基于多维数据的高速公路服务水平实时判别方法

A Real-time Identification Method of Highway Service Level Based on Multi-dimension Data

随着高速公路的不断发展,交通信息逐渐呈现多元化,而将这些信息进行整合并对高速公路服务水平进行实时判别,对于后期的交通控制管理及实时高效提供道路信息至关重要.本文考虑了时间、天气、出行日期等多因素的影响,选用车速、密度作为高速公路服务水平评价指标,结合最小二乘支持向量机与聚类分析相关算法,提出了一种基于多维数据的高速公路服务水平实时判别模型,以实现对高速公路服务水平的参数预测及实时判别.选用河北省237 360组训练样本数据和6 048组预测样本数据验证了该模型的有效性.结果显示,该模型获得了较高的判别精度和较好的预测效果,是一种有效的高速公路服务水平实时判别方法.

With the continuous development of highway,traffic information of highway is diversified gradually.Integration of the traffic of highway and real-time identification of the service level of highway is very important for the later traffic controlling and management and provision of real-time and efficient road information.Combing least squares support vector machine with cluster analysis algorithm,this paper proposes a real-time identification model of highway service level based on multi-dimension data in order to achieve the parameter prediction and real-time identification of highway service level.The model includes many influence factors,such as time,weather,travel dates and so on.In addition,it selects vehicle speed and density to evaluate highway service level.The validity of the model is tested by using 237 360 training samples and 6 048 prediction samples of Hebei province.The results show that the model produces the higher identification accuracy and better prediction effect,which is a method for real-time identification of highway service level.