Adaptive pedestrian behaviour for the preservation of group cohesion

摘要: A crowd of pedestrians is a complex system in which individuals exhibit conflicting behavioural mechanisms leading to self-organisation phenomena. Computer models for the simulation crowds represent consolidated type application, employed on day-to-day basis support designers and decision makers. Most state art models, however, generally do not consider explicit representation aggregations (groups) their implications overall dynamics. This work aimed at discussing research effort systematically exploring potential implication presence groups different situations (e.g. changing density, spatial configurations environment). The paper describes an agent-based model encompassing both traditional individual motivations (i.e. tendency stay away from other while moving towards goal) adaptive mechanism representing influence group simulated population. designed preserve cohesion specific types families friends) even high density turbulent situations. tested simplified scenarios evaluate modelling choices groups. produces results tune with available evidences literature, perspective pedestrian flows space utilisation, comprising groups; when are present, able challenging counterflow), it interesting that call further observations experiments gather empirical data. introduced effective qualitatively reproducing related phenomena stimulates efforts gathering evidences, one hand, additional leader-follower movement patterns).