Modeling ant colony optimization for multi-agent based intelligent transportation system

Abstract

This paper focuses on Sumo Urban Mobility Simulation (SUMO) and real-time Traffic Management System (TMS) simulation for evaluation, management, and design of Intelligent Transportation Systems (ITS). Such simulations are expected to offer the prediction and on-the-fly feedback for better decision-making. In these regards, a new Intelligent Traffic Management System (ITMS) was proposed and implemented - where a path from source to destination was selected by Dijkstra algorithm, and the road segment weights were calculated using real-time analyses (Deep-Neuro-Fuzzy framework) of data collected from infrastructure systems, mobile, distributed technologies, and socially-build systems. We aim to simulate the ITMS in pragmatic style with micro traffic, open-source traffic simulation model (SUMO), and discuss the challenges related to modeling and simulation for ITMS. Also, we expose a new model- Ant Colony Optimization (ACO) in SUMO tool to support a multi-agent-based collaborative decision-making environment for ITMS. Beside we evaluate ACO model performance with exiting built-in optimum route-finding SUMO models (Contraction Hierarchies Wrapper) -CHWrapper, A-star (A*), and Dijkstra) for optimum route choice. The results highlight that ACO performs better than other algorithms.