An optimal control model for assessing human agility trajectories

Agility is typically defined as the ability to rapidly change velocity or direction. However, measurement of agility is often experimentally limited to completion time in a planned agility course, which does not reveal the underlying biomechanics contributing to performance. Additional contributing factors to interpreting agility include understanding the trajectory of the path and the technique used to achieve that path. In selecting a motor strategy, previous research has shown that human motion planning can be a function of kinematic, dynamic, and time criteria. It is unclear how these criteria may affect the trajectory in a planned agility course. In this paper, an agility task is formulated as an optimal control problem and the relationship between estimated path trajectories and the selected objective function is investigated. Here we specifically consider the criterion of minimizing the magnitude of the squared jerk and minimizing final time, with constraints on speed, acceleration, and maximum ground reaction force that can be produced while running without slipping. Since this frictional constraint takes gravity into account, the trajectories are examined for Earth, as well as reduced gravity environments such as the Moon and Mars. The computed optimal trajectories for the agility task are compared to previously collected experimental data. By comparing the experimental and optimal trajectories, insight is gained on participant strategy. Extending to reduced gravity conditions provides quantitative insights on limitations for astronaut locomotion.