Using Deep Traffic Prediction for EMFAC Emission Estimation and Visualization

Tarek Bin Zahid, Brendan Morris

Abstract: Accurate traffic emission prediction is crucial for developing effective traffic management strategies and reducing the environmental impact of transportation. In this study, we propose an approach to traffic emission prediction for Las Vegas using a graph-based transformer model and an environmental factors simulator (EMFAC) model. The transformer model is employed to predict traffic speed and flow and the predicted values are then used as input parameters to the EMFAC model to estimate traffic emissions. To address the lack of EMFAC model parameters for Las Vegas, we utilize the traffic profile data from San Bernardino, California, which exhibits similar traffic characteristics to Las Vegas. The proposed method is evaluated using real-world traffic data and demonstrates promising results in terms of accuracy and effectiveness.

External IDs:dblp:conf/itsc/ZahidM24