COMPUTER MODELLING SIMULATION AND ROBUST MULTIVARIABLE CONTROL OF POWER SYSTEMS

Youssef  A. SMAILI

Electric power systems are faced with dynamic stability problems that impose undesired limitation on their power transfer capability. These problems are experienced in the form of low frequency oscillations that could lead to system instability and loss of synchronism of some units. The modeling of the power system for dynamic stability studies as well as the control of related problems was considered in this study.

The nonlinear mathematical model of a single machine infinite bus system and a multi‑machine power system were derived. In these models the loads were represented with static models as frequency and voltage dependent. Also, the full order nonlinear current model of the synchronous machine was used. These models were used to perform the nonlinear simulation for the controller evaluation. The nonlinear simulation was carried out using a program written in Advanced Continuous Simulation Language (ACSL). The linear models were also derived and analyzed using MATLAB routines.

The dynamic stability margins of the power system are traditionally enhanced by controllers referred to as the power system stabilizer (PSS). These controllers were designed for a single machine and/or multi‑machine power systems. This type of controller utilizes the speed, AC frequency, or accelerating power or a combination of these signals to derive a control signal injected at the input of the excitation and/or the governor/turbine systems. In this study controllers are designed not only for stability considerations but also to enhance the performance properties of the excitation and governor/turbine loops which were designed based on single‑input single‑output techniques. Moreover, the proposed controllers were designed using robust multivariable control theory via the linear quadratic Gaussian with loop transfer recovery (LQG/LTR) methodology. The issue of model reduction was also investigated in this study. The resulting controllers were implemented in the full nonlinear model of the power system and their performances were evaluated under different types of disturbances that are often encountered in an actual power system.