Adaptive multi-model controller using trust

During the past years, the applications of fuzzy logic have increased. Most of the fuzzy logic controllers (FLC) are based on linguistic rules which represent the control strategy of an expert. A new approach is the multi-model controller (MMC). Its concept is that, when one attempts to represent a complex system, it will be easier to consider a set of local models than a single one that should be valid for every operating point of the state space. Previously, the authors have developed a new scheme of MMC, using possibilities, and an adaptive mechanism that improves the performance of this kind of controllers.