Experiment of bi-direction pedestrian flow with three-dimensional cellular automata

• A novel bi-direction pedestrian flow model is presented with stair factor.
• The system evolution rules are given according to direction gain, collision gain and vacancy gain.
• Fermi rule is introduced to depict the competitive behavior.

In order to effectively depict bi-directional pedestrian flow characteristic in the three-dimensional space, a novel pedestrian model is proposed based on cellular automata. In the model, at first, the calculation formula of target position in the next time step is given according to direction gain and collision gain, and the system evolution rules are defined. Second, the experiment is conducted with the simulation platform to study the relationships of average system velocity, flow, system scale, pedestrian density and stair series for obtaining the bi-directional pedestrian flow characteristic. The results of numerical analysis had shown that when the pedestrian density reached the critical value, the pedestrian flow was changed from free flow status to blocking flow status, and the distribution ratio of bi-directional pedestrians and channel width had great effect on the average system velocity and the flow. Moreover, the more the stair series was, the more obvious the “faster-is-slower” effect was.