计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (31): 11-15.

• 博士论坛 • 上一篇    下一篇

基于灰色加权马尔可夫SCGM(1,1)c的交通事故预测

赵  玲1,2,许宏科1   

  1. 1.长安大学 电子与控制工程学院,西安 710064
    2.西安邮电大学 通信与信息工程学院,西安 710121
  • 出版日期:2012-11-01 发布日期:2012-10-30

Traffic accident prediction based on gray weighted Markov SCGM(1,1)c

ZHAO Ling1,2,XU Hongke1   

  1. 1.School of Electronic and Control Engineering, Chang’an University, Xi’an 710064, China
    2.School of Communication and Information Engineering, Xi’an University of Posts and Telecommunications, Xi’an 710121, China
  • Online:2012-11-01 Published:2012-10-30

摘要: 交通事故预测是交通安全评价、规划和决策的基础。基于灰色系统理论和马尔可夫链理论,应用系统云灰色模型SCGM(1,1)c拟合道路交通时序数据的总体趋势,所得拟合指标是随机波动的。马尔可夫链原理适合处理波动性大的系统过程,因此选用能更好解决随机波动性的加权马尔可夫链预测方法,提出一种用于道路交通事故次数预测的灰色加权马尔可夫SCGM(1,1)c模型,它适用于时间序列短,数据量少且随机波动不太大的动态过程预测。以某市1975—2010年道路交通事故次数为例进行了预测分析,结果表明该模型既能揭示交通事故次数变化的总体趋势,又能克服随机波动性数据对预测精度的影响,具有较强的工程实用性。

关键词: 交通安全, 交通事故预测, 单因子系统云灰色模型, 加权马尔可夫链, 随机波动

Abstract: The prediction of traffic accident is the basis of transportation safety assessment, planning and decision-making. According to grey system theory and Markov chain principle, applying a single factor system cloud grey SCGM(1,1)c model to fit the tendency of the road traffic time series, its fitting index is random fluctuation. Markov chain method is suitable for forecasting stochastic fluctuating dynamic process, selecting weight Markov chain to predict the fitting index. Combining the advantages of two models, found a weighted SCGM(1,1)c model for road traffic accident frequency prediction, the new model is suitable for forecasting such kinds of system with short time, few data and not too large random fluctuation. Finally, the new model is applied to predict the traffic accident times of Beijing from 1975 to 2010. The results show that the new model not only discovers the trend of the traffic accident time but also overcomes the random fluctuation data of affecting precision accuracy, having a strong engineering applicability.

Key words: transportation security, traffic accident prediction, SCGM(1, 1)c, weighted Markov chain, random fluctuation