Fuzzy logic based formulation of the organizer of intelligent robotic systems

Abstract

A fuzzy logic based methodology is proposed to model the organization level of an intelligent robotic system. The user input commands to the system organizer are linguistic in nature, and the primitive events-tasks from the task domain of the system are in general interpreted via fuzzy sets. Fuzzy relations are introduced to connect every event with a specific user input command. Approximate reasoning is accomplished via a modifier and the compositional rule of inference, whereas the application of the conjunction rule generates those fuzzy sets with elements all possible (crisp) plans. The most possible plan among all those generated that is optimal under an application dependent criterion is chosen and communicated to the coordination level.<>