RH, when relative humidity is low, ? > 0, while the relative humidity is high, ? <0, ? becomes almost zero when RH is about 80%. The relationship between the waveform distortion of leakage current and the total harmonic distortion (THD), low frequency harmonic components and high frequency harmonic components were analyzed by FFT. The low frequency 3rd harmonic component is main component of the distortion of leakage current waveform correlated with arc discharge, high frequency harmonic components of LC, e.g. 13th, is main component of random noise. The THD increases slowly with the increasing of RH under light contamination, while the THD decreases firstly then increases with the increasing of RH under moderate and heavy contamination, there is a minimum THD when RH with increasing of RH under moderate and heavy contamination, the discharge appears on the surface of insulators when THD get at the THDmin.Comprehensive consideration of the characteristic parameters of time domain, frequency domain and phase of leakage current, it would be used to evaluate the safety level on surface of polluted insulators.The Ih, ? and THD were proposed as characteristic parameters to evaluate the safety condition of polluted insulators for the first time. An ANN with fuzzy output was developed to evaluate the safety condition of polluted insulators. The output parameters were fuzzified into four fuzzy subsets, the capability of ANN was validated by testing samples.The characters of leakage current XP-160 and LXY4-160 insulator strings are investigated by Hilbert transform, non-parametric density estimating through histogram method and the level crossing activity based on the test results according to that the stochastic distributing characteristic of envelop of leakage current. The flashover prediction criterion is found based on above analysis for the first time in China, such as, to XP-160,Ï„135 is 93 ms which indicates that the mean time duration of 93 ms for the leakage current envelop staying above 135 mA will sufficiently cause flashover; to LXY4-160,Ï„110 is 105 ms, which indicates that the mean time duration of 105 ms for the leakage current envelop staying above 110 mA will sufficiently cause flashover.