Fault Detection, Isolation and Reconfiguration of Four-Bar Mechanism-Based Knee Exoskeleton.

Area of Science:

• Robotics
• Biomechanics
• Control Systems Engineering

Background:

• Knee exoskeletons enhance mobility but face reliability challenges.
• Actuator faults can compromise exoskeleton safety and continuous operation.
• Existing systems may lack robust fault management for critical components.

Purpose of the Study:

• To develop and validate a fault-tolerant control system for knee exoskeletons.
• To enhance the reliability and safety of knee exoskeleton devices.
• To ensure uninterrupted exoskeleton functionality despite actuator failures.

Main Methods:

• Implemented a fault detection, isolation, and reconfiguration (FDI) technique.
• Derived analytical redundancy relations (ARRs) from a bond graph model for fault detection.
• Simulated actuator fault scenarios and reconfiguration during sit-to-stand motion.

Main Results:

• The proposed FDI system effectively detected and isolated actuator faults.
• The system successfully reconfigured the faulty actuator, maintaining exoskeleton function.
• Simulations demonstrated improved robustness of the knee exoskeleton during sit-to-stand tasks.

Conclusions:

• The developed fault-tolerant control system significantly enhances knee exoskeleton reliability.
• The FDI technique ensures continuous operation and user safety in the presence of actuator faults.
• This approach offers a robust solution for dependable exoskeleton performance in rehabilitation and assistance.