Neural Modelling and Fuzzy Control: Regulation of the Temperature of an Oven

摘要

The paper describes a method for the design of a generic fuzzy control system. The features of the approach proposed are essentially the following: a multilayer perceptron neural network is used for the process model identification phase, and the model thus obtained is used by the cell-to-cell algorithm for automatic synthesis of a fuzzy controller. The approach described is applied to the control of a physical system represented by a non-linear thermal process which is widely used in industrial environments. The validity of the method proposed, which requires no knowledge of the process to be controlled, will be clearly illustrated.