摘要
为预防人为因素引起的交通事故,从驾驶员行为可靠性角度研究交通安全行为风险。利用加权综合评价法,选取风险指标,并建立安全行为风险度量模型。再采用模糊聚类算法(FCM),结合驾驶员可靠性测试数据,划分风险等级,并界定合理的风险告警阈值,避免人为设定风险告警阈值的主观性。结果表明,不良的情绪和生理状态、侥幸的违章心理是引起人为交通事故的主要因素。基于交通安全风险度量模型的告警阈值合理,可避免因阈值设置过松或过严造成的漏警和虚警问题。
To prevent traffic accidents caused by human factors, driversj traffic behavior reliability risk research were studied. Risk indicators were selected using weighting comprehensive evaluation method. A model was built for measuring the risks. The risks were classified into levels using the fuzzy c-means meth- od according to measured data on the drivers reliability. Risk warning threshold values were defined. The results show that the main factors in man-made traffic accidents are the fluke mind of traffic violation and the bad mood or poor physiological state, and that the risk warning threshold values based on the model are reasonable and contribute to reducing both false positive and negative warning caused by loose or strict threshold setting.
出处
《中国安全科学学报》
CAS
CSCD
北大核心
2013年第6期103-109,共7页
China Safety Science Journal
关键词
道路交通
交通事故
驾驶行为失误率
可靠性
安全行为风险
风险告警阈值
road traffic
traffic accident
error rate of drivers＇ behavior
reliability
safety behavior risk
risk warning threshold
作者简介
郭洪洋（1982-），男，辽宁沈阳人，西南交通大学交通运输学院博士生研究生，专业方向为交通运输规划与管理。E-mail：hongyang．guo@163．com．
