Abstract: In the paper, an improved immune algorithm is used to generate the rule-base of fuzzy controller. Simulation results show that using immune algorithm to optimize fuzzy controller helps to solve premature problems and to quicken evolution course by referring to the existing experience.