ROBUST DECENTRALIZED CONTROL DESIGN USING GENETIC ALGORITHMS IN POWER-SYSTEM DAMPING CONTROL

The paper presents the design of linear robust decentralised fixed-str ucture power system damping controllers using genetic algorithms (GA). The designed controllers follow a classical structure consisting of a gain, a wash-out stage and two lead-lag stages. To each controller is associated a set of three parameters representing the controller gain and the controller phase characteristics. The GA searches for an opti mum solution over the parameter space. Controller robustness is taken into account as the design procedure considers a prespecified set of o perating conditions to be either stabilised or improved in the sense o f damping ratio enhancement. A truly decentralised control design is a chieved as the loop control channels are closed simultaneously. The ap proach is used to design SVC and TCSC damping controllers to enhance t he damping of the interarea modes in a three-area six-machine system. Local voltage and current measurements are used to synthesise remote f eedback signals.