Power System Small Disturbance Stability Analysis And Control With Saturation Element Of Excitation System

Generator excitation system plays a fundamental role in power system. Saturation element of excitation system not only affects on transient stability but also leads to the low-frequency oscillations which result in static instability problems possibly. In studies of system stability, the so-called "region" approach has been proposed and is gradually being accepted by many scholars because the "region" can provide some important information which can't be got by using time-domain simulation approach and frequency-domain approach. This dissertation thoroughly studies on the stability problem with saturation element of excitation system from the point of view of the region of attraction.The global stability problem of single-machine infinite-bus is studied in this dissertation by using linearized model. The mathematic models are established with saturation element of excitation system under three typical control strategies:automatic voltage regulator (AVR), automatic voltage regulator plus power system stabilizer (AVR+PSS) and linear optimal excitation control (LOEC). Based on impact maps and quadratic surface Lyapunov functions, the global stability of three linearized systems under different damping factors and line impedance values is recognized. The conclusion shows that conditions of global stability can't be satisfied entirely for the linear system after considering saturation element even if the real components of all eigenvalues are negative. So, it is realistic and feasible to study on the local stability of saturation system. Furthermore, the switching time is estimated in the dissertation. After studying on the effect of damping factor on the switching time of saturation system, it is concluded that smaller damping factor is, longer switching time is.The region of attraction (RA) of power system is estimated with saturation element of excitation system. The linear matrix inequalities (LMIs) are applied to estimate RA of the system with saturation element of excitation system under linear optimal excitation control after transforming the problem of estimating RA of the system under saturated linear feedback control law into LMIs formulations. The developed method is applied to choice the preferable weighting matrix using the volume of n full-dimensional ellipsoids as the criteria, and the expectable result is shown in the simulation test. Based on the nonlinear relation between the excitation limit and the volume of ellipsoidal RA, it is recognized that the volume of ellipsoidal RA is proportional to the cube of the excitation limit. To visualize the high dimension ellipsoidal RA approximately, a novel projection approach based on singular value decomposition for high dimension ellipsoidal RA is proposed in multi-machine case.Based on RA of saturation system, a novel algorithm is proposed to calculate the boundary of the small signal stability region (SSSR), and used to study on effect of saturation element on SSSR. The case demonstrates that the gained SSSR based on the proposed algorithm is equal approximately to the one using Hopf bifurcation when the criteria of RA is very small, and the proposed algorithm gets smaller SSSR when the criteria of RA is enlarged. These results show that SSSR may consider of the disturbance, which enlarges the facet of studies on SSSR. Moreover, this dissertation analyzes effect of weighting matrix, damping factor, criteria of RA and load model on SSSR, and derives some important conclusions.Finally, the multi-objective saturation excitation controller for nonlinear system is designed in this dissertation. In order to guarantee the quadratic stability in the given RA, the relation between RA and initial state region is considered of in the procedure of designing the saturation controller. Numerical simulations are provided to validate the effectiveness of the proposed controller.