MODELING THE MOVEMENT OF CENTRALLY CONTROLLED VEHICLES AT INTERSECTIONS USING ARTIFICIAL INTELLIGENCE METHODS

Abstract

Urban intersections are critical bottlenecks in road networks, where traffic delays and conflicts concentrate. This paper presents a modeling framework for centrally controlled vehicles at a signalized intersection that integrates dynamic simulation with artificial-intelligence-driven decision policies. We address trajectory and time-window prediction, data-driven adaptive signal control, and multiobjective optimization criteria including average delay, throughput, stability, and safety indicators. The framework is grounded in contemporary traffic-flow models and optimization methods, while also drawing on mathematical advances in differential equations and numerical schemes developed by the research school of Osh State University. From an applied perspective, coordinated centralized control and predictive policies can meaningfully reduce delays and conflict probabilities. The results are intended for deployment within intelligent transportation systems and for building intersection-level digital twins to support adaptive controllers in municipal operations. The proposed methodology and evaluation protocol aim to bridge theoretical models and practical requirements, enabling scalable and sustainable traffic management.