Experiments on Impedance Control to Derive Adaptive Strategies

In this paper, we present some experimental results on impedance control applied to an industrial manip-ulator, while considering the constraints and perturbations that occur during a typical assembly task (unknown and varying environment parameters, impacts, friction, noises,...). An experimental and theorical analysis is used rst to derive a simpliied scheme from the original Hogan's controller. Simpliications concern the dynamic compensation stage, and the force feedback, which has been suppressed to deal with experimental conditions. To increase the performances of the resulting controller, a higher level is added to modify on-line the desired impedance and/or the reference trajectory. This supervisor has been developed using fuzzy logic. Experimental results on an industrial IBM Scara illustrate the ability of the system to absorb large external impacts and perturbations due to variations in the behavior of the manipulator and its environment.