Distributed Repetitive Learning Control for Cooperative Cadence Tracking in Functional Electrical Stimulation Cycling

Area of Science:

• Rehabilitation Engineering
• Control Systems
• Neuroscience

Background:

• Functional electrical stimulation (FES) cycling aids mobility in neurological conditions (NCs).
• Current FES cycling controls face limitations due to high stimulation intensity and muscle fatigue.
• Need for advanced control strategies to optimize FES cycling rehabilitation.

Purpose of the Study:

• To design and evaluate a distributed repetitive learning controller (RLC) for FES-assisted cycling.
• To improve cycling cadence tracking while minimizing muscle fatigue in individuals with NCs.
• To integrate motorized assistance strategically during periods of low muscle torque output.

Main Methods:

• Developed a switched distributed RLC with independent learning feedforward terms for six lower-limb muscle groups and an electric motor.
• Implemented a controller that activates muscles in efficient kinematic regions and provides assistance when needed.
• Utilized Lyapunov-based stability analysis to guarantee tracking performance for the nonsmooth system.

Main Results:

• Experimental evaluation with able-bodied and NC participants demonstrated effective cadence tracking.
• Achieved a mean root-mean-squared cadence error of 3.58 ± 0.43 RPM for healthy individuals.
• Obtained a mean root-mean-squared cadence error of 4.26 ± 0.84 RPM for participants with NCs.

Conclusions:

• The distributed RLC effectively controls FES cycling, offering a promising rehabilitative strategy for individuals with NCs.
• The controller's adaptive learning and switched assistance minimize fatigue and enhance cycling performance.
• This approach represents a significant advancement in FES cycling technology for mobility restoration.