Iterative Learning Control for a Soft Exoskeleton with Hip and Knee Joint Assistance.

Area of Science:

• Biomechanics
• Robotics
• Human-Computer Interaction

Background:

• Different terrains alter walking biomechanics, necessitating adaptive assistive devices.
• Existing exoskeletons often focus on single joints or lack terrain-specific adaptation.

Purpose of the Study:

• To design and evaluate a lightweight soft exoskeleton for simultaneous hip and knee assistance.
• To develop a novel control strategy for terrain-adaptive assistance.
• To quantify the metabolic benefits of the exoskeleton across different slopes.

Main Methods:

• A lightweight soft exoskeleton assisting hip flexion and knee extension was developed.
• A parameter optimal iterative learning control (POILC) strategy was implemented.
• Metabolic rates were measured in three subjects walking on downhill, flat, and uphill terrains with and without the exoskeleton.

Main Results:

• The exoskeleton reduced net metabolic rate by 9.86% (downhill), 12.48% (flat), and 22.08% (uphill).
• Assistance effectiveness increased with steeper terrain slopes.
• The POILC method mitigated wearer-specific variations.

Conclusions:

• The developed soft exoskeleton effectively reduces metabolic cost during walking on varied terrains.
• The terrain-adaptive control strategy enhances assistance performance.
• This technology shows promise for improving mobility and reducing energy expenditure for diverse users.