Dual Control for Jerk-Driven Robotics in Rehabilitative Planar Applications

Abstract

This study compares a set of strategies to plan and control the trajectory of a robotic device in a planar workspace. These strategies are based on an affective application of jerk-laws able to indicate undesirable conditions (e.g., vibrations) facilitating the device control. The jerk is the time derivative of acceleration, and this solution provides an indirect means to control the variation rate of the actuator torques, while avoiding the complex robot dynamic models and their algorithms for computing the dynamics. In order to obtain a smooth trajectory, a regulator to control a robotic device has been developed and validated. It consists of the implementation of two control modules able to (i) generate the predefined trajectory and (ii) guarantee the path tracking, reducing unwanted effects. In this case a simple S-shaped path has been originated by the “trajectory generator module” as a reference movement to rehabilitate upper limb functionality. The numerical simulation and the results of preliminary tests show the efficacy of the proposed approach through the vibration smoothness appraisal associated with the motion profile.