We propose in this talk to introduce a new paradigm for kinematic control of highly redundant robots that provides sparse control vector in the joint space (joint velocity). Indeed, when a robot is highly redundant in comparison to the task to be performed, current control techniques are not “economic” in the sense that they demand, most of the time unnecessarily, all the joints to move. Such a behaviour can be undesirable for some applications. In this direction, this work proposes a new control paradigm based on linear programming (LP) that intrinsically provides a parsimonious control strategy, that is, one in which few joints move. We will present the method as well as simulation and experimental results on the HOAP-3 humanoid robot. Finally, a comparison with the Pseudo-Inverse will be detailed.
Philippe Fraisse is currently professor at the Université de Montpellier (UM-LIRMM), France. He received the Master of Electrical Engineering of Ecole Normale Supérieure de Cachan in 1988. He received Ph.D. degree in Automatic Control of Université de Montpellier on two-arm robotics, France, in 1994. His research interests are humanoid robotic, human motion, human-robot interaction and assistive robotics.