Research On Power Quality Disturbance Recognition Algoriyhm Based On S-transform

With the rapid development of science and technology and industrial intelligence,the influence of power quality on aerospace,automobile,electronics manufacturing is becoming more and more significant.Therefore,the management of power quality problem is widely concerned and studied.The key step of power quality management is the identification of power quality disturbances,which is power quality disturbance detection and classification recognition.This paper sums up the hazards caused by power quality problems,and according to the power quality related standards and the development status at home and abroad,the power quality disturbance identification algorithm is studied in two aspects,which are power quality disturbance detection and classification recognition.Aiming at the problem of power quality disturbance detection,12 kinds of power quality disturbance mathematical models are established according to the power quality problems common in power system.This paper studies the commonly used power quality detection algorithm(short time Fourier transform and wavelet transform),and uses MATLAB to carry on the simulation experiment.And the characteristics of each algorithm and the advantages and disadvantages of each algorithm in detecting different types of power quality disturbances are described.Aiming at the shortage of common power quality detection algorithms,this paper focuses on the method of power quality disturbance detection using S-Transform.Based on the analysis of the principle of S-Transform,this paper proposes a Segmented and Modified S-Transform detection method.The method overcomes the disadvantage that the difference method,rms method and Fourier transform can not detect all power quality disturbance types,the wavelet transform is affected by noise and the short time Fourier transform window width is fixed.In this method,the detection signal is divided into low frequency,intermediate frequency and high frequency bands,the corresponding window width adjustment factor exists in each band,and the time-frequency resolution of S-Transform can be changed by changing the window width adjustment factor.In order to get the optimal window width adjustment factor,the relationship between the disturbance parameter estimation errors and the kurtosis at disturbance area is analyzed.The detection ability of Segmented and Modified S-Transform is better than that of common detection algorithm by simulation.For the classification and identification of disturbance signals,this paper firstly uses Segmented Modified S-Transform to analyze various disturbance signals,and the analysis shows that the time-amplitude and frequency-amplitude eigenvectors can reflect the main characteristics of disturbance signals in time domain and frequency domain.Secondly,extracts 7 eigenvalues from the eigenvectors,and builds the DecisionTree Classifier of 14 classification rules by using the eigenvalues.The Decision Tree Classifier that is realizes the automatic classification of disturbance signals,and has high recognition accuracy rate.Based on the proposed new method of Power Quality Disturbance identification,this paper designs a kind of Power Quality Disturbance identification device that is based on dual core CPU(DSP+ARM)chip OMAP-L138.The device has the functions of basic power parameter measurement,Power Quality Disturbance identification,Power Quality parameter measurement,historical data storage and network communication,etc.Finally,the device is tested experimentally.The test results show that the device basically achieves the designed functional requirements.