ADAPTIVE POWER SYSTEM STABILIZER BASED ON LINEAR QUADRATIC CONCEPT.

An adaptive power system stabilizer is described in this paper. The stabilizer algorithm consists of an explicit identifier using recursive least squares identification technique. Control is computed using the linear-quadratic control algorithm. Deterioration in the quality of identification immediately following a large disturbance and the possibility of loss of stability is countered by retaining the prefault parameters for a short period after the inception of a large disturbance. The parameter update is reinitiated after this period. The proposed technique was first tested by simulation. It was then implemented on a microcomputer. Simulation and real-time experimental results show that the adaptive stabilizer is very effective in damping out synchronous machine oscillations following system disturbances.