Multi-parameter Sensitivity Analysis And Optimization Research Of Power System Stabilizer

At present,long-distance large-capacity transmission and interconnection of power grids have great impact on the stability of power systems.The low-frequency oscillation that belong to the small-signal stability has an increasingly greater impact on the stability of the power system.A large number of research shows that power system stabilizer(PSS)is the most commonly used excitation controller in modern excitation system.It can effectively suppress low frequency oscillation by properly setting PSS parameters.This paper is mainly divided into two parts: introducing the global sensitivity analysis into the tuning of PSS parameters;optimizing PSS parameters based on the sensitivity analysis results.First,the principle of phase compensation by PSS is analyzed,and the superiority of Prony method in analyzing low frequency oscillation signal is verified.Build a singlemachine infinite-bus system with excitation voltage regulator(AVR)and PSS.Selecting ITAE of angular velocity,power angle and active power as an index of performance in singlemachine infinite-bus system.The Latin hypercube sampling and Monte Carlo method based on variance are used to calculate the global sensitivity of the PSS parameters.With multiparameters changing at the same time,the contribution of each variable to the output variance is calculated,and the sensitive parameters and sensitive intervals in the PSS under lowfrequency oscillation are found.The sensitivity analysis results are preliminarily verified by dynamic simulation.The interactive effects of the frequency and the PSS parameters on the unit ITAE value in the entire low-frequency oscillation frequency domain(0.1-2.5 Hz)are analyzed.In order to solve the problem that the classical genetic algorithm is difficult to break through the local optimal,an improved genetic algorithm is proposed,which uses the cross operator of adaptive probability and the mutation operator with different mutation probability to different fitness.In the single-machine infinite-bus system,I simulate low frequency oscillation,mechanical power disturbance and three-phase short-circuit operation,and use the improved genetic algorithm to optimize the PSS parameters.The oscillation characteristics of the infinite system are analyzed by using the unit model dynamic simulation and Prony analysis.The simulation results show that after the Sobol sensitivity analysis of the PSS parameters obtains the sensitive parameters and the design range,using the improved genetic algorithm to optimize the PSS can effectively suppress the low-frequency oscillation.Finally,simulate and extract the unit oscillation amplitude at different frequencies in the entire low-frequency oscillation frequency domain.The dynamic mechanism of PSS parameter design is explained.