Study On Modal Identification And Suppression Of Low Frequency Oscillation In Power System

With the continuous expansion of the interconnection scale of China’s power system,at the same time,the increasing power consumption in the power industry and the extensive application of fast excitation systems,all these factors will lead to the decline of the damping level of the system,so the probability of low-frequency oscillation is increasing day by day.To study the problem of low-frequency oscillation,the first task is to accurately identify the data of low-frequency oscillation.If the low-frequency oscillation is not analyzed quickly and accurately and corresponding suppression measures are taken,the stability of power system may be affected and certain economic losses may be caused.Therefore,how to quickly and accurately identify the low-frequency oscillation mode and take corresponding damping suppression measures is one of the keys to realize the stable operation of power system.In this paper,firstly,the concept of low-frequency oscillation is expounded in detail and its generating mechanism is discussed.Then,the mathematical models of single machine infinite and multi-machine systems are established,and the essence of low-frequency oscillation is analyzed by using negative damping mechanism.Then,the modal analysis method of low-frequency oscillation and the main measures to suppress it are discussed.Finally,the functions of each link in power system stabilizer(PSS)are analyzed,and the essence that it can improve system damping and suppress low-frequency oscillation is revealed by using damping torque method.Aiming at the problems of noise interference and inaccurate identification parameters in modal identification of low-frequency oscillation signals,this paper proposes a new method based on stein unbiased risk estimate(SURE)wavelet denoising and TLS-ESPRIT to extract low-frequency oscillation parameters.Firstly,the low-frequency oscillation signal is filtered by SURE denoising technology to effectively filter out the interference of random noise,and then the modal identification of the filtered signal is carried out by TLS-ESPRIT algorithm.In the key order determination of the algorithm,this paper proposes an order determination method based on normalized singular entropy to improve TLS-ESPRIT method.The effectiveness of the improved method is verified by the simulation analysis of numerical signal examples and PMU data examples measured in North America power grid,and compared with the traditional identification algorithm.In the aspect of low-frequency oscillation suppression,this paper simulates and analyzes the high efficiency of PSS for low-frequency oscillation suppression under two different working conditions: excitation voltage step and single-phase short circuit,and makes statistics on related characteristics.Subsequently,this paper proposes a method to suppress low-frequency oscillation of power system based on modal identification,and gives the related flow chart.Finally,based on PSASP8-machine 36-node system,an example is simulated,and different PSS parameters are selected to simulate and analyze the suppression effect of generator power angle,active power and other different outputs,which verifies the feasibility of this suppression strategy,and compares and analyzes the suppression effect of the same PSS parameter in different installation positions.