Cooperative Multi-agent for The End-Effector Position of Robotic Arm Based on Consensus and PID Controller

Abstract

In a multi-agent case, we consider not only an agent but also several agents. The problem of multi-agent is that each of the agents has a randomly different initial state. The existence of multi-agent in a system needs to be managed in order to be a structured system for making certain decision cooperatively. This paper presents the implementation of cooperative multi-agent for synchronizing the end-effector position of the robotic arm based on consensus and PID controller. Physically, the robotic arm has three servo motors that mean the robotic arm has three joints. Thus, to synchronize the end-effector position of the robotic arm among agents, all of the servo motors in all agents must be controlled in order to achieve the same point of view. In addition, this paper expounds how to derive the consensus equations and PID controller based on designed graph topology. The presence of two proposed methods is intended to compare whether or not the problem of multi-agent regarding randomly different initial state can be overcome by using consensus and PID controller. From the experimental results, the derived consensus equations and PID controller can be implemented to synchronize all of the joints possessed by the robotic arm. As a result, the end-effector of the robotic arm in all of the agents successfully pointed to the same position in three-dimensional space. It indicates that the derived consensus equations and PID controller can handle the different state of each joint in all of the agents.