驾驶员隧道段夜间行车动视点特征模糊聚类评价  

Fuzzy Cluster Evaluation Driver's Eye Movement Characteristic of Tunnel Section at Night

作　　者：国威[1] 潘晓东[1] 邓其[1] 喻泽文[2] 

机构地区：[1]同济大学交通运输工程学院,上海201804 [2]湖南省交通科学研究院,湖南长沙410015

出　　处：《公路工程》2013年第6期97-101,共5页Highway Engineering

基　　金：国家自然科学基金资助项目(50878157);浙江省交通厅科技资助项目(2009H10)

摘　　要：通过隧道路段行车试验,利用Smart Eye AB型眼动仪对驾驶员夜间隧道动视点数据进行采集并进行处理。选取视点停留比率、显著可见区域、扫视幅度、扫视峰值速度、扫视持续时间5项指标作为驾驶员动视点特征主要影响因素。利用模糊聚类分析中的传递闭包聚类方法对驾驶员夜间隧道路段行车视觉特点进行了分析与评价。通过驾驶员对路段的熟悉程度可将动视点特征分为6类。结果表明:该模糊聚类评价方法能够有效反映驾驶员夜间路段视觉信息获取与加工特征,可为保障夜间行车安全提供理论与实践依据。Through the driving test of the tunnel sections, using of the Smart Eye AB eye tracker to collect the driver＇s dynamic viewpoint data at night of the tunnel sections acquisition and processing. Then, selected five characterized visual parameters： viewpoint stays ratio, conspicuity area, saccade am- plitude, saccade peak velocity, saccade velocity as the main factors which influenced the driver＇ s moving viewpoint. Finally, used transitive closure clustering method of fuzzy clustering analysis to analysis and e- valuate driver＇ s visual characteristics of tunnel sections at night. By the familiarity of the driver on the road section, the dynamic viewpoint can be characterized into six categories. The results show that the fuzzy clustering evaluation method can effectively reflect the characteristics of the driver＇s visual informa- tion acquisition and processing of tunnel sections at night, and it can provided theoretical and practical basis for the protection of driving safety at night.

关 键 词：隧道路段 夜间行车 动视点指标 视觉信息 模糊等价关系 传递闭包聚类 

分 类 号：U453.7[建筑科学—桥梁与隧道工程]