Robot Movement Planning and Control Based on Equilibrium Point Hypothesis

摘要

Despite those the inverse dynamics methods in traditional robotics, few are eligible to be applied to systems with as high degrees of freedom (DOFs) as humans. Few tackle the intricacies of the human musculoskeletal system itself. We propose a two-phase motor control model based on the equilibrium point hypothesis, which takes advantage of the muscle spring system to control human movements. The motor planning algorithm calculates the solution in the joint space, given a simple or complex task in Cartesian space. Then the spring model simulating the muscles takes charge of the movement execution. This model greatly reduces the amount of computation, compared to the inverse dynamics. We demonstrate the model in various motions as reaching, walking, sitting and rising