Adaptive Fuzzy Logic Controllers with Sensitivity Learning Rules
Thananchai Leephakpreeda · 1999
This paper presents a design sensitivity method for leaming rules of adaptive Fuzzy Logic Controllers (FLCs) ' FLCs can be made to adapt their initial knowlJdge bases to ihe changes in the operating conditions. With a model reference control strategy, the performance index is defined as the error function between the desired and actual outputs of the proiess. It is to be minimized by thegradient-based optimization which acquires a knowledge of the sensitivity of the performance index with respect o the parameters of FLCs. The tracking problem of a continuous stirred tank reactor isinvestigated as a realistic non-linear case study for the-viability of the proposed methodology. Up to now, FLCs have been widely used in many applications of process control. Compared to conventional control stratesies. FLCs can yield superior controller perlormince