Research On Detection And Recognition Method Of Power Quality Disturbance Signal

With the increasing scale of modern power grids,the development of intelligence is becoming more and more obvious,power companies and users pay more attention to power quality issues.The power quality problem is manifested as the power quality disturbance signal generated in the power system.In the past,the power quality disturbance analysis is mainly the steady-state power quality disturbance signal detection.In recent years,with the impact load connected to the power grid,transient power quality disturbances such as transient oscillation,transient pulse and voltage sag occur.The identification of power quality disturbance signal in power system is the basis of power quality analysis.Only by accurately identifying the type of power quality disturbance signal can we select a more accurate and efficient detection method to analyze the disturbance signal and obtain the characteristic parameters of the disturbance signal,so as to effectively take improvement measures to improve the quality of power in the power system.However,the development of the theory of power quality disturbance signal recognition is not mature enough.There is no unified recognition model;the characteristics of various detection methods are different,and the effect of disturbance signal detection is also different.In this paper,the support vector machine(SVM)is applied to the identification of power quality disturbance signals.On this basis,the Hilbert-Huang transform(HHT)is used to detect the disturbance signals,the corresponding shortcomings are improved,and the classification and detection effects are verified by simulation.The main research contents of this paper are as follows:Firstly,the power quality disturbance signal detection method is studied,and the basic theory of HHT algorithm is emphasized,including the two core contents of empirical mode decomposition(EMD)and Hilbert transform.The HHT is used to analyze the disturbance signal from both the time domain and the frequency domain,and accurately detect the specific parameters such as the start and end time and the amplitude change of the disturbance signal.Secondly,the SVM-based power quality disturbance signal recognition method is deeply studied.According to the signal characteristics of the disturbance,the wavelet packet decomposition method is used to extract the disturbance signal feature quantity.These feature quantities are used as the SVM input signal for the SVM to automatically identify the disturbance type.The improved particle swarm optimization algorithm(IPSO)is used to optimize the penalty parameter and kernel function parameter of SVM.The simulation study of the IPSO-SVM classifier identifies the disturbance signal and compares it with the classification effect of the SVM and PSO-SVM classifiers.Finally,based on the identification of disturbance signals,the application of HHT in steady-state,transient and composite power quality disturbance signal detection is studied.For the disturbance signal such as transient oscillation and voltage sag,the simulation results verify the detection effect.