Optimal frequency regulation of a two-area power system

Load frequency controllers are required to operate effectively under increasingly restrictive conditions. These include highly varying random fluctuations in power demand which changes the operating point of the power system, in turn resulting in parametric changes and uncertainties in the power system. The frequency and tie-line power deviations need to be regulated in minimum time within narrow limits taking into account the associated constraints on the actuators. In this paper, we consider the application of an Linear Quadratic Gaussian (LQG) controller to a two-area system which suffers from big changes in the parameters of the plant. Using appropriate model reduction techniques, the performance of a reduced-order controller is evaluated under sudden load changes represented as constant disturbances and under parametric changes. We show that the LQG controller performs significantly better and provides greater margin of stability than a pair of optimized PID controller, designed for each area separately.