EMD-SVR: A Hybrid Machine Learning Method to Improve the Forecasting Accuracy of Highway Tollgates Traveling Time to Improve the Road Safety
Paper in proceeding, 2021

Tollgates are known as the bottleneck of the highways, which cause long waiting queues in rush-hour times of the day. This brings many undesirable consequences such as higher carbon emission and road safety issues. To avoid this scenario, traffic control authorities need accurate travel time forecasts at tollgates to take effective action to monitor potential traffic load and improve traffic safety. Accurate forecasting of the traffic travel time will help traffic regulators to prevent arising problems by taking action. The main objective of this study is to improve the short-term forecasting (minutes) of the traffic flow on highway tollgates by improving a novel hybrid forecasting method that combines Empirical Mode Decomposition with Support Vector Regression (EMD-SVR). Results claim that compared with SVR, the new proposed hybrid prediction model, EMD-SVR, can effectively improve prediction accuracy. Better forecasting of the traffic load will provide safer roads but will also lower the carbon emissions caused by longer traveling times.

Empirical Mode Decomposition

Forecasting

SVR

Machine learning

Author

Atilla Altintas

Chalmers, Mechanics and Maritime Sciences, Fluid Dynamics

Lars Davidson

Chalmers, Mechanics and Maritime Sciences, Fluid Dynamics

Lecture Notes of the Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering

1867-8211 (ISSN)

Vol. 364 LNICST 241-251

4th EAI International Conference on Intelligent Transport Systems, INTSYS 2020
Porto, Portugal,

Subject Categories

Transport Systems and Logistics

Infrastructure Engineering

Other Civil Engineering

DOI

10.1007/978-3-030-71454-3_15

More information

Latest update

5/6/2021 1