A robust power system control design combining system identification and genetic algorithms

Presents a hybrid control design method that combines features from a system identification problem with meta-heuristics of genetic algorithms (GA). The main objective of the method is to find robust controllers that make the closed-loop system less sensitive to parametric uncertainties and also force the closed-loop spectrum to be located in a subregion that represents minimum damping. Minimum-phase low-order decentralized dynamic-output-feedback controller structure is readily accomplished in the proposed method. The approach is devised to be applied in the design of power system damping controllers. However, the mathematical treatment employed is general to other engineering applications.