Performance estimation techniques for power system dynamic stability using least squares, Kalman filtering and genetic algorithms

Abstract

This paper presents performance comparison of three optimal estimation techniques for on-line assessment of power system dynamic stability of single-machine infinite-bus system. The stability assessment approach is based on estimating the synchronizing and damping torque coefficients of the synchronous machine using three optimum estimation techniques including least squares (LS), Kalman filtering (KF) and genetic algorithms (GA). The coefficients are estimated from time responses of the changes in the rotor angle, rotor speed, and electromagnetic torque. The performances of the above three optimal estimation techniques were examined. Compared with the LS and GA techniques, the paper shows that KF technique offers several advantages. This includes significant reduction in computing time and storage needed for the estimation of the synchronizing and damping torque coefficients besides its robustness in dealing with noisy measurements. Thus, KF approach results in a remarkable reduction in the computational complexity associated with this problem and hence allow for on-line implementation needed for continuous monitoring of the dynamic stability indices. On the other hand, though GA gives accurate results in comparison with LS and KF. However, it was found that the calculation by GA are very time consuming rendering it unsuitable for on-line application.