Personalized mechanism of randomization on the road based on multi-agent collaboration

Mechanism of randomization describes behaviors of velocity randomization, but ignoring individual differences in traffic such as psychosocial, cultural and other social complexity. In this paper, 1) building multi-level model of roads based on the complexity of the analytical community, and turn factors of randomization into basic elements, especially the personalized factors, expanding the architecture of Agent in the traditional homogenization cellular automata model, 2）integrating those factors by the Multi-Agent collaboration. Describing the differences of randomization of individual cellular, making it more detail in traffic psychology, culture and other social complexity, the whole randomization behaviors of simulation transport system will be closer to reality. Finally, through designing simulation systems according to the specific situations in the related literature, analyzing and integrating factors of randomization, the results of speed - density and flow – density have a higher degree of compliance with the actual situation than the data of original paper to prove the advantages of personalized mechanism of randomization which have been built in this paper.