A generalized framework to achieve coordinated admittance control for multi-joint lower limb robotic exoskeleton

Area of Science:

• Robotics
• Biomechanics
• Control Systems

Background:

• Traditional N-degree-of-freedom (DOF) robotic joint space admittance controllers present complexity and can cause disjointed joint movements.
• Ensuring coordinated and safe movement is crucial for lower limb robotic exoskeletons, especially during human-robot interaction.

Purpose of the Study:

• To introduce a central pattern generator (CPG) network into a one-dimensional joint space admittance control framework.
• To enhance coordinated movement and user security in a custom-made four-DOF lower limb robotic exoskeleton.

Main Methods:

• A central pattern generator (CPG) network was employed to generate predefined trajectories for the four joints of the robotic exoskeleton.
• Unilateral knee joint torque was measured using muscle electromyography (EMG) signals.
• The measured torque was converted into state variables for the CPG via a one-dimensional admittance controller, allowing harmonic trajectory adjustments.

Main Results:

• The CPG network successfully generated coordinated joint trajectories.
• The integration of CPG with admittance control allowed for harmonic adjustments to predefined trajectories based on real-time user biomechanical feedback (EMG-derived torque).
• The robotic exoskeleton executed the adjusted trajectories accurately using a Proportional-Integral-Derivative (PID) controller.

Conclusions:

• The proposed CPG-based admittance control strategy effectively addresses the limitations of traditional controllers in multi-DOF robotic systems.
• This approach enhances the coordinated movement and safety of lower limb robotic exoskeletons, paving the way for more intuitive and responsive assistive devices.
• The study demonstrates a viable method for integrating biological signals (EMG) to modulate robotic exoskeleton behavior for improved human-robot synergy.