Implementation of a Fuzzy Multi-Agent Model for City Evacuation Traffic Management Using Probabilistic Automata

Because of importance of quickly city evacuation during natural or unnatural happenings, it’s essential to apply an optimized control policy to prevent congestion and stop of vehicles. Existing works for traffic management in critical conditions have paid little attention to artificial intelligence approaches. Therefore, the main goal of authors in this research is offering an optimized and intelligent control policy for city evacuation traffic. This policy uses fuzzy inference system for decision making of each agent and probabilistic automata for optimizing performance of agents as for their preferences during time. To check degree of success of offered control policy, Agent Base Simulation in RStudio and Netlogo environments have been implemented using RNetlogo and frbs packages in R language. Simulation results show traffic load distribution, using maximum capacity of roads and congestion prevention by suggested policy. With regard to communication technologies such as GPS, smart phones, automatic tax payment systems in roads and … that have been developed in recent years, it is also possible to implement suggested critical traffic control policy in real world.