Abstract:At present, there are few fire risk assessment models for historical buildings with wooden components that are highly vulnerable to fires. Bayesian network (BN) is a practical risk assessment method that is helpful to risk management. Because many indicators affect the fire risk of historical buildings, the BN may have many nodes and a complex structure. As the parent nodes increase, the number of conditional probability distributions required to determine the child node's conditional probability table (CPT) grows exponentially. Therefore, a combined method is proposed to reduce the complexity of determining CPTs by integrating various methods, including historical data collection, expert knowledge, logical relationships, and empirical formulas. Based on the proposed combined method, combining fire protection codes and provisions in China, a BN-based fire risk assessment model is presented for historical buildings, which integrates static indicators, such as building parameters and cultural significance, and dynamic indicators, such as environmental factors. In the model, the variable weight synthesis theory is introduced to automatically adjust the nodes' relative importance (weights) to improve the rationality of the BN for the special scenarios and the unconventional emergency scenarios. The proposed BN model can reflect the fire development and spread trend, assess the fire risk of historical buildings, analyze key factors, and propose fire improvement measures. Taking Nanyue Temple as a case study, based on the collected field data, the assessment results have demonstrated and verified the proposed model and method are feasible. The sensitivity analysis results have verified the rationality of the method and identified the key factors.

A method to improve the determination of ignition probability in buildings based on Bayesian network

A Combined Method to Build Bayesian Network for Fire Risk Assessment of Historical Buildings

Reliability analysis of building fire protection system based on Bayesian networks

Using Bayesian Network to develop a probability assessment approach for construction safety risk

A statistical model to predict ignition probability

Electrical Fire Dynamic Risk Assessment for High-Rise Buildings Based on Variable Fuzzy Set Theory and Bayesian Network

Fire Risk Assessments of Informal Settlements Based on Fire Risk Index and Bayesian Network

A Probabilistic Model for Fire Temperature Rise in High-Rise Residential Buildings under the Action of Uncertain Factors

Fire Risk Assessment of An Energy Storage Station Based on Bayesian Network and PyroSim

Quantitative Assessment of Building Fire Risk to Life Safety

RISK ASSESSMENT OF SUBWAY STATION FIRE BY USING A BAYESIAN NETWORK-BASED SCENARIO EVOLUTION MODEL

An Improvment Methods for Building Fire Detection Based on BP Neural Network

Using Bayesian Network to Develop an Approach for Construction Safety Risk Assessment

Decentralized Inverse Model for Estimating Building Fire Source Location and Intensity

A Bayesian network model for the Asian games fire risk assessment

Uncertainty Analysis of Parameters on Occupant Evacuation in Building Fires

A Risk Analysis Method of Cable Fire in Utility Tunnel Based on a Bayesian Network

A Review of Researches on the Estimation of Building Fire Parameters

An advanced fire estimation model for decentralized building control

Risk assessment of safety level in university laboratories using questionnaire and Bayesian network

Investigation of the Probability of a Safe Evacuation to Succeed in Subway Fire Emergencies Based on Bayesian Theory