道路交通事故次数组合预测模型

The combined prediction model on road traffic accidents

道路交通事故次数预测对于掌握未来交通安全状况,合理评价交通安全措施的可行性和实施效果具有十分重要的意义。利用我国道路交通事故次数的统计数据,分别采用灰色预测模型和历史序列拟合分析的方法建立事故次数的预测模型,针对两种模型的优缺点,采用相对误差倒数分配权重的方法建立组合预测模型,计算结果表明,组合预测模型在预测事故次数方面具有较高的稳定性和较好的预测精度,完全能够满足事故次数预测的要求。

It is of the significance for the road traffic to predict the traffic safety situation in the future, and to evaluate the feasibility and implementation of traffic safety measures. Based on the statistics on the road traffic accidents in China, a gray prediction model and a historical sequence fitting analysis method are used to establish the road traffic prediction model. According to the advantages and disadvantages of the two models, the eountdown allocation weights method of relative error is added to establish the combined prediction model. The results calculated show that the combined prediction model has higher stability and better prediction accuracy in the traffic prediction, meeting the traffic prediction requirements.