Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes

Dongjoo Park, Laurence R. Rilett, Gunhee Han

Research output: Contribution to conferencePaper

14 Citations (Scopus)

Abstract

The ability to forecast multiple period link travel times is a necessary component of real-time Route Guidance Systems (RGS). This paper examines the use of artificial neural networks (ANN) that incorporate expanded input nodes for this problem. Recent link travel time information is used as the primary input and this information is highly nonlinear which is problematic for conventional ANN. A sinusoidal transformation technique is employed for pre-mapping the input feature (i.e. the recent link travel times) in order to increase the separability of the function approximated and hence improve the prediction capabilities of the ANN. Actual link travel times from Houston, Texas that were collected as part of the automatic vehicle identification system of the Houston Transtar system were used as a test bed. The results of the best ANN with expanded input nodes were compared to a conventional ANN, a modular ANN, and other existing link travel tune forecasting models including a Kalman filtering technique, an exponential smoothing model, a historical profile, and a real-time profile. It was found that the proposed expanded input neural networks outperformed a conventional ANN and other travel time forecasting models, and gave similar results to those of a modular ANN.

Original languageEnglish
Pages325-332
Number of pages8
Publication statusPublished - 1998 Jan 1
EventProceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation - Newport Beach, CA, USA
Duration: 1998 Apr 261998 Apr 29

Other

OtherProceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation
CityNewport Beach, CA, USA
Period98/4/2698/4/29

Fingerprint

Highway systems
Travel time
Neural networks
Automatic vehicle identification

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Park, D., Rilett, L. R., & Han, G. (1998). Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes. 325-332. Paper presented at Proceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation, Newport Beach, CA, USA, .
Park, Dongjoo ; Rilett, Laurence R. ; Han, Gunhee. / Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes. Paper presented at Proceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation, Newport Beach, CA, USA, .8 p.
@conference{2a07b7a5bf4f4f668bdb79b68c1e53b0,
title = "Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes",
abstract = "The ability to forecast multiple period link travel times is a necessary component of real-time Route Guidance Systems (RGS). This paper examines the use of artificial neural networks (ANN) that incorporate expanded input nodes for this problem. Recent link travel time information is used as the primary input and this information is highly nonlinear which is problematic for conventional ANN. A sinusoidal transformation technique is employed for pre-mapping the input feature (i.e. the recent link travel times) in order to increase the separability of the function approximated and hence improve the prediction capabilities of the ANN. Actual link travel times from Houston, Texas that were collected as part of the automatic vehicle identification system of the Houston Transtar system were used as a test bed. The results of the best ANN with expanded input nodes were compared to a conventional ANN, a modular ANN, and other existing link travel tune forecasting models including a Kalman filtering technique, an exponential smoothing model, a historical profile, and a real-time profile. It was found that the proposed expanded input neural networks outperformed a conventional ANN and other travel time forecasting models, and gave similar results to those of a modular ANN.",
author = "Dongjoo Park and Rilett, {Laurence R.} and Gunhee Han",
year = "1998",
month = "1",
day = "1",
language = "English",
pages = "325--332",
note = "Proceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation ; Conference date: 26-04-1998 Through 29-04-1998",

}

Park, D, Rilett, LR & Han, G 1998, 'Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes' Paper presented at Proceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation, Newport Beach, CA, USA, 98/4/26 - 98/4/29, pp. 325-332.

Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes. / Park, Dongjoo; Rilett, Laurence R.; Han, Gunhee.

1998. 325-332 Paper presented at Proceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation, Newport Beach, CA, USA, .

Research output: Contribution to conferencePaper

TY - CONF

T1 - Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes

AU - Park, Dongjoo

AU - Rilett, Laurence R.

AU - Han, Gunhee

PY - 1998/1/1

Y1 - 1998/1/1

N2 - The ability to forecast multiple period link travel times is a necessary component of real-time Route Guidance Systems (RGS). This paper examines the use of artificial neural networks (ANN) that incorporate expanded input nodes for this problem. Recent link travel time information is used as the primary input and this information is highly nonlinear which is problematic for conventional ANN. A sinusoidal transformation technique is employed for pre-mapping the input feature (i.e. the recent link travel times) in order to increase the separability of the function approximated and hence improve the prediction capabilities of the ANN. Actual link travel times from Houston, Texas that were collected as part of the automatic vehicle identification system of the Houston Transtar system were used as a test bed. The results of the best ANN with expanded input nodes were compared to a conventional ANN, a modular ANN, and other existing link travel tune forecasting models including a Kalman filtering technique, an exponential smoothing model, a historical profile, and a real-time profile. It was found that the proposed expanded input neural networks outperformed a conventional ANN and other travel time forecasting models, and gave similar results to those of a modular ANN.

AB - The ability to forecast multiple period link travel times is a necessary component of real-time Route Guidance Systems (RGS). This paper examines the use of artificial neural networks (ANN) that incorporate expanded input nodes for this problem. Recent link travel time information is used as the primary input and this information is highly nonlinear which is problematic for conventional ANN. A sinusoidal transformation technique is employed for pre-mapping the input feature (i.e. the recent link travel times) in order to increase the separability of the function approximated and hence improve the prediction capabilities of the ANN. Actual link travel times from Houston, Texas that were collected as part of the automatic vehicle identification system of the Houston Transtar system were used as a test bed. The results of the best ANN with expanded input nodes were compared to a conventional ANN, a modular ANN, and other existing link travel tune forecasting models including a Kalman filtering technique, an exponential smoothing model, a historical profile, and a real-time profile. It was found that the proposed expanded input neural networks outperformed a conventional ANN and other travel time forecasting models, and gave similar results to those of a modular ANN.

UR - http://www.scopus.com/inward/record.url?scp=0031674181&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031674181&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:0031674181

SP - 325

EP - 332

ER -

Park D, Rilett LR, Han G. Forecasting multiple-period freeway link travel times using neural networks with expanded input nodes. 1998. Paper presented at Proceedings of the 1998 5th International Conference on Applications of Advanced Technologies in Transportation, Newport Beach, CA, USA, .