Modelling of traffic control delays at priority junctions using artificial neural network

Traffic delay is an essential aspect taken into consideration in the evaluation of operational performance of priority junctions. Delay is typically described as the excess time taken in a transportation facility in comparison to that of a reference value. Although, there are several methods availab...

Full description

Saved in:

Bibliographic Details
Main Author:	Sahraei, Mohammad Ali
Format:	Thesis
Language:	English
Published:	2018
Subjects:	TA Engineering (General) Civil engineering (General)
Online Access:	http://eprints.utm.my/id/eprint/79143/1/MohammadAliSahraeiPFKA2018.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my-utm-ep.79143
record_format	uketd_dc
spelling	my-utm-ep.791432018-09-30T08:23:17Z Modelling of traffic control delays at priority junctions using artificial neural network 2018 Sahraei, Mohammad Ali TA Engineering (General). Civil engineering (General) Traffic delay is an essential aspect taken into consideration in the evaluation of operational performance of priority junctions. Delay is typically described as the excess time taken in a transportation facility in comparison to that of a reference value. Although, there are several methods available for the estimation of traffic control delay, they can lead to different results. A comparative analysis for the estimation of the control delay using the American highway capacity manual and the Malaysian highway capacity manual showed that the theoretical models are not consistent with actual delays observed from sites. This implies that both theoretical models are not directly capable of analysing control delay at priority junctions in Malaysia. This study was carried out to model traffic control delays at priority junctions using Artificial Neural Network (ANN). In this study, data were sampled from eight priority junctions of various configurations. Data pertaining to the analysis of critical gap, follow-up time, and control delay were collected using video camera recording technique. The study was divided into two phases comprising analysis of field data, and the development of ANN and mathematical models using MATLAB software. In the course of data analysis, the research recognized and estimated various variables that influence control delay. To generate the model, an ANN with two hidden layers and several sizes of neurons in the hidden layers were developed. Several mathematical models for estimation of control delay with a reasonable accuracy were developed using the outputs from the ANN model. Findings from this research showed that the range of conflicting flow is from 130 to 2470 veh/h and 120 to 2300 pcu/h, the values of control delays predicted are 3-37 sec/veh and 4-43 sec/pcu, respectively. Accordingly, the minimum and maximum values of traffic control delay occurred for both left- and right-turning vehicles from the minor roads. The modelling results showed that the values of control delay for right-turning manoeuvre from minor road at junction with four lanes major/two lanes minor road were higher than other junctions. This is due to queue delays and stops delay behind the stop line, in order to select an appropriate gap on the major road in the far and near side. Delay values for right-turning manoeuvre from major road at junction with four lanes major/four lanes minor road were greater than other junctions. The analysis revealed that heavy vehicles had the lowest effect on the proposed models, with an increase from 10% to 50%, resulting in the values of control delay to increase from 1% to 3%. On the contrary, the movement flow and conflicting flow had the highest impact, with an increase from 10% to 50% whereby the control delay could increase to 44%. The statistical analyses revealed that the delay estimated using the formula acquired from the ANN model and those from the field studies are equal. 2018 Thesis http://eprints.utm.my/id/eprint/79143/ http://eprints.utm.my/id/eprint/79143/1/MohammadAliSahraeiPFKA2018.pdf application/pdf en public phd doctoral Universiti Teknologi Malaysia, Faculty of Civil Engineering Faculty of Civil Engineering
institution	Universiti Teknologi Malaysia
collection	UTM Institutional Repository
language	English
topic	TA Engineering (General) Civil engineering (General)
spellingShingle	TA Engineering (General) Civil engineering (General) Sahraei, Mohammad Ali Modelling of traffic control delays at priority junctions using artificial neural network
description	Traffic delay is an essential aspect taken into consideration in the evaluation of operational performance of priority junctions. Delay is typically described as the excess time taken in a transportation facility in comparison to that of a reference value. Although, there are several methods available for the estimation of traffic control delay, they can lead to different results. A comparative analysis for the estimation of the control delay using the American highway capacity manual and the Malaysian highway capacity manual showed that the theoretical models are not consistent with actual delays observed from sites. This implies that both theoretical models are not directly capable of analysing control delay at priority junctions in Malaysia. This study was carried out to model traffic control delays at priority junctions using Artificial Neural Network (ANN). In this study, data were sampled from eight priority junctions of various configurations. Data pertaining to the analysis of critical gap, follow-up time, and control delay were collected using video camera recording technique. The study was divided into two phases comprising analysis of field data, and the development of ANN and mathematical models using MATLAB software. In the course of data analysis, the research recognized and estimated various variables that influence control delay. To generate the model, an ANN with two hidden layers and several sizes of neurons in the hidden layers were developed. Several mathematical models for estimation of control delay with a reasonable accuracy were developed using the outputs from the ANN model. Findings from this research showed that the range of conflicting flow is from 130 to 2470 veh/h and 120 to 2300 pcu/h, the values of control delays predicted are 3-37 sec/veh and 4-43 sec/pcu, respectively. Accordingly, the minimum and maximum values of traffic control delay occurred for both left- and right-turning vehicles from the minor roads. The modelling results showed that the values of control delay for right-turning manoeuvre from minor road at junction with four lanes major/two lanes minor road were higher than other junctions. This is due to queue delays and stops delay behind the stop line, in order to select an appropriate gap on the major road in the far and near side. Delay values for right-turning manoeuvre from major road at junction with four lanes major/four lanes minor road were greater than other junctions. The analysis revealed that heavy vehicles had the lowest effect on the proposed models, with an increase from 10% to 50%, resulting in the values of control delay to increase from 1% to 3%. On the contrary, the movement flow and conflicting flow had the highest impact, with an increase from 10% to 50% whereby the control delay could increase to 44%. The statistical analyses revealed that the delay estimated using the formula acquired from the ANN model and those from the field studies are equal.
format	Thesis
qualification_name	Doctor of Philosophy (PhD.)
qualification_level	Doctorate
author	Sahraei, Mohammad Ali
author_facet	Sahraei, Mohammad Ali
author_sort	Sahraei, Mohammad Ali
title	Modelling of traffic control delays at priority junctions using artificial neural network
title_short	Modelling of traffic control delays at priority junctions using artificial neural network
title_full	Modelling of traffic control delays at priority junctions using artificial neural network
title_fullStr	Modelling of traffic control delays at priority junctions using artificial neural network
title_full_unstemmed	Modelling of traffic control delays at priority junctions using artificial neural network
title_sort	modelling of traffic control delays at priority junctions using artificial neural network
granting_institution	Universiti Teknologi Malaysia, Faculty of Civil Engineering
granting_department	Faculty of Civil Engineering
publishDate	2018
url	http://eprints.utm.my/id/eprint/79143/1/MohammadAliSahraeiPFKA2018.pdf
_version_	1747818156838092800

Modelling of traffic control delays at priority junctions using artificial neural network

Similar Items