Parallel-Degree Enhancement For The Large-Scale And High-Resolution Artificial Societies Oriented To Emergency Management

Artificial society computational experiment,which bases on the construction of artificial society system,is a novel approach to emergency management study.However,the understanding of artificial society is not profound and extensive currently,and related theoretical research is not complete.Moreover,problems about the elements,structures and logical relations in artificial society for emergency management remain unsolved.As a virtual laboratory of social science,artificial society is always an important issue to sociologist and scientists since it focuses on the quantification,expression and realization of the consistency of artificial society and real world.In the context of emergency management for public health events,based on the current understanding of artificial society,the theory of artificial society for emergency management is studied in this paper,and a novel concept-"artificial society parallel degree" is then proposed.In the framework of the novel concept,parallel-degree enhancement method is studied for artificial society prototype system in "KD-ACP" platform.The works and innovations of this paper can be summarized as following aspects:(1)The concept of artificial society parallel degree is proposed.By applying the concept of "scenario" in artificial society theory,the parallel degree is able to explain the consistency or coincidence of artificial scenario and real scenario,which enriches the artificial society system and gives an indicator to measure the confidence of artificial society.Calculation approach of artificial society parallel degree is deduced by formalized symbols.The result that the level of artificial society parallel degree can be upgraded via improving parallel degree of each component model is concluded.(2)Virtual residential environment generation method of high parallel degree based on GIS data is developed.The generation method is developed using urban GIS data and analysis of urban residential architecture features.In order to ensure that features of virtual residential architectures at community and rural area have high consistency with artificial society in community level and rural level respectively,the novel method can provide appropriate architecture area size and number of floors to virtual residential environment in artificial society.(3)Artificial population self-adaptive behavior modeling method based on regional epidemic perception is proposed.The regional spatial epidemic description method is designed based on three state features of epidemic.Perception model of artificial population to epidemic is then designed.Finally,artificial population self-adaptive behavior modeling method based on regional epidemic perception is constructed via setting self-adaptive behaviors in different epidemic perception conditions.This method enables artificial population to perceive surrounding epidemic.As a result,artificial population are capable to self-adaptive-protect themselves because they can take protection measures in different ways and different degrees according to the threat of surrounding epidemic.(4)Artificial population traveling behavior modeling method is proposed based on artificial population behavior model framework of temporal-spatial migration.In this paper,temporal-spatial migration of artificial population in transportation network is abstracted from traveling behavior of artificial population.The artificial population traveling behavior model based on temporal-spatial migration is then constructed using transportation network modeling,path programming and event-chain generation techniques.The novel model largely reduces the influence of the original mechanism of artificial society prototype system caused by artificial population traveling behavior modeling.Moreover,artificial population is able to travel along real transportation routes in artificial society system.(5)A "hierarchical targeted immunization" strategy aiming to control epidemic propagation is proposed.In this paper,the effect of controlling epidemic propagation by social network hierarchical structure is pointed.And the novel strategy named "hierarchical targeted immunization" is therefore proposed,which has better effect than traditional "global targeted immunization".Besides,this is capable for epidemics of both SIR and SIS spread dynamic models,and it is quite stable when controlling epidemics.In this paper,artificial society parallel degree and parallel-degree enhancement are studied from both theory and realization technique.In terms of theory,an indicator and its calculation method aiming to measure the consistency of artificial society system and real society are designed.This method can be applied to support evaluating qualities of simulation experiment in artificial society system.Besides,this method enriches the theoretical system of artificial society.As for realization technique,research works are conducted in large-scale virtual environmental modeling,population self-adaptive behavior modeling and population traveling behavior modeling respectively.This study upgrades the level of parallel degree in artificial society for public health emergency management,and it is of extensive significance for the study of epidemic spread and control.