The Design And Application Of Neural Networks Self-organizing Fuzzy Controller

It designed a neural networks self-organizing fuzzy controller, and used it in control the electro-hydraulic position servo system, for resolving the system's uncertainly, complexity and nonlinearity, an effective approach was provided.Firstly, the mathematical model of electro-hydraulic position servo system is built up, as the design basis of the controller, its non-linearity and time-variation is analyzed. Based on the control principle of the neural network and the fuzzy control, the mechanism that they combine is analyzed in detail, all feasible plans of neural network and fuzzy control are discussed and many schemes are presented. Secondly, the neural network self-organizing fuzzy controller was proposed, it adopted the method of estimating the network expectation output, and put forward a brand-new arithmetic that estimated previous moment control output offline, accordingly, the training sample of the neural network had the integrated form. Lastly, it proved through the computer emulation experiment, that for the control of electro-hydraulic position servo system, the position response curve which using neural network self-organizing fuzzy controller has less overshoots, less oscillation amplitude, and better dynamic performance than groovy fuzzy controller, and has stronger robustness.