基于脑电的实时疲劳监测算法的研究

疲劳驾驶是造成交通事故的一个重要因素。基于脑电波传感器采集到的脑电(EEG)信号,提出了一种基于特征节律波样本空间距离的实时驾驶疲劳监测算法。该算法通过提取脑电特征信号并建立相应的特征样本空间,然后通过将实时采集的脑电信号与样本空间之间的距离来判别驾驶员是否处于疲劳状态;并结合多次实验结果,使用ROC曲线分析方法获得确定疲劳判别的阈值,由此实现对驾驶者疲劳状态的实时监测。实验结果显示,该算法能有效地、实时地对疲劳进行监测,且具有较好的鲁棒性和准确性。

Fatigue monitor algorithm research based on EEG

Fatigue drive is an important factor of causing a traffic accident. Based on the electroencephalogram (EEG) signals collect by brain wave sensor. This paper proposed a real-time driving fatigue monitoring algorithm based on the distance of the sample space of feature rhythm wave.The algorithm extracts EEG signal and establish the features sample space, the compute the distance of the real-time acquisition EEG feature between sample space and determine whether the driver is fatigue,Through many experiments result to get the threshold by ROC curve analysis,thereby achieve real-time monitoring. Experimental results show that the algorithm can effectively and in real-time fatigue monitoring, and has good robustness and accuracy.