Multi Agent system Randomized in Road Traffic

In recent years, the rapid growth of road traffic density in theworld generates a rising request for tools that can be used toanalyze and control the traffic networks. Microscopic trafficsimulation[1,2,3] has proven to be one of the most useful toolsfor analysis of various traffic systems. Thus microscopic trafficsimulation has become an ever increasing field of research anddevelopment. In this paper we present a new model formicroscopic traffic simulation; Traditional traffic simulationmodels neglect some real-life factors that need to be considered,such as the effect of random distribution in the entry of lanehence ,this paper presents the development of a concept and anassociated architecture called the MAS2RT (Multi Agentsystem Randomized in Road Traffic), which is based on thecombination of Multi-Agent Systems (MAS) [4,5], Poissondistribution in urban Traffic and a stochastic model [6,7] tocapture the randomness of individual agent. The secondcontribution of this paper is about the internal structure ofmobile agents[8], initially reacts according to the instructions ofthe Main agent (MA) and in the case of a lack of dynamicinformation, the mobile agents take decisions based on theirexperiences accumulated during previous interactions.The obtained results illustrate that using the randomness in thereaction of agent enhanced greatly the performance ofsimulation.