基于脑电信号的正负情绪分类方法研究

Method Research on Positive and Negative Emotions Classification Based on EEG

驾驶员情绪状态的实时识别与预警,对保证道路交通安全系统的正常运行有着重要的作用与意义。本研究基于便携式脑电设备采集了16位被试前额双通道脑电数据,分别从时域和频域上进行特征提取,使用集成学习分类的方法对正负情绪进行分类。结果显示频域特征以及特征的不对称指数在正负性情绪的分类起到了关键的作用,得到基于梯度提升决策树(GBDT)分类器的正负情绪识别准确率最佳,为92.4%。本研究提出了一种对驾驶员正负性情绪状态识别的新方法,为后续情绪状态的实时识别奠定了基础。

The real-time recognition and early warning of driver's emotional state is of great significance to ensure the normal operation of road traffic safety system.Based on the portable EEG device,the EEG data of 16 subjects are collected,and the EEG features are extracted from time-domain and frequency-domain,then the positive and negative emotions are classified by different integrated learning classifiers.The results show that the frequency-domain features and the asymmetry index of all features play a key role in the classification of positive and negative emotions,and the accuracy of positive and negative emotions recognition based on the Gradient Boosting Decision Tree(GBDT)classifier is the best,which is 92.4%.This study proposes a new method for drivers’positive and negative emotional state recognition,which lays a foundation for real-time recognition of emotions.