Research On Transient Power Quality Disturbance Detection Analysis And Evaluation

With the rapid development of technology, process of production become automation and equipment become intelligent, the users have higher demands for power quality, the power generation and supply side are more concerned about how to improve power quality. Power system disturbances are divided into steady-state and transient. Relative to the transient power quality, the algorithm of steady-state power quality analysis has been more mature, transient disturbances is rapid and harmful, its theory and application of software and hardware are still in a preliminary stage. Transient power quality disturbance detection and analysis system needs to detect and locate the disturbance quickly, to determine the cause of the disturbance and resume promptly, needs to identify the type of transient power quality disturbances. The analysis of the disturbance characteristics under the specific operating environment formulating measures could improve the transient power quality.The model of transient power quality detection and analysis include locating the time, extracting the eigenvalues and identifying the type of disturbances. Position disturbances with wavelet transform modulus maxima. Proposed the new algorithm based on wavelet transform modulus maxima perturbation theory, mark the type with the modulus maxima, according to the classification results to select the best wavelet packet transform or wavelet coefficients as feature vectors. Put the feature vector in SVM to recognize disturbances. Using MATLAB, compared to the algorithm based on wavelet (packet) transform and SVM, verify the accuracy of the new algorithm by calculate the amount of errors, etc., cited four examples to illustrate the validity of the simulation method.Proposed the new algorithm based on wavelet transform modulus maxima theory and BT-SVM, using different methods of analysis based on the differences of transient power quality disturbance characteristics for the feature extraction and the disturbance recognition, the algorithm ensure real-time, while improve the accuracy of detecting transient high-frequency disturbance analysis.