Adaptive Fuzzy Neural Network Control Based on Genetic Algorithm

The conventional PID controller parameter tuning method needs the precise mathematical model of the controlled object, and the tuning parameters cannot be adjusted online. Then, in view of the fuzziness of control system and its elements and the complexity of their relationship, this paper combines the characteristics of good expression of fuzzy mathematics knowledge with the strong learning ability of artificial neural network model, and studies the application of fuzzy neural network to realize the control of nonlinear system. Genetic algorithm is used to optimize the parameters of membership function and the weight of neural network, so that the controller has adaptive learning ability, which can avoid the shortcomings of the original BP algorithm easily falling into the local optimal value. Finally, the control of a single inverted pendulum is simulated. The experimental results and data analysis show that it has obvious advantages in control time, stability and anti-interference, and the control effect is better than the traditional fuzzy controller and PID control method.