The pedestrian flow dynamics are very complex. Even the fundamental diagrams of uni-directional pedestrian flow exhibit diversity. In the empirical data of Mori-Tsukaguchi and Helbing et al, the flows drop… Click to show full abstract
The pedestrian flow dynamics are very complex. Even the fundamental diagrams of uni-directional pedestrian flow exhibit diversity. In the empirical data of Mori-Tsukaguchi and Helbing et al, the flows drop drastically with medium densities and nearly keep constant with high densities, which is different from many other empirical and experimental results. To simulate these special phenomena with cellular automaton (CA) models, we use the new configuration, in which the occupied area of each pedestrian is 0.4 m * 0.2 m. The modeling framework is based on the ITP model, which is a two-process CA model presented in our previous paper, and one new rule is added: when the local density behind is below a threshold, pedestrians prefer to keep a certain front gap. With the simple rule, the new fundamental diagrams can qualitatively coincide with the empirical facts of Mori-Tsukaguchi and Helbing et al. Some other results, including simulation patterns and velocities of pedestrians can further confirm the validity of the new rule, and give some explanations for the diversity of different empirical data. These results are useful contributions for pedestrian flow modeling.
               
Click one of the above tabs to view related content.