Research For Orientation Of Pollution Source Of Voltage Flicker Based On Wavelet

In recent years, various types of loads emerge along with rapid economic growth in China. Power quality of grids has been deteriorated severely. Many factors contribute to this deterioration. Those loads with jerky and non-linear characterizations are usually deemed as main pollutants to power grids in sense of quality problems. Power quality problems like voltage distortion, fluctuation and flicker etc. appear more frequently. Deterioration of power quality has make power grids more liable to contingencies and may cause damages to customers. There exist numerous power quality problems and their features are quite different from each other. For certain power quality problems, effective tools for indices surveillance and pollution sources location are still not available until now. It is not possible to find out causes of pollution and responsible entities without such tools. These lead to increasingly enlarged contradictions between power suppliers and power customers and make it difficult to design power quality improvement measures appropriate to the situation.Research work presented in this paper focus on issues related to voltage flicker, one of the most important power quality problems. Orientation of sources of pollution is investigated based on the non-linear load theory. And, square demodulation is inducted into wavelet transform for extraction of the envelope of voltage flicker. The proposed method can not only avoid uncertain errors caused by synchronized signals, but also keep the time-domain resolution of modulation signals. The method detects the envelope of voltage flicker after separating high harmonics in case that the two kinds of pollution coexist. This improves the accuracy of voltage flicker detection and realizes the separation of two pollutions. The proposed algorithm is programmed in Matlab and tested in two systems (a simulation system constructed in the RTDS environment and a real system). Results from both systems illustrate validity and practicality of the proposed algorithm.