Aiming at automatic generation control (AGC) design of a class of interconnected power systems when the expansion is done on line

an optimized selfhealing control method

which can guarantee the expanded system to steady run when the topological structure is changed suddenly

is proposed by adopting dynamic output feedback and genetic algorithm. By means of modeling to the expansion system

the robust decentralized controller keeping the system structure to be stable is designed by using LMI method. Genetic algorithm is used to optimize the related parameters emerged in the process to solve

and the blindness to choose the parameters is avoided. In the optimal process

the fitness function evaluating synthetically the time domain response performance is chosen to optimize the interconnected constraint matrix in the design. The performance indexes of the system are improved. The simulation research is done against a kind of interconnected power system model. Consider a new area subsystem is add to original system on line

the automatic generation control law of the new area is designed by using the proposed method. The simulation results show that the proposed method reaches the selfhealing control request

which the structure of the expansion system is stable. The control performances optimized by genetic algorithm are better than the before

and the effectiveness of this method is proved.