基于脉搏波特征融合的驾驶疲劳检测方法

疲劳驾驶是交通事故的主要诱因之一，精确检测驾驶人的疲劳程度是主动预防疲劳驾驶事故的核心内容之一。通过开展自然驾驶试验，以驾驶人的生物信号脉搏波（Blood Pressure Waveform，BPW）为数据源，使用脉搏波波形分析方法从中提取有效表征驾驶疲劳的特征指标，构建用于检测驾驶疲劳等级的BPW特征指标集，在此基础上引入D-S证据理论建立了基于BPW特征融合的驾驶疲劳检测模型。结果表明：该模型对测试数据的疲劳驾驶理论检测精度达到了91.8%，优于贝叶斯网络模型的81.4%和支持向量机模型的84.3%，能够满足实际应用的需求，但与决策回归树检测模型99.7%的精度相比较还有差距。研究获得的基于生物信息融合的驾驶疲劳检查模型和方法在驾驶疲劳检测与监测中具有很好的应用前景，可为辅助安全驾驶和疲劳预警及主动干预提供新的技术方案。

Driver Fatigue Detection Model Based on BPW Feature Fusion

Fatigue driving is one of the leading causes of traffic accidents. Therefore, accurately detecting a driver's fatigue level using technology is crucial for saving drowsy drivers from accidents. In this paper, biological signals relating to a driver's blood pressure waveform (BPW) were collected by naturalistic driving tests, and the BPW analysis method was employed to extract relevant features to detect driving fatigue effectively. A dataset of BPW features for driving fatigue detection was then constructed. Meanwhile, the Dempster-Shafer (D-S) evidence theory was introduced to build a driver fatigue detection model using those BPW features. Compared with other detection models employing Bayes Network, SVM, and REPTree with accuracies of 81.4%, 84.3% and 99.7%, respectively, the detection model based on D-S evidence theory has a theoretical accuracy of 91.8% under the current test conditions, which satisfies the current application but still falls short of the model that employs REPTree. The driver fatigue detection model, based on biological information fusion, described in this paper, will have good application prospects in detecting and monitoring driving fatigue, providing a novel technical solution for driver safety assistance systems for fatigue warning and intervention.