To achieve the effective disaster prevention and emergency disposal decision-making of ancient buildings,the multi-level architecture of"data-processing-business-service"and the BN-KM algorithm(Bayesian network and K-Means clustering)were adopted,and an intelligent method for the fire risk assessment and decision-making suggestion of ancient buildings based on the digital twin model was established.The multi-regional applicability of this method was verified through eight historical cases of ancient building fires,and the accuracy was 87.50％.The time-dynamic applicability of this method was verified through the fire risk evolution law of G ancient building(group)in four time periods.The results show that the multi-level architecture can realize the integration and quantification of heterogeneous data from multiple sources(e.g.histor-ical data,monitoring data,etc.),and the second-by-second computation of fire risk assessment.The digital twin method can calculate and deduce the high-risk areas before the accident occurs or without causing important losses.Meanwhile,it can pro-vide professional services for different objects(decision-makers,rescuers,and the public).The research results describe the assessment and decision-making suggestion process of the new intelligent method.This methodology may be applied to other disaster scenarios(earthquake,etc.),and provide opinion support for disaster prevention and emergency disposal of ancient buildings.
维普
万方数据