The Detection Of Power Quality Disturbances Based On EEMD Decomposition

In recent years, the load structure in modern power system has undergone major changes. The proportion of nonlinear, volatility, impact and unbalanced load in the power grid is increasing gradually. The problem of power quality has become more and more serious. In order to improve the power quality, it is necessary to detect and analyze the power quality disturbance signal effectively. In this paper, we study the digital processing method of power quality disturbance signal. In view of its deficiency, we propose a Ensemble Empirical Mode Decomposition(EEMD), and improve the algorithm, then apply them to the power quality disturbance detection.First of all, this paper introduces the research status of power quality disturbance detection, expounds the principle and implementation steps of EEMD algorithm, Aiming at the deficiency of EEMD algorithm in power quality disturbance detection, the possibility of the improvement of EEMD algorithm in power quality disturbance detection is discussed.Secondly, we propose a power quality disturbance detection method based on denoising EEMD algorithm. According to the characteristics of power quality disturbance signal, we improve the EEMD algorithm. The addition of a single auxiliary noise is changed to a pair of positive and negative white noise with equal amplitude to improve the accuracy of decomposition; The amplitude of adding white noise and the integration of the average number for the EEMD algorithm is artificial determination. We adopt a parameter adaptive method based on probability statistics to solve this problem; For the decomposition of the various IMF components which is easily impacted by the noise, we propose a threshold denoising method. Combining these three processes, we obtain an denoising EEMD method, and apply the method to the detection of power quality disturbances. The effective detection of single disturbance and complex disturbance is realized.Finally, we propose a power quality disturbance detection method based on adaptive EEMD algorithm. The added white noise in the screening of the intrinsic mode functions in EEMD algorithm is the same. The high-frequency components and low-frequency components are different in sensitivity to noise, adding the same white noise may cause the decomposition mode mixing phenomenon. To solve this problem, we analyze the problem of adding white noise to the original algorithm, and construct an auxiliary white noise whose amplitude and frequency is changed by the sine rule, then replace Gauss white noise in the original algorithm to white noise we constructed. Combining the denoising EEMD algorithm,we get an adaptive EEMD algorithm. The method is applied to the analysis of power quality disturbance signal to improve the adaptive and accuracy of the disturbance signal decomposition. It is advantageous to obtain the characteristic of power quality disturbance.This two improved EEMD algorithms are used to detect harmonics, voltage sag, temporary rise, oscillation, pulse and single power quality disturbance signal, and a variety of complex disturbance signals are detected. The simulation results show that both the denoising EEMD algorithm and adaptive EEMD algorithm can accurately detect various types of disturbances'starting and stopping time and the magnitude of the disturbance; The adaptive EEMD algorithm has better detection precision than denoising EEMD algorithm in fewer selecting times. Therefore, the test results are better.