Adaptive interaction torque-based AAN control for lower limb rehabilitation exoskeleton

Area of Science:

• Robotics
• Biomechanics
• Control Systems

Background:

• Lower limb rehabilitation exoskeletons require precise control to provide effective assistance.
• Existing methods may over-assist or under-assist, impacting patient recovery.
• Adaptive control strategies are crucial for personalized rehabilitation.

Purpose of the Study:

• To propose an adaptive interaction torque-based assist-as-needed (AITAAN) control method for lower limb rehabilitation exoskeletons.
• To accurately estimate wearer's muscle torque and provide only the necessary assistance.
• To achieve fast and accurate exoskeleton trajectory tracking for improved rehabilitation outcomes.

Main Methods:

• Designed a desired input torque using computed torque control (CTC).
• Employed a nonlinear disturbance observer (NDO) to estimate lower limb muscle torque.
• Utilized a spring-damper dynamics model to convert interaction torque tracking to trajectory tracking.
• Developed a flexible boundary prescribed performance controller (PPC) for exoskeleton trajectory control.
• Validated the controller through co-simulations of a coupled wearer-exoskeleton system in MATLAB/Simulink.

Main Results:

• The AITAAN controller effectively estimated wearer's muscle torque.
• The exoskeleton provided assistance only for the residual torque, adapting to the wearer's effort.
• Fast and accurate trajectory tracking of the exoskeleton was achieved.
• Co-simulations demonstrated the controller's effectiveness and superiority over other methods.

Conclusions:

• The proposed AITAAN control method offers a promising approach for personalized lower limb exoskeleton rehabilitation.
• Accurate muscle torque estimation and adaptive assistance enhance rehabilitation efficiency and safety.
• The controller's ability to achieve precise trajectory tracking is critical for effective therapeutic intervention.