Harmonic And Voltage Flicker Detection Of Power Quality Base On Independent Component

With the development of electric power technology and the wide application of electric & electronic devices in power system, power quality problems have already become the focus to which both suppliers and users of electricity have paid close attention. Since the increase of non-linear loads, the power system harmonic has serious deterioration of power quality, endangering the power system security, and impacting the user's normal work. So the harmonic problem became the focus of study, especially the power system harmonics measure is of great significance. In addition, with the improvement of power quality and the consideration of flicker-sensitive users, It has become more and more important to detect flicker in order to reducing the influencing.Based on the summary of the harmonic and the algorithm of the flicker detection, A method of Power Quality's harmonic detection using improved Fast-ICA and the voltage flicker detection using dynamic independent have been proposed. The main contents are the following:1,The improved Fast-ICA to detect the harmonic and flicker has been dicussed. The basic principles of Fast-ICA algorithm have been introduced at first, then Newton's iteration has been improved to reduce the number of iterations and speed the convergence rate. According to the principles of independence of Negentrope Maximum, the amplitude of the harmonic signal has been revised to obtain the estimation of real signal. FFT has been used to calculate the amplitude and frequency of the separative signal. In the end, the harmonic and flicker detection have been achieved.2,The method of the voltage flicker detection using dynamic independent has been discussed. Based on the summary of the'static'independent component analysis, the algorithm of dynamic independent has been proposed by changing the iteration form. Then according to the principles of independence of Kurtosis Maximum, the amplitude of the harmonic signal has been revised to obtain the estimation of real signal. FFT has been used to calculate the separative signal, therefore, the accurate amplitude and frequency have been obtained and the voltage flicker's detection has been completed.3,Based on the analysis of the two algorithms, simulation experiment has been made separately using the Fast-ICA algorithm, the improved Fast-ICA algorithm and dynamic independent component algorithm. Experiment results show that the improved Fast-ICA and dynamic independent component algorithm on this paper have good performance in separating the harmonic and flicker signals and in detecting the amplitude and frequency. The correctness and effectiveness of the two algorithms have been proved.