Should we use neural networks or statistical models for short-term motorway traffic forecasting?

被引：146

作者：

Kirby, HR

Watson, SM

Dougherty, MS

机构：

[1] Institute for Transport Studies, University of Leeds, Leeds

[2] Ctr. for Res. on Transp. and Society, Högskolan Dalarna

来源：

INTERNATIONAL JOURNAL OF FORECASTING | 1997年 / 13卷 / 01期

基金：

英国工程与自然科学研究理事会;

关键词：

traffic forecasting; ARIMA models; neural networks; ATHENA model; motorway flows;

D O I：

10.1016/S0169-2070(96)00699-1

中图分类号：

F [经济];

学科分类号：

02 ;

摘要：

This article discusses the relative merits of neural networks and time series methods for traffic forecasting and summarises the findings from a comparative study of their performance for motorway traffic in France. Whilst it was possible to get a good performance with both neural networks and traditional Auto-Regressive Integrated Moving Average (ARIMA) models when forecasting up to an hour ahead using data supplied in 30-min intervals, a purpose-built pattern based forecasting model known as ATHENA, developed by INRETS, out-performed both these methods somewhat. The ways in which these models relate to the structure of traffic data are discussed and alternative paradigms are proposed.

引用

页码：43 / 50

页数：8