基于EEG信号的脑力疲劳检测方法的研究

发布时间：2018-06-05 14:11

本文选题：脑电信号 + 脑力疲劳　；参考：《广西大学》2014年硕士论文

【摘要】：脑力疲劳是由于长时间工作或学习造成的一种常见生理现象,不及时调整和恢复会降低人们工作和学习的效率,严重的会威胁到人们的健康与生命。脑电(Electroencephalogram,简称EEG)信号直接反映了大脑组织的电活动,利用其来评估脑力疲劳已经成为脑力疲劳检测的研究热点。目前,基于EEG信号脑力疲劳检测中都采用多通道脑电采集设备,由于该类设备的局限性导致基于EEG信号脑力疲劳检测只能在实验室进行为了克服该类设备在实际应用中携带不便、操作复杂成本高等局限性,本文探讨了使用便携式脑电采集设备采集单通道EEG信号对脑力疲劳进行检测的方法。在前人研究的基础上,主要完成了以下工作： 1、对基于EEG信号脑力疲劳检测研究中所使用的脑电采集设备、电极选取以及特征提取方法的国内外研究现状作了较全面的介绍。 2、针对小波包快速算法固有的频率混淆缺陷,提出了一种单子带重构小波包改进算法(Improved Single Sub-band Reconstruction of Wavelet Packet Algorithm,简称ISSBR-WPA),该算法的思路是引入两个算子来消除小波包分解过程与重构过程中各子带多余的频率成分,从而有效地克服频率混淆现象的产生。实验结果表明,与小波包快速算法相比,ISSBR-WPA算法较为准确地提取EEG信号中的δ、θ、αα和p四个节律,为准确计算脑力疲劳特征参数提供了保障。 3、采用两种特征参数评估大脑是否处于疲劳状态。两种特征参数分别为基于EEG信号四个节律相关能量比的8个特征F1～F8和基于各子带小波包系数的方差。对便携式设备采集大脑FP1处的EEG信号进行特征提取与分析,实验结果表明,8个能量比中特征F2、F3、F4、F6和F7可作为评估脑力疲劳的有效指标,其中特征F2更为有效；低频部分子带的小波包系数方差能有效地区分清醒和脑力疲劳两种精神状态。 4、为了验证基于便携式脑电采集设备的特征提取结果的有效性,利用多通道脑电采集设备采集得到FP1和01导联处的EEG信号分别进行相同的特征提取和分析。实验结果表明,多通道脑电采集设备采集FP1处EEG信号所提取的特征参数表征的脑力疲劳状态和便携式设备所表征的脑力疲劳状态一致。说明采用便携式脑电采集设备采集大脑FP1处的EEG信号能检测大脑的疲劳状态。实验结果还表明,多通道设备采集大脑O1处的EEG信号也可以作为分析大脑疲劳状态的信号。
[Abstract]:Mental fatigue is a common physiological phenomenon caused by long hours of work or study. Not adjusting and recovering in time will reduce the efficiency of work and study, and will seriously threaten people's health and life. Electroencephalograms (EEGG) signals directly reflect the electrical activity of brain tissue, and the use of these signals to assess brain fatigue has become a research hotspot in the detection of brain fatigue. At present, multichannel EEG acquisition devices are used in EEG signal brainpower fatigue detection. Because of the limitation of this kind of equipment, the brainpower fatigue detection based on EEG signal can only be carried out in the laboratory in order to overcome the limitation of the equipment carrying inconvenience in practical application and the high cost of complex operation, etc. In this paper, the method of detecting mental fatigue by using portable EEG acquisition equipment to collect single channel EEG signals is discussed. The main works are as follows: 1. The EEG acquisition equipment which is used in the research of EEG signal brain fatigue detection is studied. The research status of electrode selection and feature extraction methods at home and abroad is introduced. 2. Aiming at the inherent frequency confusion defect of wavelet packet fast algorithm, In this paper, an improved single Sub-band Reconstruction of Wavelet packet algorithm (ISSBR-WPAA) is proposed. The idea of this algorithm is to introduce two operators to eliminate the redundant frequency components of each sub-band in the process of wavelet packet decomposition and reconstruction. Thus the frequency confusion can be effectively overcome. The experimental results show that the ISSBR-WPA algorithm is more accurate than the wavelet packet algorithm in extracting the 未, 胃, 伪 and p rhythms of EEG signals. It provides a guarantee for calculating the characteristic parameters of mental fatigue. 3. Two characteristic parameters are used to evaluate whether the brain is in a fatigue state. The two characteristic parameters are eight feature F1F8 based on the four rhythmic correlation energy ratio of EEG signals and the variance of wavelet packet coefficients based on each sub-band. The EEG signals collected from FP1 of the brain by portable equipment were extracted and analyzed. The experimental results show that the characteristics F2F3F4F4F6 and F7 can be used as effective indexes for evaluating mental fatigue, among which feature F2 is more effective. The variance of wavelet packet coefficients of the low frequency subbands can effectively distinguish two mental states: awake and mental fatigue. 4. In order to verify the validity of feature extraction based on portable EEG acquisition equipment, The EEG signals in leads FP1 and 01 were extracted and analyzed respectively by using multi-channel EEG acquisition equipment. The experimental results show that the characteristic parameters extracted from EEG signals collected from FP1 by multi-channel EEG acquisition equipment are consistent with the mental fatigue state represented by portable devices. It is concluded that EEG signals collected from FP1 can be used to detect the fatigue state of the brain. The experimental results also show that EEG signals at O1 can also be used to analyze the fatigue state of the brain.
【学位授予单位】：广西大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7

【参考文献】