考虑结伴行为的行人流场域元胞自动机模型  

An extended floor field model based on cellular automata for pedestrian and group behavior

作　　者：童蔚苹[1] 程琳[1] 

机构地区：[1]东南大学交通学院,南京210096

出　　处：《系统工程理论与实践》2014年第9期2386-2391,共6页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(51178110;51378119);住建部城市综合交通技术(2013-K5-8)

摘　　要：与行人流疏散情景不同,在行人流聚集过程中,行人动力学特性具有一些显著的特点,如对目标位置的选择相对模糊不清,结伴同行的行人组凝聚力明显,组内行人存在着明显的跟随行为.论文以行人流聚集过程为研究对象,分析结伴行为对行人流聚集动力学的影响,扩展场域思想,提出一种扩展的行人流场域元胞自动机模型.模型运用感知区表示行人对周边环境的感知能力,提出用组内场对同伴间彼此凝聚、相互牵制的结伴行为进行建模.行人转移概率的计算新添了组内场因子并考虑了行人忽略场值的微小变化而保持原地不动的情景.模拟实验成功再现了行人对独立空间领域的需求和结伴行人的聚簇成团现象,讨论了行人组尺寸对聚集时间、行人间平均间距的影响.Compared to in evacuation process,the characteristics of pedestrian dynamics have some obvious feature in inflow process.In inflow process,the destination is not clear;group cohesion among members is obviously;and there is a clear following behavior among group members.In this paper,pedestrian dynamics in inflow process is mainly studied,and group behavior effect on inflow dynamics is analyzed in detail.Then,an extended floor field model based on cellular automata for pedestrian and group behavior is proposed.In the model,perception capabilities of pedestrian are modeled with the idea of perception fan,and the group floor field is introduced to represent the effect of cohesion force,which is a wish to preserve the group cohesion and minimize the distances among group members.The transition probabilities are modified by the group floor field and threshold which is related to the decision making about walk or stop：when the gradient of total floor fields is lower than the threshold,pedestrians tend to ignore the small gradient and stay on their present cells.Simulation results successfully capture the phenomena of pedestrians’ requirement for independent space and clustered together among group members.Inflow time and the minimal distance with other group members versus the impact factor of group size are discussed.

关 键 词：行人聚集 结伴行为 组内场 场域元胞自动机模型 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]