Nonlinear predictive control based on the extraction of step-response models from Takagi-Sugeno fuzzy systems

This paper deals with nonlinear predictive control based on higher order Takagi-Sugeno fuzzy systems which can also be interpreted as generalized radial basis function networks. We investigate how the fuzzy models can be linked to a special type of model based predictive control algorithm, namely the dynamic matrix control (DMC). Previously, purely linear step response models were used for long-range prediction. Here, the method is extended to nonlinear processes. Therefore, various step responses for different operating points are extracted from the fuzzy model. For performance evaluation, a heat exchanger is identified by means of the local linear model tree algorithm and controlled by the modified DMC.