A Multi-Agent Approach Based on Fuzzy Logic For a Robot Manipulator.

摘要

The use of the numeric variables for the modeling of evolution space of the robot manipulators requires the use of very specific and very effective tools. The purpose of this paper is to propose a method of modeling and control of a manipulator arm using fuzzy logic by exploiting the intrinsic properties of the agent paradigm. This method is directed so as to regulate the problem of difficulty and sometimes the lack of information through the use of fuzzy variables and this for the representation of the robot’s environment. The suggested method can help the robot to join a well defined goal. We also propose a simple fuzzy controller and a set of geometric models to follow up the best paths for a robot manipulator to three degrees of freedom. The simulation results show the effectiveness of the approach.