Project Details
Projekt Print View

Hybrid Models for High-precision sEMG-based Joint Torque / Movement Prediciton for Wearable Robotics

Subject Area Biomedical Systems Technology
Term from 2018 to 2023
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 398081449
 
Final Report Year 2023

Final Report Abstract

The JuMP project has produced a new, application-oriented, hybrid sEMG-based motion prediction model (domain model) that is easily adaptable to users despite its high parameter count. This domain model was contrasted with a black-box motion prediction model based on different sEMG features. Both types of models will be integrated in a further step. For the myotendinous actuator, as part of an improved motion prediction model, two approaches were proposed to determine the length distribution between muscle and tendon, one based on ultrasonic imaging and the other based on tracking the innervation zone center position of a muscle using a beamforming-like algorithm.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung