DocumentCode :
828096
Title :
The identification of time series models of lower extremity EMG for the control of prostheses using Box-Jenkins criteria
Author :
Triolo, R.J. ; Nash, D.H. ; Moskowitz, Gordon D.
Author_Institution :
Shriners Hosp. for Crippled Children, Philadelphia, PA, USA
Volume :
35
Issue :
8
fYear :
1988
Firstpage :
584
Lastpage :
594
Abstract :
The Box-Jenkins method of model identification and diagnostic checking was used on the EMG of the posterio-lateral thigh while contraction level, electrode position, and limb function were varied. Models remained exclusively autoregressive (AR) of order typically less than five for isometric contractions ranging from 25 to 50% of the maximum voluntary knee flexions and extensions, and for recording sites located within 30 mm of the original electrode position. Diagnostics performed on the residuals from AR(4) models indicate that a fourth-order model is adequate for myoelectric control applications under many circumstances.
Keywords :
artificial limbs; bioelectric potentials; identification; muscle; physiological models; Box-Jenkins criteria; autoregressive models; contraction level; diagnostic checking; electrode position; isometric contractions; knee extensions; limb function; lower extremity EMG; maximum voluntary knee flexions; model identification; myoelectric control; posterio-lateral thigh; prosthesis control; time series models; Electrodes; Electromyography; Extremities; Filters; Knee; Predictive models; Prosthetics; Signal processing; Thigh; White noise; Electromyography; Humans; Isometric Contraction; Knee Prosthesis; Leg; Models, Biological; Muscles; Prosthesis Design; Self-Help Devices; Time Factors;
fLanguage :
English
Journal_Title :
Biomedical Engineering, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9294
Type :
jour
DOI :
10.1109/10.4591
Filename :
4591
Link To Document :
بازگشت