Multi Agent System Inspired Distributed Control of a Manipulator

Abstract

Robotic manipulators are used in a wide variety of applications. In all the newlineapplications, the end-effector or the tool of the manipulator needs to be moved newlinealong a desired trajectory in its workspace. In this thesis we present model-based newlinecontrol schemes for robotic manipulators using a distributed architecture. newlineInspired by multi-agent/robotic systems, first we perceive a manipulator, newlinewhich is MIMO multi-body system, as a multi-agent system with the joints (or newlinethe joint-link pairs) as sub-systems or agents, which interact with each other in newlinea distributed manner. Here, the interaction between the joint-link agents is in newlinethe form of interactive forces and moments that lead to dynamic coupling. As newlinethe adjacency graph formed by the joint-link agents as nodes and links between newlinetwo joints as edges is connected, the direct interactions between the immediate newlineneighbors result in interaction (in the form of dynamic coupling) between any two newlinejoint-link agents. newlineWe carry out an analysis of the computational cost associated with the newlinemodel-based control law for planar serial-link manipulators with newlinedegrees-of-freedom varying from 2 to 6 using Maple. Using this analysis, we newlineestablish the fact that the total computational cost associated with the newlinemodel-based control law increases with the degrees-of-freedom. Toward newlinemitigating the computational overhead associated with the conventional newlinemodel-based control scheme, we propose a distributed architecture for the newlinemotion control of manipulator exploiting its multi-agent nature. Here, each newlinejoint-link agent is controlled by a dedicated controller, and the joint-level newlinecontrollers communicate and cooperate among themselves. Though one of the newlineprimary motivation for the proposed distributed control scheme is to reduce the newlinecomputational overhead, in this thesis we rely on the natural distributed nature newlineof the manipulator dynamics rather than the program optimization or operation newlineoptimization techniques that are used at the algorithmic level. newline