徐蒙蒙++严凌

摘 要：自2012年国庆节以来国家实施了高速公路对7座以下含7座的小客车免费政策后，高速公路交通量巨大，导致了一系列问题，如：拥堵、事故等。因此事先做好交通量预测工作很重要，预测精度显得尤为重要，为高速公路管理部门提前做好交通流组织和预防准备工作。文章利用matlab软件在传统GM1，1模型的基础上稍作改进建立了修正灰色马尔科夫模型，再对残差绝对值序列进行滑动平均处理建立无偏的残差绝对值GM1，1模型，再结合马尔科夫链过程将残差符号划分为3类，利用马尔科夫转移矩阵来判定未来残差的符号的概率和残差状态，最终得到完整的修正预测模型。通过对比传统GM1，1模型与修正模型，发现修正模型的精度更高，具有很大的可信度。对北京市高速公路近几年的历史数据利用已建立的灰色马尔科夫预测模型，分别预测了4个节假日下一年的数据。此种方法可为高速公路管理部门提供数据支持，同样适用于其它地区免费节假日时高速公路交通量的预测。

关键词：交通量预测；灰色马尔科夫；免费节假日；高速公路

中图分类号：F570 文献标识码：A

Abstract： Since the National Day in 2012 since the implementation of the national highway on the 7 following 7 small passenger car free policy， the highway traffic volume is huge， leading to a series of problems， such as： congestion， accidents and so on. Therefore， it is very important to do a good job in forecasting the traffic volume in advance. The accuracy of forecasting is very important， and it is necessary for the highway management department to prepare the traffic flow and prevent the preparation in advance. In this paper， a modified gray Markov model is established based on the traditional GM1，1 model， and then the residual error of the residual absolute value sequence is established. The absolute value GM1，1 model， combined with the Markov chain process to divide the residual symbols into three categories， using the Markov transfer matrix to determine the probability and residual state of symbols for future residuals， and finally get a complete correction prediction model. By comparing the traditional GM1，1 model with the modified model， it is found that the modified model is more accurate and has great credibility. The historical data of the expressway in Beijing in recent years have been forecasted by the gray Markov forecasting model which has been established， and the data of the four holidays are predicted respectively. This method provides data support for expressway management， and is also applicable to the forecast of expressway traffic at free holidays in other areas.

Key words： traffic forecast； gray Markov； holidays and festivals； highway

0 引 言

交通量預测是进行交通量现状评价、综合分析建设项目的必要性和可行性的基础，是确定公路建设项目的技术等级、工程规模的主要依据。还是提高交通运输管理水平、降低运输成本的重要手段之一。周荣康等[1]人基于灰色残差GM（1，1）模型预测了道路交通量，考虑应用残差GM1，1模型对原有的预测值进行修正，用于交通量预测中，结果表明，方法是可行的；……