Model-Based Self-Abstracting Rule Fuzzy Control

Aiming at increasing the performance of fuzzy control and keeping a fuzzy controller maintain an ideal given track to its objective value, and also aiming at finding an easy way to abstract fuzzy control rule set for the complex controlled process, this paper presents a new fuzzy control method, namely model based self-abstracting rule fuzzy control method (MBSAR). The MBSAR takes the one order inertia system as the reference model and abstracts the fuzzy control rules by self-organizing and self-tuning the consequences of each functioning fuzzy control rule both offline and online. A self-tuning rule set of the self-abstracting rules has been built by analyzing the dynamical response process of a controlled object. The simulations show that the MBSAR can decrease the overshoot significantly and has a wonderful control performance, which are better than that of the self-abstracting rule fuzzy control method without model and PI control method.