交通事故预测是交通安全评价、规划和决策的基础.为准确预测道路交通事故死亡人数和有效解决道路交通安全问题,以我国1991—2015年交通事故死亡人数作为研究对象,选取指数平滑法作为预测方法,并介绍指数平滑法平滑次数选取的依据、初值确定的方法,以及平滑系数的选取方法.在平滑系数的选取中,采用比经验估计法和试算法理论性更强的Levenberg-Marquardt算法,并证明该方法在平滑系数选取中的实用性.预测结果显示,2016年后我国交通事故死亡人数又有较明显的回升趋势,建议有关当局进一步加强对交通安全的重视,以避免此现象的发生.%Traffic accident prediction is the basis of traffic safety evaluation, planning and decision-making. This paper,in order to accurately predict the death toll of road traffic accidents and effectively solve the problem of road traffic safety,took the death toll of traffic accidents in China from 1991 to 2015 as the object of study,the exponential smoothing method is selected as the forecasting method. The authors introduces the basis of the selection of the smoothing times of the exponential smoothing method, the method of determining the initial value and the selection method of the smoothing coefficient. In the selection of smoothing coefficients,the Levenberg-Marquardt algorithm is adopted,which is more effective than the empirical method and the experimental algorithm, and the practicability of the method in the selection of smoothing coefficients is proved. The forecast results show that there is a significant increase in the number of traffic accidents in China after 2016,and it is suggested that the relevant authorities should pay more attention to traffic safety in order to avoid this phenomenon.
展开▼
