Abstract
Many mathematical models have been proposed for analysis of traffic jams, and applied to simulation of traffic flow. Grouping cars based on the properties of driving, and then modeling each group eneables us to obtain an appropriate model. The simulation with the model will be more precise than a single model for whole cars. In this paper, we focus on modeling by TASEP, which is the most basic cellular automaton. Applying a multi-species TASEP to spatio-temporal data, we propose statistical clustering methods of cars.
