Network level spatial temporal traffic forecasting with Hierarchical-Attention-LSTM

Abstract

Traffic state data, such as speed, density, volume, and travel time collected from ubiquitous roadway detectors require advanced network level analytics for forecasting and identifying significant traffic patterns. This paper leverages diverse traffic state datasets from the Caltrans Performance Measurement System (PeMS) hosted on the open benchmark and achieved promising performance compared to well-recognized spatial-temporal prediction models. Drawing inspiration from the success of hierarchical architectures in various Artificial Intelligence (AI) tasks, cell and hidden states were integrated from low-level to high-level Long Short-Term Memory (LSTM) networks with the attention pooling mechanism, similar to human perception systems. The developed hierarchical structure is designed to account for dependencies across different time scales, capturing the spatial-temporal correlations of network-level traffic states, and enabling the prediction of traffic states for all corridors rather than a single link or route. The efficiency of the designed hierarchical LSTM is analyzed by ablation study, demonstrating that the attention-pooling mechanism in both cell and hidden states not only provides higher prediction accuracy but also effectively forecasts unusual congestion patterns. Data and code are made publicly available to support reproducible scientific research.