虚拟现实驾驶仿真环境下的驾驶分心研究

发布时间：2018-04-05 02:08

本文选题：交通安全　切入点：视觉分心　出处：《北方工业大学》2017年硕士论文

【摘要】：调研数据显示,驾驶分心是引发交通事故的主要原因之一。随着车载智能设备的增加和日益复杂的交通状况,占用了驾驶员越来越多的注意力资源,容易引发驾驶分心现象,从而导致驾驶员感知、决策能力降低,甚至造成驾驶失误等一系列不安全驾驶行为,因此驾驶分心已经成为交通安全领域重要研究问题。本文根据人眼瞳孔的变化特征与自主神经相互关系的医学基础,利用眼动测量法记录驾驶员眼动信息,选定眼睛本身指标—瞳孔直径进行视觉分心和认知分心的识别方法研究.首先,在对国内外研究成果广泛调研基础上,开展了预实验和预处理。课题从驾驶员眼动角度开展研究,确定研究内容和思路。通过预实验选择了合适的实验被试、有效的研究指标和适宜的环境,为正式实验的开做铺垫。原始数据的处理过程分为两步,第一步剔除两眼瞳孔直径绝对差值大于阈值的数据,第二步利用滑动标准差查找异常数据,结合线性插值法对异常数据进行修复,并取得了较好的处理效果。其次,课题设计了视觉分心实验方案,通过设置标识牌使被试发生驾驶状态改变,每位实验被试根据提示先后进行跟驰驾驶、超车驾驶和视觉分心驾驶。实验结束后提取三种驾驶状态下瞳孔直径信息,利用方差分析从数值上进行视觉分心识别方法研究,均方根分析和递归图分析分别从能量和图像上可进行视觉分心状态的观测。接下来利用车道中心偏离距离进行驾驶安全性分析,发现视觉分心状态下存在驾驶安全问题。最后,课题设计了认知分心实验方案,通过询问计算题的方式使被试发生认知分心,并利用包络分析观测认知分心水平变化。结果表明,认知分心过程中瞳孔直径逐渐升高,并随着认知分心结束而回归正常水平。随后对认知分心状态进行驾驶安全性分析,选用方向盘旋转率指标,通过与正常驾驶状态进行对比,发现认知分心状态下方向盘旋转率变化平缓,表明此时横向驾驶绩效降低。
[Abstract]:Research data show that driving distraction is one of the main causes of traffic accidents.With the increase of vehicle intelligent devices and the increasingly complex traffic conditions, drivers are occupying more and more attention resources, which can easily lead to driving distractions, thus leading to the driver's perception and decision-making ability.Even causes a series of unsafe driving behavior such as driving error, so driving distraction has become an important research problem in the field of traffic safety.Based on the medical basis of the relationship between the characteristics of eye pupil change and autonomic nerve, the eye movement information of driver was recorded by eye movement measurement method, and the visual distraction and cognitive distraction were studied by selecting the index of eye itself-pupil diameter.Firstly, on the basis of extensive research results at home and abroad, pre-experiment and pre-treatment were carried out.The research is carried out from the angle of driver's eye movement, and the research content and train of thought are determined.The suitable experimental subjects were selected by pre-experiment, the effective research index and the suitable environment were selected, which paved the way for the formal experiment.The process of processing the original data is divided into two steps. The first step is to eliminate the data whose absolute difference of pupil diameter is greater than the threshold. The second step is to search the abnormal data by sliding standard deviation, and to repair the abnormal data with linear interpolation method.Good results have been obtained.Secondly, a visual distraction experiment scheme is designed, which changes the driving state of the subjects by setting up a signboard. Each subject carries on following driving, overtaking driving and visual distracted driving according to the prompts.At the end of the experiment, the information of pupil diameter in three driving states was extracted, the visual distraction recognition method was studied numerically by variance analysis, and the visual distraction state could be observed from energy and image by root mean square analysis and recursive graph analysis, respectively.Then the driving safety analysis is carried out by using the driveway center deviation distance, and it is found that there are driving safety problems in visual distraction.Finally, a cognitive distraction experiment scheme was designed, in which cognitive distraction was induced by asking and calculating questions, and the change of cognitive distraction level was observed by envelope analysis.The results showed that the pupil diameter increased gradually during cognitive distraction and returned to normal level with the end of cognitive distraction.Then the driving safety of cognitive distraction state was analyzed and the steering wheel rotation rate index was selected. By comparing with the normal driving state, it was found that the steering wheel rotation rate changed slowly under cognitive distraction state, which indicated that the lateral driving performance decreased.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.25;TP391.9

【参考文献】