Control of human arm movement in machine-human cooperative welding process

• This work models and controls human welder׳s hand movement.
• The work is done in a virtualized welding system linking human and robot.
• Welder adjusts his hand movement per a visual instruction from control system.
• Welder׳s hand movement is measured by virtualized welding system’ sensors.
• The robot follows the hand movement and performs actual welding.

This paper studies accurate control of human arm movement in machine-human cooperative control of GTAW process. An innovative teleoperated virtualized welding platform is utilized to conduct dynamic experiments to correlate the human welder arm movement with the visual signal input. An adaptive ANFIS model is proposed to model the intrinsic nonlinear and time-varying characteristic of the human welder response. A model based predictive control algorithm is then proposed and an analytical solution is derived. Human control experimental results verify that the proposed controller is able to track varying set-points and is robust under measurement and input disturbances.