Model Free Human‐Robot Interaction Control

This chapter addresses different model‐free controllers in both task space and joint space. Proportional‐Derivative (PD) control is one of the simplest control techniques for admittance control schemes. PD and Proportional‐integral‐derivative control improve the robustness of the robot‐interaction behaviors. Both controllers guarantee convergence to a bounded zone. The chapter considers task space dynamics for the controllers with the virtual work principle. It discusses experiments of the model‐based and model‐free admittance controllers on 2‐degrees of freedom (DOF) o pan and tilt robot and 4‐DOF exoskeleton robot prototypes with a force/torque sensor mounted at its end effector. The model‐free controllers have good tracking results without the robot dynamics. The accuracy and robustness dilemma of impedance control is solved without the robot dynamics. Convergence and stability of the model‐free controllers is presented using Lyapunov stability theory.