Towards Personalized Control for Powered Knee Prostheses: Continuous Impedance Functions and PCA-Based Tuning Method.

Area of Science:

• Biomedical Engineering
• Robotics
• Control Systems

Background:

• Personalizing prosthetic devices requires optimizing control parameters, but current finite state machine impedance control (FSM-IC) methods involve time-consuming manual tuning.
• Efficient tuning is essential for improving user interaction and the performance of prosthetic limbs.

Purpose of the Study:

• To develop a novel, efficient approach for tuning knee prostheses using continuous impedance functions (CIFs) and Principal Component Analysis (PCA).
• To reduce the manual effort and time required for optimizing prosthetic control parameters.

Main Methods:

• Modeled CIFs (stiffness, damping, equilibrium angle) as fourth-order polynomials and optimized them using convex optimization.
• Applied PCA to the CIFs to extract principal components (PCs) representing common features.
• Utilized PC weights as tuning parameters for reconstructing various impedance functions.

Main Results:

• Successfully generated CIFs through convex optimization, creating a new, efficient tuning space.
• Demonstrated the feasibility of the proposed tuning space using walking data from 10 able-bodied individuals.
• Validated the ability to reconstruct diverse impedance functions using the principal component weights.

Conclusions:

• The proposed CIFs and PCA-based tuning method offers a more efficient and systematic approach to personalizing knee prostheses.
• This novel tuning space simplifies the optimization process, potentially leading to improved prosthetic performance and user experience.
• The approach is validated and shows promise for future applications in prosthetic control system development.