Green Vehicle Routing Optimization in Urban Traffic Congestion Environment Based On K-Means Clustering and Genetic Algorithm

Abstract

As urban traffic congestion becomes increasingly severe, supply chain transportation faces the dual challenges of declining efficiency and mounting environmental burdens. To address this, this study constructs a Green Vehicle Routing Problem (GVRP) framework that integrates real-time traffic information. First, K-means clustering technology is employed to classify urban road networks by congestion levels, establishing a three-tier classification system of low, medium, and high congestion zones. Subsequently, polynomial regression methods are utilized to establish a quantitative relationship model between vehicle speed and carbon emission intensity. Based on this theoretical foundation, a multi-objective optimization framework is designed that comprehensively considers environmental costs and traffic impedance factors, with performance comparison tests conducted between genetic algorithms and classical shortest path algorithms. Experimental results demonstrate that genetic algorithms perform excellently when handling high-congestion road segments, significantly reducing carbon footprint while shortening transportation cycles. This research provides scientific basis and operational paradigms for enterprises to construct sustainable supply chain networks based on dynamic traffic information.