A new approach for multi-objective-optimization-based fuzzy-PID control

A parameter self-tuning PID controller based on genetic optimization is proposed in this paper. A mathematical model of the fuzzy-PID controller, which parameters are tunable and can be updated timely, is set up. Membership function, fuzzy control rules and parameters of the PID controller are optimized by improved GA. And self-adaptive crossover and mutation operators are used to improve the performance of global searching and converging. Therefore, the proposed controller avoids disadvantages of conventional fuzzy-PID controller with invariable inference rules, and has higher accuracy. Experiments indicate that the proposed controller has better performance than that the conventional one, and can meet with requirements of the given servo system. © 2012 Springer Science+Business Media, LLC.