Please use this identifier to cite or link to this item:
Title: Robust models for large scale traffic estimation and prediction
Authors: Ma, Zunjing
Keywords: DRNTU::Engineering
Issue Date: 2014
Abstract: Urban mobility is an important driver for economic growth. However, many urban cities today are suffering from traffic congestions worldwide. To solve this, traffic prediction models are highly demanded to build Intelligent Transportation Systems (ITS) to control and reduce traffic jams. Data mining techniques and statistical models are widely used for traffic forecasting and the current researches focus on performing predictions individually for each link. However, for large-scale networks, this approach is not scalable. In this study, we will focus on recently proposed method for large-scale prediction termed as compressed prediction. In this method, the large network is first represented by a small subset of road segments and prediction is performed on those roads only. The traffic state for the entire network is then predicted by extrapolation. In this report, we study the robustness of the model different traffic conditions (rush hours, weekdays/weekends etc.). We further analyze the impact of non-nominal conditions such as accidents and road works. Analysis of networks conditions during such events can potentially lead to the development of more robust prediction algorithms. In addition, traffic incidents were evaluated to investigate the impact of incidents on the prediction model, as well as to enhance the understanding of traffic congestion and incident duration. Numerical results of this research demonstrated the accuracy and robustness of the proposed model.
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
FYP Final Report_MaZunjing.pdf
  Restricted Access
4.5 MBAdobe PDFView/Open

Page view(s)

Updated on Nov 28, 2020


Updated on Nov 28, 2020

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.