Fuzzy-Neural Networks Application In Speed Control Of AC Motor Vector-Control System

Fuzzy-neural network application in vector control system is studied in this paper. Intelligent control is getting a lot of attention in the area of electric drive system, as a new and important branch, fuzzy-neural network is used widely in all kinds of control system at the same time. Because of many good advantages, vector control is widely used for industrial application, but it has many problems, such as complex system structure, nonlinearity , parameter changes. Based on studying a great deal of papers and analyzing the application of intelligent control in modern electric drive area, fuzzy-neural application in speed control of AC motor vector control system is studied through simulation experiment in this paper, and some work has been done for the further experiment.In this paper, a simulation study is made for both the vector-convertion model of a squirrel-cage induction motor and vector control system which takes the model as its control object. In the simulation model, many controllers are used and deeply studied, they are classical PID controller, simple fuzzy control, fuzzy controller with integration factor and fuzzy + PI dual mode controller. In the end of this thesis, fuzzy-neural controller application is stressed. ANFIS(Adapt!ve-Network-based Fuzzy Inference System )is used in Fuzzy neural network. ANFIS that can produce fuzzy rules from sample data directly is organically combination of fuzzy inference system and neural network. Through sample data, it can optimize fuzzy rule. Using the optimized fuzzy rule, a lot of simulation experiments are done. System performance is studied, such as steady-state characteristics, dynamic characteristics, the effects of parameter variations and the effects of disturbance inputs. The necessary structure diagrams, algorithm, and results from the simulation have been shown in this paper. Simulation results demonstrates that the benefits of using fuzzy-neural network controller are restraining nonlinearity and parameter changes to vector control system, improving system robustness. So that fuzzy neural network has good prospect in vector control system.