Bus line optimization using multi-agent simulation model of urban traffic behavior of inhabitants applying branch and bound techniques

In this paper, a multi-agent simulation model of urban traffic behavior using Origin-Destination data and an example of bus line optimization as an application of the model are showed. In the model, inhabitants´ behaviors of travel are expressed by decision of inhabitant agents. They select one of means of mobility and route by optimizing their evaluation functions. There are too many search candidates in their decisions to calculate in practical time, thus some assumptions and mathematical optimization techniques are applied to reduce calculation time. Route search algorithms of agents are divided into two parts of common and individual. The result of the common part is jointly used by all agents, thus the algorithm of the common part is executed only once. On the other hand, the individual part is built as parallelizable structure. Some potential of the proposed model is demonstrated by simulation examples of two scenarios in which a part of traffic networks differ each other. Furthermore, it is showed that an application example in which route of community buses in Imizu-city is improved using the proposed model.