{"title":"Physical-informed random field technique for virtual modelling based building probabilistic vulnerability assessment","authors":"Zhiyi Shi , Yuan Feng , Mark G. Stewart , Wei Gao","doi":"10.1016/j.strusafe.2025.102595","DOIUrl":null,"url":null,"abstract":"<div><div>Developing a probabilistic vulnerability assessment framework for bushfire-prone buildings is a critical measure to reduce bushfire-induced risks to life safety and economic losses to an acceptable level. A reliable assessment approach should include multiple probability-based macro indicators by considering their inherent uncertainties. These macro indicators can incorporate the efficiency of bushfire-damaged transportation network at specified moments, the geographical position of buildings, among others. A Physics-Informed Random Field-Virtual Modelling (PIRF-VM) framework for probabilistic vulnerability assessment of bushfire-prone buildings in large-scale bushfire incidents is proposed. The PIRF generates a random field-based, multi-physical information-fusion model for the simulation of bushfire spread in a large-scale approximate natural environment. The integrated physical information includes the spatially varying vegetation characteristics, the Digital Elevation Model (DEM)-based terrain, the terrain-shaped time-dependent wind field, the geographical coordinates of roads and buildings. To mitigate the computational burden posed by stochastic bushfire simulations in PIRF, the VM is introduced. It can establish an explicit functional relationship between input physical information and output responses of interest, such as the remaining time for bushfire reaching a specified location. As a result, for any new input physical information, the output responses can be directly predicted without time-consuming simulations. Benefiting from the efficient predictions of the PIRF-VM, several probability-based macro indicators are simultaneously considered when assessing the probabilistic vulnerability for bushfire-prone buildings in large-scale bushfire incidents. The Australian community of Cowan serves as an example to illustrate the practical application of the proposed scheme, demonstrating potential in constructing more bushfire-resilient communities in the face of bushfire hazards.</div></div>","PeriodicalId":21978,"journal":{"name":"Structural Safety","volume":"115 ","pages":"Article 102595"},"PeriodicalIF":5.7000,"publicationDate":"2025-03-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Structural Safety","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0167473025000232","RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}
引用次数: 0
Abstract
Developing a probabilistic vulnerability assessment framework for bushfire-prone buildings is a critical measure to reduce bushfire-induced risks to life safety and economic losses to an acceptable level. A reliable assessment approach should include multiple probability-based macro indicators by considering their inherent uncertainties. These macro indicators can incorporate the efficiency of bushfire-damaged transportation network at specified moments, the geographical position of buildings, among others. A Physics-Informed Random Field-Virtual Modelling (PIRF-VM) framework for probabilistic vulnerability assessment of bushfire-prone buildings in large-scale bushfire incidents is proposed. The PIRF generates a random field-based, multi-physical information-fusion model for the simulation of bushfire spread in a large-scale approximate natural environment. The integrated physical information includes the spatially varying vegetation characteristics, the Digital Elevation Model (DEM)-based terrain, the terrain-shaped time-dependent wind field, the geographical coordinates of roads and buildings. To mitigate the computational burden posed by stochastic bushfire simulations in PIRF, the VM is introduced. It can establish an explicit functional relationship between input physical information and output responses of interest, such as the remaining time for bushfire reaching a specified location. As a result, for any new input physical information, the output responses can be directly predicted without time-consuming simulations. Benefiting from the efficient predictions of the PIRF-VM, several probability-based macro indicators are simultaneously considered when assessing the probabilistic vulnerability for bushfire-prone buildings in large-scale bushfire incidents. The Australian community of Cowan serves as an example to illustrate the practical application of the proposed scheme, demonstrating potential in constructing more bushfire-resilient communities in the face of bushfire hazards.
期刊介绍:
Structural Safety is an international journal devoted to integrated risk assessment for a wide range of constructed facilities such as buildings, bridges, earth structures, offshore facilities, dams, lifelines and nuclear structural systems. Its purpose is to foster communication about risk and reliability among technical disciplines involved in design and construction, and to enhance the use of risk management in the constructed environment