A Crowd Model with Multiple Individual Parameters to Represent Individual Behaviour in Crowd Simulation

This paper presents a crowd model which aims to simulate the individual behaviour and increase the heterogeneity in crowd simulation thus to achieve a more realistic simulation result. This model builds upon the physics foundation by referencing both Reynolds’ Boids model and Helbing’s Social Force model. The mechanism of behaviour processing is based on the Reynolds’ steering behaviours model. The basic individual behaviours such as seek, wander and avoid obstacles are identified and serve as the basis to achieve the more complex behaviours, such as follow the leader (grouping), queue at doorway (clogging) and so on. Fine network model is used to obtain the precise positions of each individual and the multi-agent system approach is adopted to support the decision making process. A set of key personal parameters (both physical and psychological) are identified for the individuals, which act as independent intelligent agents and conduct their own decision and behaviour. The group/collective behaviour can also be observed through the individuals which react to the surrounding crowd (those personal parameters also determine how they would response). All the individual parameters are configurable in order to provide the ability to fine-tune the model to fit various needs and scenarios. A series of test simulations with various individual parameters have been conducted during the research and the results are presented in the paper.