Fuzzy controller is an important part of the fuzzy control system. To improve the parameters of the fuzzy controller effectively, this paper uses DC motor as the controlled object. Use three improved intelligent optimization algorithm respectively,named the genetic algorithm, the particle swarm optimization algorithm and the neural network to achieve the parameters of self-optimizing of the fuzzy controller.Firstly, this paper designs an improved genetic algorithm based on the research of genetic algorithm: Optimize the initial population and use the real-coded way, introduce the concept of similar to avoid the algorithm trapped in local optimal solution, adjust the crossover and mutation operator accordingly, and then use the improved genetic algorithm to achieve the parameters of self-optimizing of the fuzzy controller. By MATLAB simulation,compared with traditional genetic algorithm, we verify that, the fuzzy controller based on this algorithm, the convergence speed is faster and the effect is better.Then, this paer designs an improved particle swarm optimization algorithm by studying swarm optimization algorithm: Optimize the initial population and adjust the learning factors by the way of dynamic, introduce genetic operations to improve the diversity of particles and convergence speed, and use the improved particle swarm optimization algorithm to achieve the parameters of self-optimizing of the fuzzy controller. By MATLAB simulation, compared with the standard particle swarm algorithm, we verify that, the fuzzy controller based on this algorithm has the better control effect.Finally, this paper designs an improved fuzzy neural network controller based on the new BP algorithm: In order to obtain good initial parameter values, this paper uses the uniform random number and normalized processing method for parameter initialization. In order to improve the convergence of the algorithm, this paper uses adaptive adjusting learning rate to improve the fuzzy neural network controller, so as to optimize membership functions and fuzzy rules. And the proof process of the stability of the system are given based on the lyapunov theorem. By MATLAB simulation, we verify that, the fuzzy controller based on this algorithm has the satisfactory control effect. Finally, use the dc motor as a controlled object,compared the simulation of the three different methods to optimize the fuzzy controller, the simulation results show that the new type of fuzzy neural network controller to control dc motor speed control system has the most ideal control effect of the three, and the system has good control performance. |