Generalising Traffic Forecasting to Regions without Traffic Observations
This addresses the challenge of deploying intelligent transportation systems in sensor-scarce regions, though it is incremental as it builds on existing traffic forecasting methods.
The paper tackles the problem of traffic forecasting for regions without traffic sensors by proposing GenCast, a model that integrates physics-informed neural networks and external signals like weather, resulting in consistently reduced forecasting errors on multiple real-world datasets.
Traffic forecasting is essential for intelligent transportation systems. Accurate forecasting relies on continuous observations collected by traffic sensors. However, due to high deployment and maintenance costs, not all regions are equipped with such sensors. This paper aims to forecast for regions without traffic sensors, where the lack of historical traffic observations challenges the generalisability of existing models. We propose a model named GenCast, the core idea of which is to exploit external knowledge to compensate for the missing observations and to enhance generalisation. We integrate physics-informed neural networks into GenCast, enabling physical principles to regularise the learning process. We introduce an external signal learning module to explore correlations between traffic states and external signals such as weather conditions, further improving model generalisability. Additionally, we design a spatial grouping module to filter localised features that hinder model generalisability. Extensive experiments show that GenCast consistently reduces forecasting errors on multiple real-world datasets.