Development of a Neural-Fuzzy-Based Variable Admittance Control Strategy for an Upper Limb Rehabilitation Exoskeleton

Area of Science:

• Robotics
• Neurorehabilitation
• Biomedical Engineering

Background:

• Stroke-induced upper limb motor dysfunction necessitates advanced rehabilitation.
• Current exoskeletons lack adaptive control, multi-limb training, and personalized adjustments.

Purpose of the Study:

• To develop a novel three-degree-of-freedom upper limb exoskeleton.
• To enhance robot-assisted stroke rehabilitation through adaptive control and bilateral training.

Main Methods:

• Implemented a neuro-fuzzy adaptive admittance control architecture with dual inputs.
• Utilized a Brunnstrom stage-specific fuzzy rule base for adaptive parameter adjustment.
• Developed a bilateral adaptable mechanical structure for dual-limb training.

Main Results:

• Achieved a maximum trajectory tracking error of less than 1.2° and RMS error of ≤0.13°.
• Demonstrated an RMS error of 2.99 mm for circular trajectory tracing.
• Validated effective balance between tracking accuracy and human-machine compliance.

Conclusions:

• The developed exoskeleton offers a promising solution for upper limb stroke rehabilitation.
• The adaptive control strategy enhances personalized and effective robotic assistance.
• Bilateral training capability expands the exoskeleton's clinical applicability.