基于BP神经网络的驾驶疲劳预测模型及其有效性分析  

Driving fatigue prediction model with validity analysis based on BP neural network

作　　者：贾安琪 邓超 JIA Anqi;DENG Chao(School of Automobile and Traffic Engineering,Wuhan University of Science and Technology,Wuhan 430065,Hubei,China;Intelligent Automobile Engineering Research Institute,Wuhan University of Science and Technology,Wuhan 430065,Hubei,China;Unmanned System Intelligent Perception Control Technology Engineering Laboratory of Sichuan Province,Chengdu 610225,Sichuan,China;Laboratory of Cloud IoT Intelligent Equipment in Expressway Construction and Maintenance,Jinan 250357,Shandong,China)

机构地区：[1]武汉科技大学汽车与交通工程学院,湖北武汉430065 [2]武汉科技大学智能汽车工程研究院,湖北武汉430065 [3]四川省无人系统智能感知控制技术工程实验室,四川成都610225 [4]云基物联网高速公路建养设备智能化实验室,山东济南250357

出　　处：《农业装备与车辆工程》2024年第2期40-43,共4页Agricultural Equipment & Vehicle Engineering

基　　金：国家自然科学基金青年基金项目(52002298);教育部产学合作协同育人项目(202102580026);四川省无人系统智能感知控制技术工程实验室开放课题(WRXT2022-001);云基物联网高速公路建养设备智能化实验室开放课题;武昌工学院科学研究项目(2022KY24)。

摘　　要：疲劳驾驶是导致交通事故的重要原因之一,对疲劳驾驶行为进行预测可以有效减少事故发生。通过设计高速路段实车疲劳驾驶实验,获取相关特征指标,对特征参数进行研究,建立基于BP神经网络的驾驶员疲劳驾驶状态预测模型,对驾驶员的疲劳状态进行预测,结果表明精度达到90%以上;并对不同缺失率下的实验数据进行疲劳预测,判断该模型在数据缺失时是否有效,该模型在数据缺失率达到20%时预测精度仍达到80%,基于BP神经网络进行疲劳预测具有较高有效性。Fatigue driving is one of the most important reasons causing traffic accidents,which brings serious casualties and property losses,so the prediction of fatigue driving behavior is pretty necessary to reduce the occurrence of accidents.By designing the fatigue driving experiment of real vehicles in the high-speed section,the relevant characteristic indicators were obtained,the characteristic parameters were studied,and a driver fatigue driving state prediction model based on BP neural network was established to predict the driver's fatigue state with an accuracy of more than 90%.The fatigue prediction of the experimental data under different deletion rates was carried out to determine whether the model was effective when the data was missing,and the prediction accuracy of the model still reached 80%when the data deletion rate reached 20%,so fatigue prediction based on BP neural network was highly effective.

关 键 词：疲劳驾驶 BP神经网络 缺失数据 有效性 

分 类 号：U461[机械工程—车辆工程] TP273[交通运输工程—载运工具运用工程]