Related Experiment Video
Updated: Jul 19, 2026
11:16
Engineering Platform and Experimental Protocol for Design and Evaluation of a Neurally-controlled Powered Transfemoral Prosthesis
Published on: July 22, 2014
Real-time myoprocessors for a neural controlled powered exoskeleton arm
Ettore E Cavallaro1, Jacob Rosen, Joel C Perry
1Department of Electrical Engineering, University of Washington, Seattle, WA 98185, USA. cavallaro@sssup.it
IEEE Transactions on Bio-Medical Engineering
|November 1, 2006
Summary
This study developed a muscle model, the "myoprocessor," for exoskeleton control. Optimized using genetic algorithms, it accurately predicts joint torques, enabling intuitive exoskeleton arm integration for rehabilitation.
Area of Science:
- Robotics
- Biomechanics
- Neuroscience
Background:
- Exoskeleton robots offer promising assistive and rehabilitative solutions for individuals with force deficits or post-stroke recovery.
- Effective control of exoskeletons relies heavily on the human-machine interface (HMI).
- A neuro-muscular level HMI can achieve seamless integration, allowing exoskeleton control as a natural extension of the human body.
Purpose of the Study:
- To develop and validate myoprocessors for upper limb exoskeleton control.
- To implement a real-time muscle model (myoprocessor) for predicting joint torques based on kinematics and neural activation.
- To optimize myoprocessor parameters using genetic algorithms and experimental data.
Main Methods:
- Development of myoprocessors based on the Hill phenomenological muscle model for the upper limb.
- Optimization of internal myoprocessor parameters using genetic algorithms.
- Validation against an experimental database correlating model inputs and performance.
Main Results:
- High correlation observed between predicted joint moments from the myoprocessor model and measured experimental data.
- Demonstrated robustness of the myoprocessor model for real-time applications.
- Successful optimization of model parameters through genetic algorithms.
Conclusions:
- The developed myoprocessor model is a robust and adequate component for exoskeleton HMI.
- The myoprocessor facilitates intuitive control of upper limb exoskeletons.
- Further integration into exoskeleton control systems is warranted based on promising results.
