Physiologically based control laws featuring antagonistic muscle co-activation for stable compliant joint drives

The development of robot joints ensuring controllable compliant behaviors during interaction tasks has gained the interest of the robot community in the recent years. In biological systems, while intrinsic muscle nonlinear elasticity allows the regulation of joint compliance through antagonistic co-activation, physiological control mechanisms ensure robust stability during interaction tasks both in static cases and during motion. This work presents a novel control approach for the simultaneous regulation of position and stiffness in a hinge joint driven by two muscles which combines biological findings like co-activation and reproduces reflex actions and additional influences to ensure stability. Based on the analysis of the stiffness generated in the antagonistic system and on the evaluation of stability issues when the musculoskeletal setup is loaded by external forces, a stiffness controller is proposed which integrates a stiffness computation block and an adaptive mechanism for the control of stability during interactions. The position controller, modeling physiological properties and motor-neurons, relies on reciprocal activation while the stiffness controller implements a co-activation strategy. The controller is tested in a numerical simulation using Matlab/Simulink (R) for different task conditions. Simulation results demonstrate the ability of the controller to simultaneously regulate position and adapt joint compliance to different external perturbations.