Automated fuzzy knowledge base generation and tuning

The authors present an approach to generating and tuning a knowledge base for fuzzy logic control (FLC) of an inverted pendulum. They used a modified self-organizing control procedure under typical FLC design choices with a very crude plant model to quickly converge on a rule base appropriate for the plant. A FLC using the derived rule base showed smaller percent overshoot and shorter settling time than a simple modern controller. The knowledge base was tuned by dynamically changing the controller gain according to a thresholding parameter. The best threshold/gain value was obtained by a gradient search algorithm driven by a step-response performance cost function. The same FLC using the tuned scaling factors exhibited critically damped step response.<
>