Semester

Summer

Date of Graduation

2020

Document Type

Thesis

Degree Type

MS

College

Statler College of Engineering and Mineral Resources

Department

Industrial and Managements Systems Engineering

Committee Chair

Ashish D. Nimbarte

Committee Member

Warren R. Myers

Committee Member

Gary L. Winn

Abstract

A reduction in the % stationarity of surface electromyography (SEMG) signals with respect to the initial or fresh condition is used to predict localized muscle fatigue. However, factors other than muscle fatigue can also influence the stationarity of SEMG signals. This study was aimed at analyzing the effect of various work/task related factors on the stationarity of SEMG signals obtained from non-fatigued shoulder muscles. Twelve participants were recruited for data collection and each one performed 120 trials characterized by the combination of 2 shoulder angles (60° and 120°), 2 planes of exertions (sagittal and scapular), 3 force levels (0lb, 2.5lb, 5lb), 5 force directions (pull back, pull up, pull down, pull right and pull left) and 2 repetitions. The SEMG data were recorded from seven shoulder muscles (supraspinatus, infraspinatus, middle deltoid, anterior deltoid, posterior deltoid, biceps, triceps). Modified Reverse Arrangement Test with five window sizes (128, 256, 512, 768, and 1024 millisecond(ms)) was used to process the SEMG data. The effects of work/task related factors on % stationarity of shoulder muscles was analyzed using ANOVA. The mean stationarity of SEMG signals ranged from 87.8% to 94.9%. Among the work/task related factors, the joint angle and the plane of exertion affected the % stationarity in fewer instances compared to the force level and the force direction. The exertions that produced higher activation (SEMG amplitude) resulted in lower % stationarity, indicating an inverse relationship between % stationarity and muscle activation. The variability in % stationarity increased from 3.3% to 10.0% when the window size was increased from 128 ms to 1024 ms. The study findings could be useful in improving real-time fatigue prediction methods.

Share

COinS