Crowd Hybrid Model for Pedestrian Dynamic Prediction in a Corridor

• Main Authors: Xiaoxia Yang, Qianling Wang
• Format: Article
• Language: English
• Published: IEEE 2019-01-01
• Series: IEEE Access
• Subjects: Pedestrian flow; social force model; density prediction; hybrid pedestrian movement model
• Online Access: https://ieeexplore.ieee.org/document/8762171/

The pedestrian dynamic prediction is of great theoretical significance to provide information to staff in public buildings for decision makings. Based on the conservation law of mass and the social force model, the crowd hybrid model, in which the heterogeneity of pedestrians from the microscopic level is considered, is established to predict the dynamic characteristics of pedestrian flow in a corridor. In this model, the corridor is divided into multiple calculation black boxes in which the number of pedestrians is conserved, and the adopted density-outflow data via social force model is required to be saved in the data base in advance. The crowd dynamics in the corridor is studied, and simulation results indicate that the pedestrian density and motion state, i.e., free state and jamming state, can be predicted. The proposed crowd hybrid model combines the advantages of both macroscopic pedestrian movement model with less computation and microscopic pedestrian movement model considering the detailed interactions of individuals. This hybrid modeling method is especially suitable for the pedestrian dynamic prediction in a corridor where a camera or laser cannot satisfy the requirements of monitoring.