Towards Intuitive Control of Simultaneous Movements and Force Estimation for Myoelectric Prostheses

Rosenvang, Jakob Celander ; Horup, Ronnie

4. term

Biomedical Engineering and Informatics, Master

2012

2012-06-01

92

Håndamputation reducerer livskvaliteten markant, og selvom myoelektriske proteser kan genskabe nogle funktioner, er styringen ofte begrænset til sekventielle bevægelser og opleves ikke altid intuitiv. Dette projekt undersøger, hvordan styringen kan forbedres, så flere bevægelser kan udføres samtidigt, og kraften kan doseres proportionalt. Vi gennemførte et eksperiment med 10 raske forsøgspersoner og indsamlede EMG-data for at udvikle og optimere et hybridsystem, der kombinerer bevægelsesklassifikation med kraftestimation. Systemet understøtter otte distinkte håndledsbevægelser samt hvile ved kontinuerte kraftniveauer. Validering viste en klassifikationsnøjagtighed på 92,2 % for ni bevægelser, og kraften kunne estimeres præcist (R^2 = 0,92). Den foreslåede kontrolstrategi muliggør samtidig styring af to frihedsgrader og er egnet til realtidskontrol, hvilket markerer et skridt mod mere naturlig og intuitiv anvendelse af myoelektriske proteser.

Hand amputation substantially reduces quality of life, and although myoelectric prostheses can restore some function, control is often limited to sequential movements and may feel unintuitive. This project addresses how to enable more intuitive control that supports simultaneous movements with proportional force. An experiment with 10 healthy participants was conducted to collect EMG data and develop an optimized hybrid control scheme that combines movement classification with force estimation. The system supports eight distinct wrist movements plus rest at continuous force levels. Validation showed a classification accuracy of 92.2% across nine movements, and force was estimated with high precision (R^2 = 0.92). The proposed approach enables simultaneous control of two degrees of freedom and is suitable for real-time use, representing a step toward more natural, intuitive control of myoelectric prostheses.

[This summary has been generated with the help of AI directly from the project (PDF)]

Download
View record in AAU Student Projects