论文部分内容阅读
为了克服单一通道信息在驾驶疲劳行为判定中的局限性,提出了综合运用多通道信息融合共同判定驾驶疲劳行为的方法。该方法在充分考虑各信息源相关性和互补性的基础上,优化采用驾驶人疲劳特征ECD、车道偏离程度SAAE、方向盘转动程度SWA等疲劳判别指标,运用MVAR进行多维特征向量提取,以有向无环支持向量机为融合算法,建立了基于多分类支持向量机的驾驶疲劳行为判定模型。结果表明,运用DAG-SVM进行多通道信息决策提高了疲劳驾驶行为检测的准确性和可靠性。
In order to overcome the limitations of single-channel information in judging driver fatigue behavior, a method of synthetically using multi-channel information fusion to jointly determine driving fatigue behavior is proposed. Based on the correlation and complementarity of all information sources, this method optimizes the fatigue discrimination indexes such as driver fatigue characteristics ECD, lane departure degree SAAE and steering wheel rotation degree SWA, and uses MVAR to extract the multidimensional eigenvectors. The acyclic support vector machine is a fusion algorithm, and a driving fatigue behavior determination model based on multi-class support vector machine is established. The results show that using DAG-SVM for multi-channel information decision-making improves the accuracy and reliability of fatigue driving behavior detection.