Authors: K. Jeevalakshmi, M. Gokul
Abstract: With the growing trend in urban development along with advanced intelligent transportation systems, there is a demand for advanced solutions for the management of the transport network. Since transportation networks can easily be modeled using graphs, graph theory is a useful math-ematical tool for optimizing complex networks. In this paper, we present an advanced study of graph theoretic methods applied to the smart transportation network optimization problem. The proposed hybrid approach combines the shortest path techniques with traffic prediction based on graph neural networks to optimize adaptive routing of vehicles. Our approach uses the graph with temporal load and makes collective route optimization in order to minimize the congestion in the network. As a result, the proposed methodology outperforms the classical heuristics with a 56% reduction in average passenger waiting time and a 26% reduction in energy consumption.
DOI: http://doi.org/10.5281/zenodo.22724496
