Improved H2/H∞ control based on multi-objective genetic algorithm

A novel approach to H-2/H-infinity optimal control is presented based on multi-objective genetic algorithm (MOGA). To design H-2/H-infinity controller with less conservativeness, a kind of MOGA for H-2/H-infinity control (HHMOGA) is especially developed. HHMOGA takes the solutions of linear matrix inequality (LMI) method as initial population. Non-dominated sorting, niche, and elitist strategy are employed in order to ensure a better design. Simulation results show that HHMOGA can achieve better performances as compared with LMI method.