Developing a fuzzy logic controller in case of sparse testimonies (Q678772)

This study addresses an issue on the design of a fuzzy controller in the presence of sparse rulebases. Such collections of rules emerge in situations of a limited collection of ``if-then'' statements forming sparse testimonies in the condition space of the rules. They lead to challenging design problems that call for a conceptual augmentation of the original constructs of the fuzzy controllers in order to support reasoning for inputs where the original pieces of evidence (rules) are not available. The mechanisms employed in this paper exploit the concepts of linguistic extrapolation and interpolation. Detailed design algorithms are summarized.