计算机工程与应用 ›› 2014, Vol. 50 ›› Issue (5): 247-251.

• 工程与应用 • 上一篇    下一篇

适应交通流演化的伽马分布形状参数估计

王晓原,张敬磊,马立云   

  1. 山东理工大学 交通与车辆工程学院,山东 淄博 255049
  • 出版日期:2014-03-01 发布日期:2015-05-12

Estimation of gamma distribution shape parameter adapting to traffic flow evolvement

WANG Xiaoyuan, ZHANG Jinglei, MA Liyun   

  1. School of Transportation and Vehicle Engineering, Shandong University of Technology, Zibo, Shandong 255049, China
  • Online:2014-03-01 Published:2015-05-12

摘要: 研究交通流演化规律对交通安全措施的实施具有重要意义。交通流的演化过程一般可分为交通流的量变与质变。鉴于交通自由流状态、拥挤流状态及间歇流状态的分布均可由伽马分布表达,利用伽马分布及其变点检验算法进行变点搜索及检验,确定变点前后的交通流状态。结合柯尔莫哥洛夫—斯米尔诺夫检验方法对观测数据是否服从伽马分布进行拟合优度检验。为较直观说明交通流的演化规律,结合极大似然估计,给出了适应交通流演化的伽马分布形状参数估计。实验结果表明,伽马分布形状参数估计的研究是探索交通流演化规律的有利途径。

关键词: 交通流, 伽马分布, 柯尔莫哥洛夫&mdash, 斯米尔诺夫检验, 极大似然估计, 演化规律

Abstract: It is significant to research traffic flow evolvement for the traffic safety implementation. The evolution process of traffic flow is generally classified as quantitative change and qualitative change. Distributions of free, congested and intermittent flow all can be replaced by gamma distribution. So, in this paper, the traffic flow states before and after change-point are confirmed by searching and testing change-point with gamma distribution and change-point testing method. Kolmogorov-Smirnov test method is used to test whether or not the observation data is in line with gamma distribution. Gamma distribution shape parameter adapting to traffic flow evolvement is estimated with maximum likely estimation method in order to explain evolvement directly. The experimental results indicate that studying gamma distribution shape parameter is a favorable way to research traffic flow evolvement.

Key words: traffic flow, gamma distribution, Kolmogorov-Smirnov test, maximum likely estimation, evolvement