Research On Power Quality Disturbance Detection And Compression Technology

With the rapid development of electric power industry,all kinds of power electronic devices and non-linear load are widely used in power system,which make the power supply unstable in power grids,and the problems of various power quality(PQ)disturbances are increasingly serious.Meanwhile,some microelectronic devices are sensitive to voltage fluctuation.Therefore,a large number of scholars have done deep researches in this area and achieved some achievements.Power quality analysis mainly includes three aspects: power data compression,power quality disturbances classification and disturbance parameters estimation.The existing algorithms have achieved better performances than the traditional algorithm.However,these algorithms need to be improved in terms of parameter estimation accuracy,disturbance detection accuracy and real-time performances.It is necessary to have a further study on PQ analysis methods with high efficiency and stability.Power quality disturbance detection is the key to improve PQ,which is also an important part of PQ data compression.The research work of this thesis mainly involves two aspects: power quality disturbance detection and PQ data compression.The main achievements of this thesis are as follows.(1)A high-speed and multi-channel PQ signal sampling system based on FPGA and AD7606 is designed,which provide a reliable data source for PQ disturbance signal acquisision.Field test show that the data acquisition of sampling system can work normally.(2)For the respect of PQ transient disturbance detection in the strong tracking filter(STF),there exist some problems that the low parameter estimation accuracy under strong nonlinear system and high calculation complexity of multidimensional filter model.This thesis proposed a new disturbance detection method based on modified strong tracking Kalman filter combining STF with unscented Kalman filter(UKF).This thesis uses the suboptimal fade factor to adjust the weight of the process noise covariance matrix adaptively when the state variable is mutated,which deduce the solution of the fading factor under the orthogonal condition of the residual sequence at different times.The proposed algorithm improves the estimation accuracy of the filter while enhancing the robustness of STUKF under the mismatch model.Secondly,Comparing with the standard STUKF,The algorithm has only two UT transforms,and the computational complexity is reduced.The experimental results show that the proposed algorithm is more accurate than those of the STUKF in locating the starting and ending time of voltage swell and sag and tracking the amplitude and phase changes in a variety of complex situations,which verifies that the modified STUKF is a solution with good real-time and reliable PQ transient disturbance detection.(3)For the existing problems of the low accuracy in steady parameters estimation and week robustness for different types of disturbances in parametric power data compression method,this thesis proposes a parametric power data compression algorithm based on three-parameter sine-fit and dynamics.Dynamics can detect transient disturbance and steady disturbance in noise,so that the steady and transient disturbances can be processed in different way,which avoid the low compression ratio of uniform processing method.The 3PSF can calculate the amplitude and phase of steady component by separating each harmonic successively with high precision.What's more,the adaptive notch filter and interpolated Fourier transform ensure the accuracy of the fundamental frequency estimation in various disturbances.Combining of the above makes the steady and transient components of PQ disturbance signal separate perfectly,which improve the overall compression performance of the method.The experimental results show that the proposed method has a better compression performance than the existing parametric compression method,and it is more suitable for single and composite PQ disturbance signal.