Uncertainty solution of robot parameters using fuzzy position control applied for an automotive cracked exhaust system inspection

Abstract

This paper presents a new strategy for robotic manipulator inverse kinematics, which senses the final position error of the end-effector through an observational error model based on an Internet Protocol (IP) camera and calculates the joint variable errors through two fuzzy control techniques. Single Stage Fuzzy Controller (SSFC) and Three Stages Fuzzy Controller (TSFC) served as an inverse kinematic model to 3 degrees of freedom (DOF) robot manipulator. The inputs to the controller were the errors in the end-effector position Δx, Δy, Δz and the outputs of the controller were the controlled changes in the angular displacements of the joints. The fuzzy controller was utilized to compensate for geometric errors and non-geometric errors in the robot manipulator. Autodesk Inventor software was used for modeling the robot manipulator and make the transition to Simulink through the SimMechanics link for trajectory simulation. The simulation during the uncertainty of link length and end-effector load was implemented to show the proposed control model effectiveness. The gas leak test in an automotive exhaust system was applied to locate the crack position as an experimental validation of the proposed model in real-world applications.