Power Quality Disturbance Analysis And Identification Based On ASW-ESPRIT

With the innovation of science and the healthy development of national economy, the requirement of power quality is getting higher and higher for all walks of life. In addition, due to the growing application of power electronic equipment and the growing use of a variety of nonlinear load, impact load or fluctuation load, the power quality is influenced by various aspects and the pollution of power quality problems become increasingly serious. At the moment, for stationary signal analysis of power quality problems, the algorithms have been developed relatively mature at home and abroad. But for non-stationary signal analysis, both theoretical research and equipment development is still in the stage of trying and explore.The detection, analysis and recognition of power quality disturbance is the first guarantee to improve power quality. In this paper, the detection, analysis and identification methods of seven kinds of main power quality disturbance are studied. Power quality disturbance identification based on ASW-ESPRIT algorithm is proposed to classify the power quality disturbance. It designs and implements the power quality disturbance identification system. In this paper, the main works are as follows:(1) Because traditional sliding ESPRIT algorithm could only use the fixed sliding window to intercept data, this paper proposes adaptive sliding window to take the place of fixed sliding window and apply it to the power quality disturbance detection and analysis. According to the characteristics of the non-stationary signal of power quality disturbance, firstly, this method uses adaptive sliding window to block the signal data. Then it uses ESPRIT algorithm to deal with the data of each block, to detect the frequency and amplitude information. We receive the results obtained the time-frequency distribution of the signal. Finally, we do the simulation experiment of the nonstationary power quality disturbance signal, the simulation experiment results show that the new method is suitable for dynamic power quality disturbances detection and has practical application prospects.(2) This paper introduces the classification tree algorithm based on logic circuit to identify the PQD signal. This method firstly uses ASW-ESPRIT time-frequency analysis to analyse the disturbance signal, extracts the seven characteristics of disturbance signal. Then it generates the classification tree identification methods for the disturbance signal and the normal signal. Finally, in order to realize the classification of power quality disturbance signal, classification tree is completed by combinational logic circuits and the hardware circuit diagram. The experimental results show that the method has a simple structure, high classification accuracy and it is easy to implement.(3) This paper implements the power quality disturbance identification system based on DSP development platform which is based on the power quality analysis algorithm. The platform consists of portable PQD identification device based on DSP hardware system and the host unit. The upper machine system is implemented with VC++6.0programming language design of the software. This system is mainly applied in the transient power quality disturbance signal detection, analysis, localization and identification. It can detect equipment and line fault promptly and provide an effective tool for the practical application of power quality analysis.