This research focuses on an optimal control approach to simulate limb segment motions of a planar mechanical walking model. In this approach the optimal segment motions are a function of a performance index (such as mechanical energy) which is minimized and physically-based system constraints which must be satisfied. A Fourier-based approximation technique is applied to convert the optimal control problem into a nonlinear programming problem which can be solved using well-developed optimization algorithms. The set of nonlinear algebraic equations subject to constraints is solved for the suboptimal histories of joint angles, velocities, accelerations, and torques. By investigating different performance indices and comparing the resulting motion histories with human walking data, the approach can be used to study strategies that humans use in selecting dynamic patterns of limb motions during locomotion.
|Number of pages
|American Society of Mechanical Engineers, Applied Mechanics Division, AMD
|Published - 1 Jan 1987
ASJC Scopus subject areas
- Mechanical Engineering