Analysis of Frameworks for Traffic Agent Simulations

Abstract

Traffic jams and congestion are a global challenge for smart cities. Digital transformation is touching all aspects of modern-day life utilities, from energy to health services and from education to smart mobility. A smart city aims to cut carbon dioxide emissions while saving lives by reducing driver or passenger trip time, traffic congestion and accidents. Traffic simulations and advanced traffic management systems attempt to address the urbanization infrastructure, sustainability, and mobility issues. Many road infrastructures are decades old and are therefore prone to an excess of frequent and recurrent traffic bottlenecks. This could be attributed to an increase in driver numbers and higher travel demands, which at the same time is increasing in the overall travel duration, air pollution and road accidents. The drivers of the United States alone travel 4.8 trillion kilometers per year. A smart city targets the efficient and sustainable use of its resources and services by using intelligent and interactive management. With the rapid growth of population density, the use of private cars and public transport, scientists thrive on trying to reduce urban traffic congestion in order to efficiently facilitate access to work, education, emergency services, entertainment and other services. For this purpose, graph algorithms and multi-agent system simulations are used for traffic optimization. While urban traffic congestions cannot be modeling eliminated, several frameworks for traffic simulations were developed to reduce traffic congestion. This paper will investigate and provide a comprehensive review of several frameworks used in these traffic agent simulations that can be found in open literature. It will make a comparative analysis of these frameworks used.