摘要
为了客观、有效的评价交通拥堵的程度,采用速度指标作为反映交通流状态特征的关键变量,针对交通流速度变量存在混合分布的特点,建立了基于高斯混合分布的交通拥堵评价模型,并将EM算法运用于模型求解,通过对速度变量的聚类分析,以判别交通状态,评价交通拥堵程度。上海市快速高架路实测数据应用模型分析的结果表明:模型可以客观分析交通流的混合分布,有效评价交通拥堵。
In order to evaluate traffic congestion more objectively and effectively,velocity was employed as the key independent variable for the feature of traffic flow state,the evaluation model of traffic congestion based on Gaussian mixture distribution was developed for the reason that velocity of traffic flow shows the mixture distribution performance.The EM algorithm was used to this model for the analysis of clustering velocity variable.Depend on the result of model solution,the factual traffic flow state was estimated and the traffic congestion degree was evaluated better.The analytic result of measured data of Shanghai expressway shows that this model could analyze the traffic flow mixture distribution objectively and the evaluation of traffic congestion is effective.
出处
《公路交通科技》
CAS
CSCD
北大核心
2011年第2期127-132,共6页
Journal of Highway and Transportation Research and Development
关键词
交通工程
评价模型
高斯混合分布
交通拥堵
EM算法
traffic engineering
evaluation model
Gaussian mixture distribution
traffic congestion
EM algorithm
