Stability analysis and controller design for dynamic fuzzy systems based on a new fuzzy inference approach

Qualitative modeling of technical processes may be accomplished by dynamic fuzzy systems. A new inference method with interpolating rules is proposed as an essential basis for the analysis of this class of systems. Using this approach, the system output is dependent on both an interpolating rule derived from the fuzzy input and the fuzzy input itself. A simple example shows the typical behavior of such dynamic fuzzy systems and leads to a stability definition. Furthermore, it is shown how a fuzzy controller is designed using a set of fuzzy rules describing the process behavior. The design method is applied to a continuous stirred tank reactor.