Hierarchical Flow Decomposition for Turning Movement Prediction at Signalized Intersections

arXiv cs.LG / 4/13/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces HFD-TM (Hierarchical Flow-Decomposition for Turning Movement Prediction), a hierarchical deep learning approach that predicts corridor through-movements first and then derives turning-movement predictions from them.
It is motivated by traffic-structure findings that corridor flows cover 65.1% of volume, have lower volatility than turning flows, and explain 35.5% of turning-movement variance.
A physics-informed loss is used to enforce flow conservation, aiming to keep the predicted traffic structure consistent.
Experiments on six intersections using six months of 15-minute interval LiDAR data from Nashville, Tennessee show a mean absolute error of 2.49 vehicles per interval and improvements over a Transformer (5.7%) and a GRU (27.0%).
The authors report that hierarchical decomposition yields the biggest gains and that training time is 12.8× lower than DCRNN, suggesting suitability for real-time adaptive signal control.

Abstract

Accurate prediction of intersection turning movements is essential for adaptive signal control but remains difficult due to the high volatility of directional flows. This study proposes HFD-TM (Hierarchical Flow-Decomposition for Turning Movement Prediction), a hierarchical deep learning framework that predicts turning movements by first forecasting corridor through-movements and then expanding these predictions to individual turning streams. This design is motivated by empirical traffic structure, where corridor flows account for 65.1% of total volume, exhibit lower volatility than turning movements, and explain 35.5% of turning-movement variance. A physics-informed loss function enforces flow conservation to maintain structural consistency. Evaluated on six months of 15-minute interval LiDAR (Light Detection and Ranging) data from a six-intersection corridor in Nashville, Tennessee, HFD-TM achieves a mean absolute error of 2.49 vehicles per interval, reducing MAE by 5.7% compared to a Transformer and by 27.0% compared to a GRU (Gated Recurrent Unit). Ablation results show that hierarchical decomposition provides the largest performance gain, while training time is 12.8 times lower than DCRNN (Diffusion Convolutional Recurrent Neural Network), demonstrating suitability for real-time traffic applications.