Related Experiment Videos
A parameter optimization approach for the optimal control of large-scale musculoskeletal systems
M G Pandy1, F C Anderson, D G Hull
1Department of Kinesiology and Health Education, University of Texas, Austin 78712.
Journal of Biomechanical Engineering
|November 1, 1992
Summary
This study presents a novel computational method for solving complex optimal control problems. By converting these problems into parameter optimization, it simplifies calculations for nonlinear dynamical systems and human movement analysis.
Area of Science:
- Biomechanics
- Computational Science
- Robotics
Background:
- Optimal control problems in large-scale, nonlinear dynamical systems are computationally challenging.
- Traditional methods involving two-point boundary value problems can suffer from convergence issues due to nonlinearities.
Purpose of the Study:
- To introduce a computational method that transforms optimal control problems into standard nonlinear programming problems.
- To demonstrate the method's efficacy in solving human movement optimal control problems, specifically maximum-height jumping.
Main Methods:
- Parameterizing control histories using nodal points to convert optimal control problems into parameter optimization problems.
- Utilizing established nonlinear programming algorithms to compute near-optimal control trajectories.
- Applying the method to an optimal control model of maximum-height human jumping.
Main Results:
- The parameter optimization approach successfully computed near-optimal control trajectories, bypassing the need to solve complex two-point boundary value problems.
- The model's predictions for maximum-height jumping closely matched experimental data for various biomechanical parameters.
- Key features of human jumping, including jump height and ground contact time, were accurately reproduced.
Conclusions:
- The parameter optimization method offers a robust and efficient alternative for solving optimal control problems in complex dynamical systems.
- This approach is particularly valuable for biomechanical analyses of human movement, providing accurate and experimentally validated results.