基于CREAM扩展法的信号交叉口驾驶差错概率预测  

Prediction of Driving Error Probability at Signalized Intersection Based on Extended CREAM Method

作　　者：陈美如[1] 裴玉龙[2] 

机构地区：[1]哈尔滨工业大学交通科学与工程学院,哈尔滨150090 [2]东北林业大学交通学院,哈尔滨150040

出　　处：《武汉理工大学学报（交通科学与工程版）》2017年第3期375-378,384,共5页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：国家自然科学基金项目资助(51078113)

摘　　要：通过认知可靠性与差错分析方法扩展法计算驾驶差错概率,综合考虑了影响驾驶差错的外部因素和内部因素.建立信号交叉口驾驶事件序列,识别每个事件涉及到的认知活动;确定影响驾驶差错概率的共同绩效因子CPCs和各水平因子对认知活动的影响系数;根据交叉口的CPCs水平,确定对应水平的影响系数和可能的驾驶失效模式,计算驾驶差错概率基本值,并对其进行修正得到修正值.根据驾驶事件之间的关系,计算整个驾驶任务的驾驶差错概率.Driving error probability （DEP） is calculated by the extended CREAM method which consid- ers both the exterior and interior influence factors of driving errors. Firstly, the sequence of driving e- vents at signalized intersections is established and cognitive activities that each step of the event se- quence is involved are identified. Next the CPCs that affect DEP are proposed and the impact factors at all levels are calculated. According to the CPCs levels of the intersection, credible driving error modes are identified and the total influences of CPCs on cognitive function errors are assessed. Then, the nominal DEP of each step is calculated and revised by impact factors to gain adjusted DEPs. Final- ly, in accordance with the relationships of driving events, DEP of the whole driving task is calculated.

关 键 词：驾驶差错概率 CREAM扩展法 共同绩效因子 认知活动 失效模式 

分 类 号：U491.31[交通运输工程—交通运输规划与管理]