Abstract:
Accurate spatio-temporal traffic forecasting underpins dynamic strategy and applications for intelligent transportation systems, improving traffic safety and reducing road congestion. Deep learning methods like CNN have improved traffic flow forecasting. These CNN-based methods learn traffic as images to model spatial correlation, which only works for Euclidean grid map data, not multi-sensor data. We propose a graph-based temporal attention framework GTA that considers spatial and temporal correlation to forecast traffic flow using data from multiple sensors. Because it preserves algorithm details, GTA can better capture spatial dependencies using graph embedding techniques on sensor networks. We add an attention mechanism to adaptively identify temporal submodule relations. Transportation networks’ topological properties enable better spatial-temporal integration. GTA is tested with a large English traffic dataset and topology information. Our approach outperforms several state-of-the-art baselines in experiments.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
