A new heuristics model of simulating pedestrian dynamics based on Voronoi diagram*

Abstract

A new heuristics model based on the Voronoi diagram is presented to simulate pedestrian dynamics with the non-crowded state, in which these mechanisms of preference demand evading and surpassing, microscopic anti-deadlock, and site-fine-tuning are considered. The preference demand describes the willingness determination of detouring or following other pedestrians. In the evading and surpassing mechanisms, in order to achieve a balance between avoiding conflicts and minimizing detour distances, a new pair of concepts: “allow-areas and denial-areas” are introduced to divide the feasible region for pedestrians detour behaviors, in which the direction and magnitude of detour velocity are determined. A microscopic anti-deadlock mechanism is inserted to avoid deadlock problem of the counter-directional pedestrian. A site-fine-tuning mechanism is introduced to describe the behavior of avoiding getting too close to the neighbors in pedestrian movement. The presented model is verified through multiple scenarios, including the uni- or bi-direction pedestrian flow in the corridor without obstacles, the uni-direction pedestrian flow in the corridor with obstacles, and the pedestrian evacuation from a room with single-exit. The simulation results show that the velocity–density relationship is consistent with empirical data. Some self-organizing phenomena, such as lanes formation and arching are observed in the simulation. When pedestrians detour an obstacle, the avoiding area before the obstacle and the unoccupied area after the obstacle can be observed. When pedestrians evacuate through a bottleneck without panic, the fan-shaped crowd can be found, which is consistent with the actual observation. It is also found that the behavior of following others in an orderly manner is more conducive to the improvement of the overall movement efficiency when the crowd moves in a limited space.

An extended social force model on unidirectional flow considering psychological and behavioral impacts of hazard source

“…Another simulation study from the same first author modelled a mixed exit selection strategy in a multi-exit scenario, and found that there is a critical density to distinguish whether a pedestrian jam occurs in the evacuation and whether an exit selection strategy is in effect. [14] Wu et al [15] introduced a heuristic model based on the Voronoi diagram and successfully reproduced some typical phenomena. Jia et al [16] used the social force model approach and found that the evading and surpassing behavior increases the evacuation time, but that a circle-shaped obstacle induces a relatively higher bottleneck passing capacity.…”

A modified heuristics-based model for simulating realistic pedestrian movement behavior

“…Because of the good performance and simple rules, the heuristics-based model has attracted increasing attention in recent years. [21][22][23][24][25][26] Although the model has been widely applied, there remain deficiencies in theory and also in practical applications. To overcome the deficiency that the original heuristics-based model fails to reproduce lane formation in the dense crowds, Guo et al [27] proposed an improved model by modifying the objective function of the target direction.…”