夜间环境下城市道路驾驶人视觉搜索模式研究

发布时间：2019-05-18 11:59

【摘要】：驾驶人因素是引发道路交通事故的主要致因,而在驾驶人因素中占据主导地位的是感知因素。驾驶人对信息的感知绝大部分来自于视觉,夜间环境下交通事故具有多发性和高死亡率的特点,所以对夜间不同环境下的眼动这一基本视觉行为的研究,有助于揭示驾驶人在夜间驾驶过程中的视觉搜索规律,对改进道路交通安全设施的设计标准,完善交通管理措施,均有着重要的理论指导和实际意义。选取7名符合实验要求的驾驶人,在驾驶模拟器上完成昼夜环境下相同城市道路交通环境的驾驶任务,使用眼动仪记录驾驶人在驾驶过程中的相关眼动数据；结合国内外对驾驶人视觉搜索模式的理论研究,确定了以累积注视时间、视角特征、注视持续时间和视角搜索广度几个指标作为视觉搜索模式的基本表征参数。运用统计学方法对采集的数据进行处理和分析。首先统计分析驾驶人的注视时间、视角特征和视觉搜索广度随昼夜环境变化的规律和差异性；其次,运用动态聚类法,研究驾驶人在右转驶入、直行驶入和左转驶入交叉口时的视野平面,依据动态聚类的结果进行注视区域划分,分析驾驶人昼夜环境下的注视区域差异：此外,对注视目标进行逐一统计,分析驾驶人在驶入信号交叉口时的注视目标频次,用马尔科夫一步转移概率矩阵将驾驶人在交叉口内的注视规律矩阵化,并求出注视平稳分布：最后,论文将上述研究结果应用于交通管理政策方法制定和交通设施优化设计。研究结果表明,昼夜不同环境下对驾驶人的视觉搜索模式会产生一定的影响。主要研究结论如下：1.驾驶人的水平视角和垂直视角随着昼夜环境变化存在显著性差异；而视觉搜索广度和注视持续时间在昼夜环境下未体现显著性差异。2.用动态聚类方法,将驾驶人在昼夜环境下采用不同方式驶入交叉口的注视区域进行划分,发现注视区域划分会根据昼夜环境的不同体现差异性。3.将驾驶人驶入交叉口时的视频逐帧回放,获得昼夜环境下驾驶人采用不同方式驶入交叉口时的注视目标频次和视觉热点,发现注视频次和视觉热点表现出差异性。4.驾驶人在交叉口内的一步转移概率矩阵和平稳分布会因为昼夜环境的不同体现差异性。
[Abstract]:The driver factor is the main cause of road traffic accidents, and the dominant factor in the driver factor is the perceptual factor. Most of the drivers' perception of information comes from vision. Traffic accidents in the night environment have the characteristics of multiple and high mortality, so the basic visual behavior of eye movement in different environments at night is studied. It is helpful to reveal the visual search law of drivers in the process of driving at night, and has important theoretical guidance and practical significance for improving the design standard of road traffic safety facilities and perfecting traffic management measures. Seven drivers who met the requirements of the experiment were selected to complete the driving task of the same urban road traffic environment under the same day and night environment on the driving simulator, and the related eye movement data of the driver in the driving process were recorded by the eye movement instrument. Combined with the theoretical research on the driver's visual search mode at home and abroad, the cumulative fixation time, visual angle characteristics, gaze duration and visual angle search breadth are determined as the basic representation parameters of the visual search pattern. The statistical method is used to process and analyze the collected data. Firstly, the regularity and difference of the driver's gaze time, visual angle characteristics and visual search breadth with the diurnal environment are statistically analyzed. Secondly, the dynamic clustering method is used to study the visual field plane of the driver when he turns right, straight into and left into the intersection, and the fixation area is divided according to the results of dynamic clustering. This paper analyzes the difference of fixation area in the driver's circadian environment: in addition, the fixation target is counted one by one, and the frequency of the gaze target when the driver enters the signalized intersection is analyzed. Markov one-step transfer probability matrix is used to matrix the fixation law of drivers at intersections, and the stationary distribution of fixation is obtained. finally, the above research results are applied to the formulation of traffic management policy methods and the optimization design of traffic facilities. The results show that the visual search pattern of drivers will be affected to a certain extent in different environments day and night. The main conclusions are as follows: 1. There was significant difference between horizontal angle of view and vertical angle of view with the change of diurnal environment, while the breadth of visual search and the duration of gaze did not show significant difference in diurnal environment. 2. Dynamic clustering method is used to divide the fixation area of the driver driving into the intersection in different ways under the diurnal environment. It is found that the fixation area division will reflect the difference according to the different diurnal environment. 3. The video of the driver entering the intersection is played back frame by frame, and the frequency of the fixation target and the visual hot spot of the driver driving into the intersection in different ways under the circadian environment are obtained, and it is found that the fixation frequency is different from the visual hot spot. 4. The one-step transition probability matrix and stationary distribution of drivers at intersections are different because of the different diurnal environment.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：U491.25

【相似文献】