A direct collocation framework for optimal control simulation of pedaling using OpenSim.

Area of Science:

• Biomechanics
• Computational modeling
• Human movement analysis

Background:

• Direct collocation (DC) offers low computational cost for movement optimization.
• DC has been underutilized for optimal control of human pedaling.
• OpenSim is a widely used environment for biomechanical modeling.

Purpose of the Study:

• Develop a direct collocation (DC) framework for optimal control simulation of human pedaling.
• Integrate the DC method within the OpenSim modeling environment.
• Assess the framework's efficiency and accuracy in simulating pedaling biomechanics.

Main Methods:

• Developed a planar bicycle-rider model in OpenSim.
• Formulated the DC method in MATLAB for optimal control using data tracking.
• Solved the optimal control pedaling problem with varying parameters and initial conditions.

Main Results:

• Successfully solved the optimal control pedaling problem within practical timeframes (24 minutes to 10 hours).
• Achieved accurate tracking of experimental data, with simulated pedal angles and forces within ±1 standard deviation.
• Observed muscle tendon unit (MTU) excitation patterns generally matched experimental EMG data in timing and shape.
• Demonstrated robustness to node density (above 31) and initial guess selection.

Conclusions:

• The proposed DC framework provides a computationally efficient and flexible tool for human pedaling simulation.
• The framework facilitates the application of optimal control methods to study pedaling biomechanics, energetics, and control.
• This approach can be readily adapted for predictive simulations, such as optimizing for pedaling rate.