中国道路交通伤害的时间序列分析  

Time-series Analysis on Road Traffic Injury in China

作　　者：文进[1] 袁萍[2] 邓振华[3] 刘宽林[4] 张跃康[5] 刘力克[2] 孔斌[4] 黄思兴[4] 

机构地区：[1]四川大学华西医院中国循证医学中心,成都610041 [2]四川大学华西公共卫生学院流行病学教研室 [3]四川大学华西基础与法医学院病理学教研室 [4]成都市公安局交通管理局事故预防处理处 [5]四川大学华西医院脑外科

出　　处：《四川大学学报（医学版）》2005年第6期866-869,共4页Journal of Sichuan University(Medical Sciences)

基　　金：四川省科技厅重点科技项目(2000-169)资助

摘　　要：目的建立我国道路交通伤害的预测模型,以期从宏观上掌握我国道路交通伤害的发生和变化趋势,为控制我国道路交通伤害提供参考依据. 方法收集我国1951～2003年的道路交通伤害资料,进行时间序列分析,建立自回归-求和-移动平均模型(ARIMA模型).结果建立了我国道路交通伤害事故数、万车死亡率和10万人死亡率各自的ARIMA模型方程,显示预测值与实际值接近. 结论时间序列模型在道路交通伤害预测中具有较好的应用价值.Objective To establish the predictive models of road traffic injury（RTI） in China, to know the trend of RTI, and to provide the reference data for controlling RTI in China, Methods The China RTI data from 1951 to 2003 were collected, and in view of the problem of missing values, the method of intrapolation was adopted. The Box-Jenkins technique was used to analyze and predict the trend of RTI in China. Following the process for stationary time-series analysis, model identification, parameter estimation and model diagnosis, the predictive equation for RTI would be established. Results A series of predictive equations on RTI were finally established based on ARIMA models. The curve fitting is effective and the predictive data of RTI in 2003 are close to the true statistical data. Conclusion The time-series model thus established proves to be of significant usefulness in RTI prediction.

关 键 词：道路交通伤害 时间序列分析 自回归-求和-移动平均模型 

分 类 号：U491.3[交通运输工程—交通运输规划与管理]