EXTRACTING CORE INFORMATION FROM INCONSISTENT FUZZY CONTROL RULES

Abstract

In many cases, inconsistent fuzzy control rules can be obtained when the information about the controlled plant is uncertain and/or the system is very complex. In most cases, inconsistent rules have been considered as ill-defined rules and, thus, not allowed to be in the same rule base. One way to handle the situation is to try to extract core information from the whole set containing the inconsistent rules. This approach can be useful especially when it is difficult to separate consistent rules from the set of inconsistent rules. In the paper, an inference scheme is presented to extract core information from such a rule base with inconsistent rules. The proposed inference scheme with the aid of a statistical concept justifies application of all the rules with inconsistent rules in the same rule base for effective control of a complicated plant. A simulation study is performed for a truck-and-trailer backer-upper control. The result shows that inconsistent rules can be effectively utilized if a proper inference method is adopted for the extracted rules.