Coupling statistical and agent-based models in the optimization of traffic signal control

Abstract

There have been two directions to target to the problem of Traffic Signal Control (TSC): macroscopic and microscopic. On one hand, macroscopic help to find the optimal solution with an assumption of homogenization (both for vehicles and environment). On the other hand, microscopic one can take into account heterogeneity in vehicles as well as in environment. Therefore, it is very important to couple the two directions in the study of TSC. In this paper, we proposed to couple statistical and agent-based models for TSC problem in one intersection. The experiment results indicated that the proposed model is sufficient good in comparison with some others TSC strategies.