Implementation of human-machine synchronization control for active rehabilitation using an inertia sensor

Area of Science:

• Neuro-rehabilitation
• Robotics in Medicine
• Biomechanics

Background:

• Active training is crucial for stroke patient recovery, especially for those with mild impairments.
• Rehabilitation devices lacking backdrivability require human-machine synchronization for effective active training.
• Upper limb exoskeleton rehabilitation devices (ULERD) offer portable and wearable solutions but often lack backdrivability.

Purpose of the Study:

• To propose and validate a method for achieving human-machine synchronization in non-backdrivable ULERD.
• To enable effective active training for stroke patients using an upper limb exoskeleton.
• To enhance the feasibility and effectiveness of robotic-assisted neuro-rehabilitation.

Main Methods:

• Utilized an inertia sensor to detect user forearm motion.
• Implemented an adaptive weighted average filtering for smooth velocity estimation.
• Developed a double closed-loop control strategy for accurate real-time tracking.

Main Results:

• Demonstrated effective detection of user forearm movement.
• Achieved smooth and accurate velocity tracking of the user's limb.
• Validated the feasibility and effectiveness of the proposed synchronization method through experiments.

Conclusions:

• The proposed method successfully establishes human-machine synchronization for non-backdrivable ULERD.
• This synchronization is a critical precondition for effective active training in neuro-rehabilitation.
• The developed system is effective and feasible for enhancing stroke rehabilitation outcomes.