Abstract:Crowds in public spaces such as subway station are generally composed by diverse types of pedestrians with various motion features and behavior preferences. This paper aims to propose a behavior-based cellular automata model that can express the heterogeneous structure of crowds and explore the influence of different crowd composition on pedestrian dynamics, especially evacuation efficiency. The degree of aggressiveness is first introduced into cellular automata model as the internal state, which affects the motion characteristics and behavior preference of pedestrians. It is closely related to the motion characteristics including free walk velocity, target cells selection and conflict resolution mechanism. In addition, it determines the balance selection between keeping shorter paths and avoiding congestion. Simulation experiments of unidirectional pedestrian flow is conducted in the corridor with single exit. Simulation results indicate that a higher shortest path tendency leads to more time spent on blocking, and a higher detours tendency will cause longer detours time. A good balance between the two tendencies can make the individual's evacuation more efficient. A higher proportion of aggressive people and moderate degree of aggressiveness results in a higher average evacuation efficiency. In the crowd, the evacuation of conservative groups (i.e. pedestrians with zero aggressiveness) will be the most advantageous when the proportion of aggressive groups and degree of aggressiveness are moderate. For aggressive groups (i.e. pedestrians with the degree of aggressiveness greater than zero), evacuation efficiency will be the highest when the proportion of aggressive groups is moderate and the degree of aggressiveness is high. The proposed model provides a new perspective for the evaluation and analysis of heterogeneous pedestrian flow.

A Dual Adaptive Cellular Automaton Model Based on a Composite Field and Pedestrian Heterogeneity

Relationship Between Behavior Aggressiveness and Pedestrian Dynamics Using Behavior-Based Cellular Automata Model.

Effects of Crowding Perception on Self-organized Pedestrian Flows Using Adaptive Agent-based Model

Exploring crowd persistent dynamism from pedestrian crossing perspective: An empirical study

A Force-Based Model for Adaptively Controlling the Spatial Configuration of Pedestrian Subgroups at Non-Extreme Densities

A cellular automata approach for modelling pedestrian-vehicle mixed traffic flow in urban city

Simulating the bi-directional pedestrian flow under high densities by a Floor Field Cellular Automaton model

An Extended Model Describing Pedestrian Evacuation Considering Pedestrian Crowding and Stampede Behavior

A Continuum Model For Pedestrian Flow With Explicit Consideration Of Crowd Force And Panic Effects

An all-densities pedestrian simulator based on a dynamic evaluation of the interpersonal distances

A macroscopic pedestrian model with variable maximal density

A collective motion model based on two-layer relationship mechanism for bi-direction pedestrian flow simulation

Homogeneity and Activeness of Crowd on Aged Pedestrian Dynamics.

Pedestrian dynamics in single-file movement of crowd with different age compositions

Crowd Dynamics of Self-Propelled Individuals with Collision Avoidance Considering Anticipation and Intrusion Aversion

Continuous agent-based modeling of adult-child pairs based on a pseudo-energy: Relevance for public safety and egress efficiency

Simulation of Evacuation Characteristics Using a 2-Dimensional Cellular Automata Model for Pedestrian Dynamics.

Cellular Automata Modeling of Pedestrian’s Crossing Dynamics

Modelling Physical Contacts to Evaluate the Individual Risk in a Dense Crowd

Dynamics characteristic of pedestrians' particular overtaking behavior based on an improved social force model