基于步态加速度信号的人体疲劳检测研究
发布时间：2018-10-05 10:51
【摘要】：工作强度增加、工作时间延长、精神压力增大等诸多因素都会导致出现疲劳感。在疲劳状态下工作效率低,易引发安全事故,会给身体带来多种疾病。对人体疲劳状态进行实时检测越来越受到学术界的关注,在体育训练、运动健身、医疗康复等领域的实际应用价值更为突出。本课题通过对人的步态加速度信号进行采集和处理分析,从以下两个方面对人处于疲劳状态下行走时步态的改变展开了研究。 课题研究初期,通过观察人体前后方向的步态加速度信号波形图,发现疲劳前后存在明显的不同,因此根据信号在时域上的变化,用相关系数法求出阈值进行疲劳判断。对9个测试样本进行实验验证,检测结果的准确率为93.06%。该方法是针对同一个个体进行的疲劳检测,检测前需要先得到正常状态下的步态加速度信号,然后和测得的信号进行比较判断。 为了能够从步态加速度信号中挖掘更多的步态特征参数进行疲劳检测进行了更为深入的研究。通过对步态加速度信号进行了数学建模,设计相应算法得到各类步态特征参数,包括步态周期、步调、步态加速度均方根、自相关系数、峰峰值、FFT等。分别计算出疲劳和非疲劳状态下各类步态参数的均值和标准差,然后进行配对T检验,进而分析人体疲劳前后步态在时域和频域范围内的变化。研究结果表明,人体疲劳以后三个方向的步态加速度信号稳定性都会变差,垂直方向上幅值、频域范围的变化非常明显。今后可以使用这些步态特征参数从多角度、多层次进行疲劳检测。 总之,以上研究均说明了步态加速度特征可以被用来进行疲劳检测。前期研究是一种具体疲劳检测方法,而后期研究则更加深入的发掘出更多的可被用来进行疲劳检测步态特征参数,对今后疲劳检测的实际应用具有一定的指导意义。
[Abstract]:Fatigue will occur due to increased work intensity, prolonged working hours, increased mental stress and so on. Under fatigue condition, work efficiency is low, easy to cause safety accident, will bring many kinds of diseases to the body. The real-time detection of human fatigue status has attracted more and more attention from academic circles, and the practical application value in sports training, sports fitness, medical rehabilitation and other fields is more prominent. By collecting and processing the gait acceleration signal, the change of gait is studied from the following two aspects. At the beginning of the study, by observing the waveform of gait acceleration signal in front and rear direction of human body, it is found that there are obvious differences before and after fatigue. Therefore, according to the change of signal in time domain, the threshold value is calculated by correlation coefficient method to judge fatigue. Nine test samples were tested and the accuracy of the test results was 93.06. This method is aimed at fatigue detection of the same individual. The gait acceleration signal in normal state should be obtained before detection, and then compared with the measured signal. In order to extract more gait characteristic parameters from gait acceleration signal for fatigue detection, more in-depth research has been done. Based on the mathematical modeling of gait acceleration signal, the corresponding algorithm is designed to obtain various gait characteristic parameters, including gait period, gait pace, gait acceleration root mean square, autocorrelation coefficient, peak and peak FFT, etc. The mean and standard deviation of gait parameters in fatigue and non-fatigue state were calculated, and then matched T test was carried out to analyze the changes of gait in time domain and frequency domain before and after fatigue. The results show that the stability of gait acceleration signals in the three directions after fatigue becomes worse and the amplitude and frequency range in the vertical direction are very obvious. These gait characteristic parameters can be used to detect fatigue from multi-angle and multi-level in the future. In a word, all the above studies show that gait acceleration characteristics can be used for fatigue detection. The previous research is a specific fatigue detection method, while the later research is more in-depth to find out more characteristic parameters can be used for fatigue detection of gait, which has a certain guiding significance for the practical application of fatigue detection in the future.
【学位授予单位】：山西大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318.0;TN911.7
[Abstract]:Fatigue will occur due to increased work intensity, prolonged working hours, increased mental stress and so on. Under fatigue condition, work efficiency is low, easy to cause safety accident, will bring many kinds of diseases to the body. The real-time detection of human fatigue status has attracted more and more attention from academic circles, and the practical application value in sports training, sports fitness, medical rehabilitation and other fields is more prominent. By collecting and processing the gait acceleration signal, the change of gait is studied from the following two aspects. At the beginning of the study, by observing the waveform of gait acceleration signal in front and rear direction of human body, it is found that there are obvious differences before and after fatigue. Therefore, according to the change of signal in time domain, the threshold value is calculated by correlation coefficient method to judge fatigue. Nine test samples were tested and the accuracy of the test results was 93.06. This method is aimed at fatigue detection of the same individual. The gait acceleration signal in normal state should be obtained before detection, and then compared with the measured signal. In order to extract more gait characteristic parameters from gait acceleration signal for fatigue detection, more in-depth research has been done. Based on the mathematical modeling of gait acceleration signal, the corresponding algorithm is designed to obtain various gait characteristic parameters, including gait period, gait pace, gait acceleration root mean square, autocorrelation coefficient, peak and peak FFT, etc. The mean and standard deviation of gait parameters in fatigue and non-fatigue state were calculated, and then matched T test was carried out to analyze the changes of gait in time domain and frequency domain before and after fatigue. The results show that the stability of gait acceleration signals in the three directions after fatigue becomes worse and the amplitude and frequency range in the vertical direction are very obvious. These gait characteristic parameters can be used to detect fatigue from multi-angle and multi-level in the future. In a word, all the above studies show that gait acceleration characteristics can be used for fatigue detection. The previous research is a specific fatigue detection method, while the later research is more in-depth to find out more characteristic parameters can be used for fatigue detection of gait, which has a certain guiding significance for the practical application of fatigue detection in the future.
【学位授予单位】：山西大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318.0;TN911.7
【参考文献】
相关期刊论文 前10条
1 蔡立羽,王志中,张海虹;基于混沌、分形理论的表面肌电信号非线性分析[J];北京生物医学工程;2000年02期
2 侯向锋;刘蓉;周兆丰;;加速度传感器MMA7260在步态特征提取中的应用[J];传感技术学报;2007年03期
3 何庆华,吴宝明,彭承琳;表面肌电信号的分析与应用[J];国外医学.生物医学工程分册;2000年05期
4 庞世华;岳旺;;疲劳的表面肌电信号特征研究[J];内江科技;2009年05期
5 王r，
本文编号：2253103
本文链接：https://www.wllwen.com/yixuelunwen/swyx/2253103.html
