Optimal Design Of Fuzzy Logic Controller Based On Adaptive Genetic Algorithm

The control effect of Fuzzy Controllers depends on two key factors: membership functions and control rules table. About this two choices, traditional methods rely mainly on people's experience, so there is a large arbitrariness and subjectivity,so it makes the control accuracy is not high. To overcome these problems, this paper use adaptive genetic algorithm to optimize the design of the fuzzy controller.Adaptive property design of genetic algorithm is improved through changed the genetic manipulation. It consists of two parts: adaptive mutation rate, adaptive crossover design and the uniformity design of selection. So the fitness of the individual can automatically adjust the size of the crossover and mutation operator, and make the rate convergence of genetic algorithm bigger, and not easy to fall into premature. In the algorithm is running, this paper implement the genetic manipulation parallel to improve the uniformity of selection.In the use of adaptive genetic algorithm to optimize the process of fuzzy controller, this paper, hybrid coding method that uses real number encoding method to encode membership functions, using the decimal coding method to encode fuzzy control rules. The article exploit the traditional second-order system as controlled object, simulation the model and verification the method, and the results show that the method for optimizing fuzzy controller is feasible and can improve the control effect.Using this method, for different objects can design a different fuzzy controller. Paper list a non-linear second-order system, time-varying second-order system as the controlled object of the simulation, the simulation results show that the fuzzy controller based on adaptive genetic algorithm has better static and dynamic performance.At the end of the paper, the fuzzy controller based on adaptive genetic algorithm applied to the three-phase induction motor slip frequency vector speed regulation system, according to observe the response curve , and compared with PI control system, the speed and stability of the controller have been improved, and have better control effect.