Title :
A hybrid EMG model for the estimation of multijoint movement in activities of daily living
Author :
Ding Qichuan ; Zhao Xingang ; Han Jianda
Author_Institution :
State Key Lab. of Robot., Shenyang Inst. of Autom. (SIA), Shenyang, China
Abstract :
Accurately identifying human´s intent of motion from electromyography (EMG) signals is the key to implement EMG-based HRI (Human-Robot Interface) systems. Human´s intent of motion includes motion modes and continuous movement variables. In this paper, a hybrid EMG-to-motion model is constructed by combining a classification model and a regression model. Based on a proper division for joints, the classification model is utilized to recognize the motion modes of `small´ joints; meanwhile, the regression model is utilized to estimate the continuous movement variables of `big´ joints. Furthermore, a Bayesian network (BN) model, which sufficiently employs context information of a task, is also involved into the hybrid model to improve its performances for motion estimation. Experiments have been conducted with three subjects to demonstrate the feasibility of the proposed methods. In these experiments, the motion modes of hand and wrist, and the continuous elbow angles are estimated with sEMG signals considering a `drinking´ task. Finally, an upper limb prosthetic is controlled to simulate human´s movement in a `drinking´ task.
Keywords :
belief networks; control engineering computing; electromyography; human-robot interaction; medical signal processing; motion estimation; prosthetics; regression analysis; BN model; Bayesian network; EMG-based HRI; classification model; daily living; electromyography; human-robot interface systems; hybrid EMG model; motion estimation; motion recognition; multijoint movement estimation; regression model; sEMG signals; upper limb prosthetic; Context modeling; Elbow; Estimation; Feature extraction; Joints; Muscles; Vectors; Human-Robot Interface; motion estimation; pattern recognition; surface electromyography;
Conference_Titel :
Multisensor Fusion and Information Integration for Intelligent Systems (MFI), 2014 International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4799-6731-5
DOI :
10.1109/MFI.2014.6997746