Enhanced Traffic Flow Forecasting on a Suburban Highway Using Empirical Mode Decomposition Preprocessed Deep Learning Model

DOI

10.2991/978-94-6463-884-4_46How to use a DOI?

Keywords

Traffic flow forecasting; Time series analysis; Empirical mode decomposition; Deep learning model; Intelligent transportation system

Abstract

Accurate traffic flow prediction is critical for efficient traffic control, urban planning, and congestion mitigation management. This study focused on using deep learning algorithms to predict short-term traffic flow on a suburban roadway. Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), Artificial Neural Network (ANN), and 1D Convolutional Neural Networks with LSTM (Conv1D-LSTM) models were used as baseline models. The study used Empirical Mode Decomposition (EMD), an analytical approach for non-stationary time series data in order to identify the outliers and to smooth the dataset in order to addressing the stochastic character of traffic flow. The dataset was derived from traffic surveillance video footage collected by the Highway Police Command Center on a two-way, four-lane suburban highway. Computer vision techniques, in collaboration with Sigmind, an AI-driven startup, were used to extract 5-min interval vehicle counts from 6:00 AM to 12:00 PM. Additionally, manual counting of vehicles from 12:00 PM to 3:00 PM extended the dataset to span 6:00 AM to 3:00 PM over 11 continuous days. Initial data analysis revealed overall trends, hourly fluctuations, and anomalies which were effectively mitigated using EMD preprocessing. According to the results, the EMD-preprocessed LSTM model demonstrated better forecasting accuracy, achieving the lowest RMSE (23.18) and MAPE (8.55) than GRU, ANN, and Conv1D-LSTM in capturing time-dependent interactions and non-linear traffic behaviors. These results highlight how crucial it is to combine deep learning techniques with sophisticated preprocessing approaches like EMD to increase prediction. This study highlights the potential of LSTM for precise short-term traffic forecasting on suburban roadways and offers insightful information for creating adaptive traffic management methods.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - MD. Ausaf Alam
PY  - 2025
DA  - 2025/11/18
TI  - Enhanced Traffic Flow Forecasting on a Suburban Highway Using Empirical Mode Decomposition Preprocessed Deep Learning Model
BT  - Proceedings of the 8th International Conference on Engineering Research, Innovation, and Education 2025 (ICERIE 2025)
PB  - Atlantis Press
SP  - 383
EP  - 392
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-884-4_46
DO  - 10.2991/978-94-6463-884-4_46
ID  - Alam2025
ER  -