Research on Traffic Forecasting Using Time Convolutional Networks Optimised by RIME Algorithm

DOI

10.2991/978-94-6239-652-4_16How to use a DOI?

Keywords

Deep learning; Traffic forecasting; Intelligent algorithm

Abstract

This paper addresses the challenges in traffic flow forecasting—namely data non-linearity, strong spatio-temporal coupling, and hyperparameter dependence on manual tuning—by proposing a hybrid forecasting model based on the Frost-Ice Optimisation Algorithm (RIME). This model integrates a Time Convolutional Network (TCN), a Bidirectional Gated Recurrent Unit (BiGRU), and a Multi-Head Attention mechanism. The model leverages TCN’s convolutional layers to capture local spatial features within traffic sequences, employs BiGRU bidirectional encoding to encode long-term temporal dependencies, and incorporates a multi-head attention mechanism to dynamically weight critical temporal step information. To further enhance performance, the RIME algorithm is applied for automatic hyperparameter tuning of learning rates, convolution kernel sizes, and neuron counts, overcoming limitations of traditional manual parameter adjustment. Experiments on real-world traffic datasets demonstrate that the proposed model achieves outstanding results in metrics such as Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE), validating its effectiveness and robustness in high-volatility traffic forecasting tasks.

Copyright

© 2026 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  - Haorui Li
PY  - 2026
DA  - 2026/04/19
TI  - Research on Traffic Forecasting Using Time Convolutional Networks Optimised by RIME Algorithm
BT  - Proceedings of the  2026 5th International Conference on Engineering Management and Information Science (EMIS 2026)
PB  - Atlantis Press
SP  - 148
EP  - 154
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6239-652-4_16
DO  - 10.2991/978-94-6239-652-4_16
ID  - Li2026
ER  -