Nested design of fuzzy controllers with partial fuzzy rule base (Q5937154)

Experts systems are used to design controllers of complex systems with unknown dynamics, and the linguistic information so involved can be described by the fuzzy if-then rules with linguistic predicates. Unfortunately the fuzzy controllers generally so obtained, by using redundant rules, are very complex and inefficient. NEWLINENEWLINENEWLINEIn this paper, a nested design approach is used to reduce this complexity by using partial fuzzy if-then rule base. This base is complete and the input space is not fully partitionned. Loosely speaking, when the performance requirements of the fuzzy controllers are not achieved, the fuzzy rule base is modified on the nested design approach to replace the unsatisfactory fuzzy rules by additional efficient fuzzy if-then rules.