摘要
为实现险性驾驶行为的有效辨识,以事故发生概率为依据,在对驾驶行为状态予以险态分级的前提下,采用单因子方差分析进行险态辨识主因子析取,并基于模糊C均值聚类法构建了险态行为辨识模型。最后,结合实例对模型予以试算,并采用样本回代对模型进行了验证。结果表明,模型误差率为2.5%。
According to the traffic accident occurence probability,the risk status of the driving behavior was classified,the main factor of the risk status idendification was extracted using the single factor variance analysis,a driving behavior risk status identification model was built based on the FCM to identify effectively the risky driving behavior.The model was used on trial with examples and tested by sample re-substitution.The result indicats that the relative error of the model is 2.5%.
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2010年第2期427-430,共4页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金项目(60879022
70422201
60870005)
西南交大青年教师科研起步项目(2009Q038)
关键词
交通运输安全工程
险态辨识
C均值聚类
驾驶行为
engineering of communications and transportation safety
identification of risk status
fuzzy C-means（FCM）
driving behavior
