Skip to Main Content
In this paper an indirect adaptive controller for single-input single-output (SISO-) systems is presented. The design procedure uses the mathematical input-output model of the plant to find a controller output which minimizes a cost function for a given prediction horizon. Therefore this predictive controller is able to work both with linear and nonlinear plants (i.e., using linear and nonlinear models of the plant); it only needs the input-output description of the plant to be controlled. In this paper, for example, results with linear ARMAX-models, as well as with a neural network description of a nonlinear plant are presented. The controller output is found using an evident search strategy which avoids computation of partial derivatives.