STTF: An Efficient Transformer Model for Traffic Congestion Prediction

Abstract

With the rapid development of economy, the sharp increase in the number of urban cars and the backwardness of urban road construction lead to serious traffic congestion of urban roads. Many scholars have tried their best to solve this problem by predicting traffic congestion. Some traditional models such as linear models and nonlinear models have been proved to have a good prediction effect. However, with the increasing complexity of urban traffic network, these models can no longer meet the higher demand of congestion prediction without considering more complex comprehensive factors, such as the spatio-temporal correlation information between roads. In this paper, we propose a traffic congestion index and devise a new traffic congestion prediction model spatio-temporal transformer (STTF) based on transformer, a deep learning model. The model comprehensively considers the traffic speed of road segments, road network structure, the spatio-temporal correlation between road sections and so on. We embed temporal and spatial information into the model through the embedding layer for learning, and use the spatio-temporal attention module to mine the hidden spatio-temporal information within the data to improve the accuracy of traffic congestion prediction. Experimental results based on real-world datasets demonstrate that the proposed model significantly outperforms state-of-the-art approaches.

Predicting vehicle travel time on city streets for trip preplanning and predicting heavy traffic for proactive control of street congestion

“…The paper of 19 suggests a vision transformer approach for traffic congestion prediction on a city-wide scale. In 20 , 21 , the authors devise a traffic congestion prediction model based on a deep learning model. The work of 22 integrates traffic science with representation learning for city-wide congestion prediction.…”

Drive Like a Machine: A Green Reflection of Red Flag Acts for Intelligent Vehicles [History and Perspectives]

“…For more than two decades, dedicated scientists and researchers have been working hard to make autonomous vehicles (AVs) drive like humans. Given that humans were once the exclusive operators and creators of AVs, it's natural to deduce that if AVs could mimic human performance, their safety and efficiency would be assured [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15]. Yet, bearing in mind that humans have never surpassed a speed of 44.722 km/h, even on an open running track, it feels somewhat unrealistic to anticipate that mimicking human behaviors and reactions could lead us to design the most skilled drivers.…”

A deep marked graph process model for citywide traffic congestion forecasting

“…Recently, a few studies have explicitly incorporated traffic patterns into deep learning models to developed pattern‐aware spatiotemporal prediction models (Di et al., 2019; Leiser & Yildirimoglu, 2021; Zheng et al., 2023). While many previous studies did not explicitly model the evolution of individual congestion events and only predicted congestion based on regular time intervals (Kumar & Raubal, 2021), some researchers have considered congestion as spatiotemporally propagating events and have attempted to perform fine‐grained congestion forecasting using graph embedding (Sun et al., 2022; Wang et al., 2023) or point process models (Zhu et al., 2022).…”