驾驶员分心监测方法探究

随着各类车载信息娱乐系统的发展,人们越来越渴望在车内与外界保持联系,注意力分散为驾驶安全带来了极大的威胁。调研了目前国际上对驾驶员的注意力进行监测的方法,发现普遍通过驾驶模拟实验对疲劳与视觉分心的捕获,但对于持续时间短、变化快的认知分心监测研究较少。在未来的分心监测研究中,需研究不同类型驾驶分心的表现特征,建立可靠、快速的驾驶分心在线监测算法。

人类司机分心行为对无人车纵向速度控制的影响

无人车通常无法探测到人类司机的分心行为,这将导致无人车延迟地采取紧急制动来避免追尾。因此,本文致力于构建无人车控制与人类司机分心监测之间的桥梁,来辅助无人车预测潜在风险并像有经验的人类司机一样避让处于分心的司机,提高无人车的智能化水平和交通系统的安全性。首先,本文提出了一种整合了司机分心监测、车对车信息交换和无人车速度控制的可行系统框架。然后,基于卷积神经网络,本文提供了一种司机分心监测实现。最后,基于模型预测控制策略,本文提出了一种考虑了司机分心行为的无人车纵向速度控制方法,并给出持续可行性分析。仿真结果验证了本文工作的有效性。

Chinese Business Cycles Identified by Coincident Indexes

In reference to the method of the Conference Board,the coincident indexes for China are constructed from a sample period between January 1990 and May 2012 and by51 chosen component indicators.The resulting coincident indexes have higher correlations with gross domestic product(GDP) growth rates than the China Economic Monitoring and Analysis Center(CEMAC) coincident index over the sample period between February2005 and May 2012.The peaks and troughs of the growth rates in several indicators are identified.The total number of peaks and troughs in the resulting coincident index is the same with the CEMAC coincident index.Unfortunately,these troughs don't signify recessions in the Chinese economy because the financial system has not seen a negative growth rate over the specific sample periods.The impacts of the Southeast Asian Financial Crisis and subprime mortgage crisis on the business cycle could be dated via a smoother index from HP filtering to the coincident index.