Design optimization platform for assistive wearable devices applied to a knee damper exoskeleton.

Area of Science:

• Biomechanics and Robotics
• Assistive Device Design
• Computational Modeling

Background:

• Designing optimal assistive wearable devices is challenging due to complex, high-dimensional design spaces.
• Traditional methods like human-in-the-loop optimization struggle with increasing design parameters.
• Predictive simulation offers a powerful, data-independent approach to anticipate device effects.

Purpose of the Study:

• To introduce a design optimization platform leveraging predictive movement simulation for assistive wearable devices.
• To address the challenges of high-dimensional design spaces in wearable device optimization.
• To identify optimal parameters for a knee exoskeleton assisting level-ground and downhill gait.

Main Methods:

• A two-layered optimization framework: inner loop for predictive movement simulation, outer loop for parameter optimization.
• Utilized predictive simulation to model human movement and device interaction without experimental data.
• Applied the framework to design a knee exoskeleton with a damping mechanism.

Main Results:

• Achieved significant reductions in normalized knee load peak for level-ground and downhill walking.
• Demonstrated a decrease in the overall cost of transport.
• Identified optimal damping torques applied during the stance phase with distinct coefficients for different gaits.

Conclusions:

• The proposed optimization platform reliably and efficiently identifies optimal parameters for assistive wearable devices.
• The framework effectively handles high-dimensional design spaces, offering valuable insights for initial device design.
• Supports designers in efficiently determining optimal parameter sets for enhanced assistive technologies.