Traffic flow prediction model based on neighbouring roads using neural network and multiple regression

Bagus Priambodo, Azlina Ahmad

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Monitoring and understanding traffic congestion seems difficult due to its complex nature. This is because the occurrence of traffic congestion is dynamic and interrelated and it depends on many factors. Traffic congestion can also propagate from one road to neighbouring roads. Recent research shows that there is a spatial correlation between neighbouring roads with different traffic flow pattern on weekdays and on weekends. Previously, prediction of traffic flow propagation was based on day and time during weekdays and on weekends. Results obtained from past studies show that further investigation is needed to reduce errors using a more efficient method. We observed from previous research that similarity of traffic condition on weekdays and weekends was not taken into account in predicting traffic flow propagation. Hence, our study is to create and evaluate a new prediction model for traffic flow propagation at neighbouring roads using similarity of traffic flow pattern on weekdays and weekends to achieve more accurate results. We exploit similarity of traffic flow pattern on weekdays and weekends by adding time cluster in our proposed model. Thus, our neural network model proposed high correlation road, time and day clusters as input factors in neural network model prediction. Our initial phase of the methodology involves investigation on correlation between neighbouring roads. This paper discusses the results of experiments we have conducted to determine relationship between roads in a neighbouring area and to determine input factors for our neural network traffic flow prediction model. To choose a particular road as a predicting factor, we calculated the distance between roads in neighbouring area to identify the nearest road. Then, we calculated correlation based on traffic condition (congestion) between roads in neighbouring area. The results were then used as input factors for prediction of traffic flow. We compared the results of the experiment using neural network without cluster parameters and multiple regression methods. We observed that neural network with time cluster parameter produced better results compared to neural network without parameter and multiple regression method in predicting average speed of vehicles on neighbouring roads.

Original languageEnglish
Pages (from-to)513-535
Number of pages23
JournalJournal of Information and Communication Technology
Volume17
Issue number4
Publication statusPublished - 1 Jan 2018

Fingerprint

Road Network
Multiple Regression
Traffic Flow
Prediction Model
Neural Networks
Model-based
Neural networks
Traffic congestion
Flow patterns
Traffic Congestion
Flow Pattern
Propagation
Neural Network Model
Prediction
Traffic
Network Flow
Spatial Correlation
Network Traffic
Experiments
Congestion

Keywords

  • Multiple regression method
  • Neural network
  • Traffic congestion
  • Traffic prediction

ASJC Scopus subject areas

  • Computer Science(all)
  • Mathematics(all)

Cite this

Traffic flow prediction model based on neighbouring roads using neural network and multiple regression. / Priambodo, Bagus; Ahmad, Azlina.

In: Journal of Information and Communication Technology, Vol. 17, No. 4, 01.01.2018, p. 513-535.

Research output: Contribution to journalArticle

@article{21d2c5475f9c45388fddda3c6e858411,
title = "Traffic flow prediction model based on neighbouring roads using neural network and multiple regression",
abstract = "Monitoring and understanding traffic congestion seems difficult due to its complex nature. This is because the occurrence of traffic congestion is dynamic and interrelated and it depends on many factors. Traffic congestion can also propagate from one road to neighbouring roads. Recent research shows that there is a spatial correlation between neighbouring roads with different traffic flow pattern on weekdays and on weekends. Previously, prediction of traffic flow propagation was based on day and time during weekdays and on weekends. Results obtained from past studies show that further investigation is needed to reduce errors using a more efficient method. We observed from previous research that similarity of traffic condition on weekdays and weekends was not taken into account in predicting traffic flow propagation. Hence, our study is to create and evaluate a new prediction model for traffic flow propagation at neighbouring roads using similarity of traffic flow pattern on weekdays and weekends to achieve more accurate results. We exploit similarity of traffic flow pattern on weekdays and weekends by adding time cluster in our proposed model. Thus, our neural network model proposed high correlation road, time and day clusters as input factors in neural network model prediction. Our initial phase of the methodology involves investigation on correlation between neighbouring roads. This paper discusses the results of experiments we have conducted to determine relationship between roads in a neighbouring area and to determine input factors for our neural network traffic flow prediction model. To choose a particular road as a predicting factor, we calculated the distance between roads in neighbouring area to identify the nearest road. Then, we calculated correlation based on traffic condition (congestion) between roads in neighbouring area. The results were then used as input factors for prediction of traffic flow. We compared the results of the experiment using neural network without cluster parameters and multiple regression methods. We observed that neural network with time cluster parameter produced better results compared to neural network without parameter and multiple regression method in predicting average speed of vehicles on neighbouring roads.",
keywords = "Multiple regression method, Neural network, Traffic congestion, Traffic prediction",
author = "Bagus Priambodo and Azlina Ahmad",
year = "2018",
month = "1",
day = "1",
language = "English",
volume = "17",
pages = "513--535",
journal = "Journal of Information and Communication Technology",
issn = "1675-414X",
publisher = "Universiti Utara Malaysia Press",
number = "4",

}

TY - JOUR

T1 - Traffic flow prediction model based on neighbouring roads using neural network and multiple regression

AU - Priambodo, Bagus

AU - Ahmad, Azlina

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Monitoring and understanding traffic congestion seems difficult due to its complex nature. This is because the occurrence of traffic congestion is dynamic and interrelated and it depends on many factors. Traffic congestion can also propagate from one road to neighbouring roads. Recent research shows that there is a spatial correlation between neighbouring roads with different traffic flow pattern on weekdays and on weekends. Previously, prediction of traffic flow propagation was based on day and time during weekdays and on weekends. Results obtained from past studies show that further investigation is needed to reduce errors using a more efficient method. We observed from previous research that similarity of traffic condition on weekdays and weekends was not taken into account in predicting traffic flow propagation. Hence, our study is to create and evaluate a new prediction model for traffic flow propagation at neighbouring roads using similarity of traffic flow pattern on weekdays and weekends to achieve more accurate results. We exploit similarity of traffic flow pattern on weekdays and weekends by adding time cluster in our proposed model. Thus, our neural network model proposed high correlation road, time and day clusters as input factors in neural network model prediction. Our initial phase of the methodology involves investigation on correlation between neighbouring roads. This paper discusses the results of experiments we have conducted to determine relationship between roads in a neighbouring area and to determine input factors for our neural network traffic flow prediction model. To choose a particular road as a predicting factor, we calculated the distance between roads in neighbouring area to identify the nearest road. Then, we calculated correlation based on traffic condition (congestion) between roads in neighbouring area. The results were then used as input factors for prediction of traffic flow. We compared the results of the experiment using neural network without cluster parameters and multiple regression methods. We observed that neural network with time cluster parameter produced better results compared to neural network without parameter and multiple regression method in predicting average speed of vehicles on neighbouring roads.

AB - Monitoring and understanding traffic congestion seems difficult due to its complex nature. This is because the occurrence of traffic congestion is dynamic and interrelated and it depends on many factors. Traffic congestion can also propagate from one road to neighbouring roads. Recent research shows that there is a spatial correlation between neighbouring roads with different traffic flow pattern on weekdays and on weekends. Previously, prediction of traffic flow propagation was based on day and time during weekdays and on weekends. Results obtained from past studies show that further investigation is needed to reduce errors using a more efficient method. We observed from previous research that similarity of traffic condition on weekdays and weekends was not taken into account in predicting traffic flow propagation. Hence, our study is to create and evaluate a new prediction model for traffic flow propagation at neighbouring roads using similarity of traffic flow pattern on weekdays and weekends to achieve more accurate results. We exploit similarity of traffic flow pattern on weekdays and weekends by adding time cluster in our proposed model. Thus, our neural network model proposed high correlation road, time and day clusters as input factors in neural network model prediction. Our initial phase of the methodology involves investigation on correlation between neighbouring roads. This paper discusses the results of experiments we have conducted to determine relationship between roads in a neighbouring area and to determine input factors for our neural network traffic flow prediction model. To choose a particular road as a predicting factor, we calculated the distance between roads in neighbouring area to identify the nearest road. Then, we calculated correlation based on traffic condition (congestion) between roads in neighbouring area. The results were then used as input factors for prediction of traffic flow. We compared the results of the experiment using neural network without cluster parameters and multiple regression methods. We observed that neural network with time cluster parameter produced better results compared to neural network without parameter and multiple regression method in predicting average speed of vehicles on neighbouring roads.

KW - Multiple regression method

KW - Neural network

KW - Traffic congestion

KW - Traffic prediction

UR - http://www.scopus.com/inward/record.url?scp=85054085363&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054085363&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:85054085363

VL - 17

SP - 513

EP - 535

JO - Journal of Information and Communication Technology

JF - Journal of Information and Communication Technology

SN - 1675-414X

IS - 4

ER -