Analogous fuzzy rule-based expert systems

摘要

In this paper, the concepts of primary fuzzy term sets and basis fuzzy sets are introduced. In order to remedy the problem of context in fuzzy reasoning environments, it is shown that a base logic, such as the main fuzzy logic paradigm proposed by Zadeh, could be employed in conjunction with three proposed primary fuzzy term sets, to model new logical operations and primitives. The proposed flexible structure introduced for fuzzy reasoning paradigms, called "soft fuzzy reasoning paradigm", provides the methodology for the implementation of all the known parameterized fuzzy reasoning paradigms. Canonical and base canonical fuzzy rule-based expert systems are introduced and employed to address the definition of analogous fuzzy systems. Three types of analogous fuzzy systems are discussed. As a problem case, two physically different systems such as a flexible robot and a heating slab which fall under type one of analogous systems, are discussed. Conditions for the transfer of knowledge across two analogous fuzzy systems are mentioned.