悉尼自适应交通控制系统线圈数据短时多步预测双层模型  

Bi-level model of multi-step forecasting for short-term data of loop in Sydney coordinated adaptive traffic system

作　　者：李琦[1,2] 姜桂艳[3] 

机构地区：[1]吉林大学交通学院,长春130022 [2]青岛市城市规划设计研究院,山东青岛266011 [3]宁波大学海运学院,浙江宁波315211

出　　处：《吉林大学学报（工学版）》2013年第4期854-860,共7页Journal of Jilin University:Engineering and Technology Edition

基　　金：国家自然科学基金项目(51278257);高等学校博士学科点专项科研基金项目(20110061110034);浙江省自然科学基金项目(LY12F01013)

摘　　要：为了进一步改善悉尼自适应交通控制系统(Sydney coordinated adaptive traffic system,SCATS)线圈数据短时多步预测的效果,在对SCATS线圈数据进行预处理的基础上,设计了一种基于动态神经网络的短时多步预测双层模型,包括基于NARX(Nonlinear autoregressive model with exogenous inputs)神经网络的多步预测方法以及基于FTD(Focused time-delay)神经网络的可预测步数在线估计方法,并采用某特大城市SCATS线圈实测数据进行了验证和对比分析。结果表明:本文方法能够进一步降低SCATS线圈数据短时多步预测的误差。In order to improve the effect of multi-step forecasting for short term traffic data collected from the loop in Sydney Coordinated Adaptive Traffic System （SCATS）, on the basis of data preprocessing, a biqevel model of multi-step forecasting using dynamic neural networks was designed. This model includes a multi-step forecasting method based on Nonlinear Autoregressive model with exogenous inputs （NARX） neural network and a predictable steps online estimation method based on Focused Time-Delay （FTD） neural network. Validation and comparative analysis were carried out using data of loop in SCATS measured from a megacity. The results indicate that the proposed method can further reduce the errors of short-term traffic multi-step forecasting in SCATS.

关 键 词：交通运输系统工程 悉尼自适应交通控制系统 短时交通预测 动态神经网络 多步预测 

分 类 号：U491[交通运输工程—交通运输规划与管理]