Research on muscle fatigue of upper limb in overhead static work

• Published in: Xibei Gongye Daxue Xuebao Vol. 42; no. 3; pp. 567 - 576
• Main Authors: YANG, Yanpu, AN, Weilan, HAN, Zhongjian, FAN, Yu, YANG, Qinxia
• Format: Journal Article
• Language: Chinese; English
• Published: EDP Sciences 01.06.2024
• Subjects: ergonomics; muscle fatigue; overhead static work; support vector machine; surface electromyography
• ISSN: 1000-2758; 2609-7125
• DOI: 10.1051/jnwpu/20244230567

To explore the muscle fatigue features of upper limb at different heights in overhead static work, an experiment was conducted to obtain the surface electromyography (sEMG) of subjects and their subjective fatigue state based on Borg CR-10 scale. The processing methods of time domain and frequency domain features of sEMG were studied and the multiclass support vector machine (SVM) was used to identify the state of muscle fatigue. By analyzing the muscular contribution, the correlation of subjective ratings and objective muscle fatigue features, ranking order of muscle fatigue accumulation, and muscular fatigue classification and identification, the results show that the muscles contribute above 10% on average are the biceps, deltoid and trapezius, and their cumulative contribution exceeds 70%; and the ranking orders of muscle fatigue accumulation in three heights are H 3 > H 2 > H 1 for biceps and trapezius and H 2 > H 3 > H 1 for deltoid; and with the time increase of overhand static operation, the muscle fatigue of upper limb gradually accumulates, resulting in the value of time domain features increases and the frequency domain features decreases, and their changes are consistent; and the accuracy of multiclass SVM is above 90% for identifying muscle fatigue of upper limb in overhead static work. 为探究手过头不同高度下静态作业的上肢肌肉疲劳特性, 通过实验设计采集了被试的表面肌电信号(surface electromyography, sEMG)及基于Borg CR-10量表的主观疲劳状态, 研究了sEMG的时域与频域特征处理方法, 并利用多分类支持向量机(support vector machine, SVM)识别肌肉疲劳状态。通过对肌肉贡献率、主客观肌肉疲劳特征的相关性、不同高度下的肌肉疲劳累积排序及肌肉疲劳分类识别进行分析, 结果表明: 肌肉平均贡献率超过10%的肌肉为肱二头肌、三角肌与斜方肌, 且其累积贡献率超过70%;对疲劳累积程度在3个高度下排序, 肱二头肌和斜方肌为 H 3 > H 2 > H 1 , 三角肌为 H 2 > H 3 > H 1 ; 随着手过头静态作业时间增加, 上肢肌肉疲劳逐渐积累, 时域特征值增加、频域特征值减小且其变化具有一致性; 多分类SVM对手过头静态作业中的上肢肌肉疲劳识别准确率大于90%。