Muhammed, Bikis (Missouri University of Science and Technology), Hurson, Ali R. (Missouri University of Science and Technology), Sedigh Sarvestani, Sahra (Missouri University of Science and Technology), Gamage, Lasanthi (Webster University)

A GAT-BiLSTMA Model for Weather-Aware Prediction of Traffic Speed

Scheduled for presentation during the Poster Session "Travel Information, Travel Guidance, and Travel Demand Management" (ThAT16), Thursday, September 26, 2024, 10:30−12:30, Foyer

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024, Edmonton, Canada

This information is tentative and subject to change. Compiled on October 14, 2024

Abstract

This paper presents a method for incorporating the effect of weather conditions in prediction of the average speed of vehicular traffic for each segment of a road network. The proposed approach utilizes two different deep learning methods: graph attention networks and bidirectional long short-term memory with attention layers. The accuracy of predictions is increased by considering the real-world driving distance between road segments, in contrast to the haversine distance used in several existing prediction methods. Categorization of input data as weekend or weekday further increased the prediction accuracy. The proposed approach was validated with two data sets published by the California Department of Transportation, PeMSD4 and PeMSD7. One year of traffic data was supplemented with weather data and used to predict the average traffic speed of each road segment for up to 60 minutes into the future. The method was shown to maintain accuracy over multiple time horizons, scale well with respect to the number of road segments, and outperform existing prediction methods in prediction accuracy.