基于ARIMA-BP组合模型的铁路行车事故数预测  

Forecast of Railway Transportation Accidents Based on ARIMA-BP Combined Model

作　　者：徐国权[1] 张佳慧 况坚 XU GUO-quan;ZHANG Jia-hui;KUANG Jian(School of Traffic and Transportation Engineering,East China Jiaotong University,Nanchang 330013,China)

机构地区：[1]华东交通大学交通运输工程学院,江西南昌330013

出　　处：《辽宁工业大学学报（自然科学版）》2023年第3期174-179,共6页Journal of Liaoning University of Technology(Natural Science Edition)

基　　金：国家社会科学基金项目(17BJY140)。

摘　　要：为更精确预测铁路行车事故数,在ARIMA模型与BP神经网络模型的基础上,利用ARIMA模型分析铁路行车事故数的线性部分;利用BP神经网络分析影响铁路行车事故数的非线性部分,如设备状况、管理状况、运输量等,构建了ARIMA-BP神经网络拉格朗日乘数法加权组合预测模型和ARIMA-BP残差修正组合模型,并对4种模型的预测精度进行比较。研究表明,ARIMA-BP神经网络残差修正组合模型预测精度最高,可为铁路部门了解事故发生趋势、有效预防事故、合理制定对策提供一定参考。In order to more accurately predict the number of railway traffic accidents,on the basis of the ARIMA model and the BP neural network model,the ARIMA model is used to analyze the linear part of the number of railway traffic accidents;the BP neural network is used to analyze the nonlinear part that affects the number of railway traffic accidents,such as equipment condition,management status,transportation volume,etc.,the ARIMA-BP neural network lagrange multiplier method weighted combination prediction model and ARIMA-BP residual correction combination model are constructed.The prediction accuracy of the four models is compared.The results show that the ARIMA-BP neural network residual correction combined model has the highest prediction accuracy.This research can provide a reference for railway departments to understand the trend of accidents,effectively prevent accidents,and formulate reasonable countermeasures.

关 键 词：铁路行车事故数 拉格朗日乘数法 残差修正 组合预测 

分 类 号：U298[交通运输工程—交通运输规划与管理]