基于灰色系统理论的道路交通事故特征与预测方法分析

以我国道路交通事故为研究对象,运用趋势图分析道路交通事故变化规律,在此基础上选择原始数据为4个的新陈代谢GM(1,1)模型和N-Verhulst模型进行道路交通事故的预测。结果表明,原始数据长度为4的新陈代谢GM(1,1)模型对道路交通事故起数、死亡人数、万车死亡率均有较高预测精度,N-Verhulst模型对万车死亡率和死亡人数有较高预测精度。

Taking the road traffic accidents in China as the research object,the trend chart is used to analyze the change law of road traffic accidents.On this basis,the metabolic GM model and N-Verhulst model with four original data are selected to predict road traffic accidents.The results show that the metabolic GM(1,1) model with the original data length of 4 has high prediction accuracy for the number of road traffic accidents,the number of deaths and the death rate of ten thousand vehicles,while the N-Verhulst model has higher prediction accuracy for the death rate of ten thousand vehicles and the number of deaths.