Model reference adaptive systems with fuzzy logic controllers

In this paper methods for adapting conventional controllers are used to adapt a fuzzy logic controller (FLC). The FLC and the plant should follow a desired trajectory which is given by a reference model. The parameters of the FLC are adapted by using a method known as model reference. Because FLCs are in general used for analytically unknown or little known plants the model reference approach must be modified. The adjustment algorithm is divided into two parts: 1) A controller output to yield model following is computed using an inverse plant model. If no analytical description of the plant exists a newly defined gradient model will be used which needs only little information about the plant. 2) The desired output of the FLC is obtained by using a gradient approach to adapt the parameters of the FLC. Simulation studies show the effectiveness of the this algorithm. Furthermore first experimental results for a hydraulic cylinder are presented.<<ETX>>