Model Predictive Control of a Feedback-Linearized Hybrid Neuroprosthetic System With a Barrier Penalty.

Area of Science:

• Biomedical Engineering
• Rehabilitation Robotics
• Control Systems

Background:

• Functional electrical stimulation (FES) is a key therapy for neurological impairments.
• FES-induced muscle fatigue limits treatment duration and effectiveness.
• Existing FES systems lack methods to mitigate fatigue dynamically.

Purpose of the Study:

• To develop an electric motor-assist system to reduce FES-induced muscle fatigue.
• To implement a model predictive control (MPC) strategy for workload allocation between FES and motor-assist.
• To enhance the usability and effectiveness of FES for motor function restoration.

Main Methods:

• Proposed an electric motor-assist system integrated with FES.
• Utilized model predictive control (MPC) for optimal control input allocation.
• Employed feedback linearization to simplify the system dynamics for MPC.
• Incorporated a barrier cost function to handle state-dependent input constraints.

Main Results:

• The proposed system effectively reduced FES-induced muscle fatigue.
• Satisfactory control performance was achieved in simulations.
• Feedback linearization significantly reduced computational load.
• Preservation of key state variables (angular position, muscle fatigue) ensured meaningful optimization.

Conclusions:

• The combined FES and electric motor-assist system shows promise for alleviating muscle fatigue.
• MPC with feedback linearization offers an efficient control strategy for assistive devices.
• This approach can improve the quality and duration of FES-based rehabilitation therapies.