Study The On-line Monitoring Technology Of Metal Oxide Arrester

To limit the lightning overvoltage and operating overvoltage, MOA are widely used because of its good nonlinear characteristics. But due to the MOA is online work, Affected by many factors that can occur gradually ageing degradation phenomenon, so it is quite necessary to electrical properties on-line monitoring. First MOA aging properties and its low current area equivalent model are analyzed. Secondly, in the predecessors of capacitive current compensation algorithm based on the further optimization, reduce the influence of harmonic voltage on online monitoring algorithm. Then, in order to further improving the accuracy of MOA on-line monitoring, Use Matlab to simulate monitoring indicators of aging sensitivity existing problems. Finally, using the relatively mature genetic algorithm to proposes a new aging on-line monitoring technology, and the leakage current have noises in practical application, a new de-noise algorithm is put forward. Study the following conclusions:The basic harmonic resistive current irl and the third resistive current ir3 generated by the basic voltage can reflect the operation state of MOA.The traditional MOA capacitive current compensation algorithm and its improved didn’t take the harmonic voltage into consideration. It will produce errors in theory. The study shows that use the proportional characteristic of Capacitive current and calculate the real capacitance, so the resistive current is deduced and in order to eliminates a part influence of harmonic voltage on the resistive current and improve the precision of irl and ir3. The results simulated by MATLAB show that improved capacitive algorithm and the algorithm to calculate irl mean error of 3.35%,2.38% respectively, the error of the ir3 were 21.11%,10.40%, has the very good stability, and can accurately response earth-sized aging degradation.In order to improving the accuracy of on-line monitor on the degradation of MOA. Use Matlab to simulating calculation about the different types of the fundamental harmonic resistant current and the third harmonic component at present. The study show that the value of fundamental harmonic resistant current and third harmonic component increases with the initial phase of harmonic voltage and the phenomenon of the third harmonic component is obviousness, but the initial phase of harmonic voltage in power grid is uncertainty, so choose the fundamental harmonic resistant current and the third harmonic component produced by the fundamental voltage can more accurately reflect the degradation of MOA.To improve the efficiency of MOA aging monitoring, the MOA on-line monitoring based on genetic algorithm (GA) is presented. The technology uses the genetic algorithm optimization computing power to solve the assessed parameters which change during the lifetime of the MOA and monitoring the degradation of MOA. The studies show that the monitoring technology is based on GA can fitting the leakage current of calculation to the actually measuring leakage current of MOA (the standard deviation is 1.498%) and calculate the parameters of simplified equivalent model k and c which change during the aging of MOA, in order to achieve the aging of MO A monitoring. Additionally, the result of Matlab simulations demonstrates that the maximum error of k and c is 0.08%,0.05%, respectively, which was calculated by this algorithm under the effects of the operating voltage harmonics、voltage fluctuation and frequency fluctuation. So the operating voltage harmonics、voltage fluctuation and frequency fluctuation in power grid have low influence on the proposed algorithm. Finally, the result of MOA on-line monitoring is improved.Achieve to monitoring the aging of MOA, we need to measure the leakage current and leakage current signal is very weak, which will be covered by interfering signal in the power system, so the leakage current should be de-noises. According to the problem that the actual measurement leakage current for MOA includes an interference signal and which covers the original leakage current, we proposed a comprehensive de-noising algorithm which composed of Kalman Filter and a Stationary Wavelet Transform. At first, using Stationary Wavelet Transform to decompose the leakage current as approach and detail signals. Second, using Kalman Filter to processes the approach and detail signals. Finally, we reconstruct the approach and detail signals, and get the de-noised leakage current. The study show that the comprehensive de-noising algorithm is better than the Kalman Filter and Stationary Wavelet Transform, the SNR of the leakage current is obvious improvement, Provides access to the "purity" leakage current.