Inconsistency detection - a powerful means for the design of MIMO fuzzy controllers

When designing a multiple input multiple output (MIMO) fuzzy controller, it is difficult to validate the correct interaction of the fuzzy IF-THEN rules. This paper deals with inconsistency detection as a powerful means for the design of MIMO fuzzy controllers. Known and novel definitions of the term inconsistency are given. Two main types of inconsistencies are distinguished, namely the structural and the effective inconsistencies. Connections between the two types are illustrated. As the foundation of a new method to detect and remove inconsistencies, an inconsistency measure is derived. This inconsistency measure enables the detection of effective inconsistencies. An example is used to show how this new method can be applied to generate consistent knowledge bases systematically