Optimization-Based Motor Control of a Paralympic Wheelchair Athlete

The following research approximated how the central nervous system of Paralympic wheelchair athletes resolve kinematic redundancies during upper-limb movements. A multibody biomechanical model of a tetraplegic Paralympic athlete was developed using subject-specific body segment parameters. The angular joint kinematics throughout a specified Paralympic sport movement (i.e., wheelchair curling) were experimentally measured using inertial measurement units. The motor control system of the Paralympian was mathematically modelled and simulated using forward dynamics optimization. The predicted kinematics from different optimization objective functions (i.e., minimizing resultant joint moments, mechanical joint power, and angular joint velocities and accelerations) were compared with those experimentally measured throughout the wheelchair curling movement. Of the optimization objective functions under consideration, minimizing angular joint accelerations produced the most accurate predictions of the kinematic trajectories (i.e., characterized with the lowest overall root mean square deviations) and the shortest optimization computation time. The implications of these control findings are discussed with regards to optimal wheelchair design through predictive dynamic simulations.