论文部分内容阅读
为提高交通事故预测的可靠性,首先应用有序聚类的方法建立交通受伤人数的分级标准;然后针对事故受伤人数为相依随机变量的特点,采取以规范化的各阶自相关系数为权重,用加权的马尔可夫链模型来预测未来交通事故受伤人数的升降变化状况;最后以北京市1970-2010年共41年的事故受伤人数为例对该方法进行了具体应用,取得了较为满意的结果,为交通事故受伤人数的预测分析提供了新的途径.
In order to improve the reliability of traffic accident prediction, firstly, an orderly clustering method is used to establish the grading standard of traffic injuries. Secondly, according to the characteristics that the accident victims are dependent random variables, taking the normalized autocorrelation coefficient of each order as the weight, Weighted Markov chain model to predict the rise and fall of the number of traffic accidents in the future; Finally, the method was specifically applied to the case of accidental injuries in Beijing for a total of 41 years from 1970 to 2010, and satisfactory results were obtained , Which provides a new way for predicting the number of injured persons in traffic accidents.