基于无线体域网和复合生理信号近似熵的驾驶疲劳研究

发布时间：2018-06-04 10:05

本文选题：无线体域网 + 复合生理信号　；参考：《中国生物医学工程学报》2017年05期

【摘要】：为有效合理地评价驾驶员在驾驶过程中的疲劳程度,通过无线体域网(WBAN)采集12名被试者的3种生理信号(脑电信号、肌电信号、呼吸信号),提取并分析这些生理信号的近似熵在驾驶过程中对疲劳程度的反映效果。试验结果表明,脑电、肌电、呼吸3种信号的近似熵均随时间的增加而逐渐下降,约90 min后下降程度变缓,表明进入比较疲劳状态。通过主成分分析可知,前两个主成分的贡献率分别为47.33%和40.26%,二者之和大于85%,其中脑电信号和肌电信号的近似熵所占权重较大。而且统计分析表明,脑电信号和肌电信号的近似熵显著性P0.05,表明脑电信号和肌电信号在表征正常与疲劳两种状态时差异明显。在此基础上,分析信号组合对驾驶疲劳的反映效果,结果表明在脑电-肌电信号的近似熵组合时,正常和疲劳状态的概率分布具有明显的界限,可以有效反映驾驶过程中的疲劳状态。通过研究,获得最佳的反映疲劳驾驶的信号组合,可为更加准确地检测和避免疲劳驾驶提供理论依据。
[Abstract]:In order to evaluate driver fatigue effectively and reasonably, three physiological signals (EEG, EMG) were collected from 12 subjects by WBAN. Respiratory signal, extraction and analysis of the approximate entropy of these physiological signals in the driving process to reflect the degree of fatigue. The experimental results showed that the approximate entropy of EEG, EMG and respiration decreased gradually with the increase of time, and slowed down after about 90 min, indicating that they had entered a relatively fatigue state. Through principal component analysis, the contribution rates of the first two principal components were 47.33% and 40.26%, respectively, and the sum of the two components was more than 85, in which the approximate entropy of EEG and EMG accounted for a large weight. The statistical analysis shows that the approximate entropy of EEG and EMG is significant (P0.05), which indicates that the difference between EEG and EMG is significant when they represent normal and fatigue states. On this basis, the effect of signal combination on driving fatigue is analyzed. The results show that the probability distribution of normal and fatigue states has obvious limits when the approximate entropy combination of EEG and EMG signals is used. It can effectively reflect the fatigue state in the driving process. Through the research, the optimal signal combination of fatigue driving can be obtained, which can provide a theoretical basis for more accurate detection and avoidance of fatigue driving.
【作者单位】：沈阳工程学院机械学院;东北大学机械工程与自动化学院;燕山大学测控技术与仪器河北省重点实验室;
【基金】：国家自然科学基金(51605419) 博士后启动基金(2016M600193) 河北省科技计划项目(152177180)
【分类号】：R318;TN911.6

【相似文献】