Overvoltage Monitoring System For Substation And Its Waveforms Analysis

Now the economic society of our country is in the stage of rapid development, demand for electricity has long maintained a rapid growth, the voltage level of the power system, transmission distance and scale expanding. Control of power grid operation mode and the corresponding system complex degree more and more high, how to reduce the risk of accidents to maintain safe and reliable delivery of electrical energy has become a major challenge to the construction of strong smart grid current. Research shows that, all kinds of equipment insulation level and mode of operation is an important factor affecting the reliability of the power system, and the main cause of failure of all kinds of equipment in the insulation system is the Internal Overvoltage and the external overvoltage. Processing for real-time monitoring and fault signal to voltage signal substation analysis helps the staff to grasp the operation of the system, find the cause of the malfunction, timely detection of insulation weak points in order to solve the rapid troubleshooting and reduce economic loss. And through the analysis of the causes of the fault voltage data can also help to provide a more reliable grid design scientific basis to determine the insulation level of power grid equipment and all kinds of protection devices to select a reasonable allocation of help.This paper introduces a ZnO lightning arrester voltage sensor, through the performance, security and other aspects of the analysis of its feasibility. In order to verify the reliability and accuracy.In this paper, using the simulation calculation, a comparison test and field measurement. By analyzing the final result, ZnO lightning arrester voltage sensor has good performance in frequency response and linearity, can use it to realize the collection of the transient overvoltage monitoring, signal data obtained with high accuracy.As for the study of transient overvoltage data collected in this paper, the transient overvoltage occurrence mechanism is analyzed in detail. Modeling of the electromagnetic transient simulation software ATPDraw, through the combination of simulation and actual measurement of the waveform of voltage characteristics, and based on the tree structure of the hierarchical classification of transient overvoltage.Feature selection plays an important role in the recognition system, based on the analysis of time and frequency domain analysis, wavelet transform and singular value theory and methods to process overvoltage data, the local characteristics of the signal to be prominent, various characteristic parameters extraction. According to the characteristics of all kinds of transient overvoltage, selecting corresponding parameters characterize the voltage of each layer classifier is established.In the process of transient over-voltage identification system, the traditional two classification support vector machine theory and multi class support vector machine combined adopted multi-layer progressive classified development program. In order to further improve the accuracy and the performance of recognition system, firstly the PCA and normalization method to preprocess the data, and then use the cross validation, genetic algorithm and particle swarm optimization algorithm of three kinds of means of each layer and the support vector machine parameters are optimized to select the optimal method to determine the parameters of SVM. Finally, using the MATLAB graphical user interface to make a transient over-voltage identification system GUI, to further enhance the interaction, visibility and operability.