Electrical Power Quality Disturbance Analysis Based On S Transform And Wavelet Denoising

Power quality disturbance will lead to equipment other than the normal work, overheating, the motor stopped etc. It will bring great damage to production and living conditions. In recent years, with voltage sag the main form of expression, transient power quality problems have caused serious losses for industrial production. Therefore, this paper focuses on the analysis of transient power quality disturbance, the disturbance signal denoising and disturbance classification research. The main research contents are as follows:First of all, according to the non-stationary of the signal, using S transform and MATLAB simulation to analyse. The height and width of the S transform window function can vary with the changes of the signal frequency and it has good time-frequency resolution. At the same time, using MATLAB simulation program to pruduce transient power quality disturbance of eight single and two complex disturbance mathematical model.S transform analysis of various disturbance signals,to extract the S transform 13 times frequency curve of S matrix, and the mean square curve line and column maximum curve, the starting and ending time of the disturbance signal are respectively, frequency and amplitude information.Then, considering the influence of noise on the results of the perturbation analysis.S transform in the noise environment can still get the accurate signal frequency information and amplitude information, but can not eliminate the influence of noise on the mutation information.For the accurate positioning of the starting and ending time of the disturbance, we use the denoising method of wavelet transform in power quality disturbance signal denoising of transient electrical noise and improved threshold function. The threshold function in the estimation of wavelet coefficients smaller errors than soft threshold function, which has better smoothnessthan the hard threshold function of reconstructed signal.The simulation results show that the improved threshold function denoising, denoising results after S transform, can get accurate information of disturbance signal disturbance.Finally, using decision tree classification method of transient electric energy to recognize quality disturbance. Using the feature extracted by S transform to build a simple rule tree.Classify and recognize eight single disturbance and two complex disturbance. The decision tree realizes the classification by using tree structure. This method is simple, easy, and has a high accuracy.