Research Of Power Quality Disturbance Detection And Identification Based On Data Mining

With the growing load capacity in power grid and development of modern power electronics technology in power system, power quality problems have become increasingly prominent. More attention has been paid to power quality disturbance detection and identification.In the dissertation, the research involves power quality disturbance detection and identification by wavelet transform and data mining. Firstly, power quality disturbance is detected by wavelet transform modulus maxima. Secondly, power quality disturbance characteristics are extracted by wavelet decomposition and power quality disturbance is identified by data mining. Experimental results validate the accuracy and efficiency of the method in accuracy and running time. Further, the impact of the identification of disturbance based on different parameters is discussed, and de-noising method of power quality disturbance based on energy features estimation is used for emerging issues in this process. Finally, combining data stream technology, on-line disturbance classification is completed.