Research On Risk Assessment Of Bridge Fire Based On Fuzzy Bayesian Network

In recent years,the span of bridge engineering has been increasing,structural technology and other innovations gradually,bridge construction and operation and maintenance has become the key to affect the economy,traffic safety,which means to improve the transport capacity at the same time to ensure the health and development of the city,improve the bridge operation risk prevention and control ability.Bridge fire is one of the gaps in bridge operation management at present.It is very necessary to build a bridge fire disaster evaluation index system,strengthen the risk management in the bridge operation stage,control fire,reduce fire incidence and ensure traffic safety to improve the bridge fire risk management.Relying on the jiangsu province taizhou Yangtze river bridge project,this paper based on the analysis summary bridge fire disaster mechanism and the advantages and disadvantages of existing Bridges fire hazards evaluation methods,scope of application,put forward suitable for Bridges small probabilistic fire disasters,sudden and fuzzy characteristics of risk factors of fuzzy bayesian network risk evaluation model,and verifies the case after the actual data research,as a bridge of fire disaster emergency decision-making,risk management to provide scientific and effective method.First of all,in this paper,the bridge fire disaster risk mechanism are analyzed,and the practical engineering with questionnaire investigation,literature research and field survey methods such as identification of risk factors,the following results: bridge system including risk source and risk factors of fire risk,the carrier and the vulnerability factors of risk and risk of disaster prevention and mitigation measures and the effective degree of the three subsystems;The target node,intermediate node and evidence node of the network topology were determined,and the bayesian network topology was established by introducing NETICA software.Secondly,the fire risk assessment model of bridge based on fuzzy bayesian network is established and inferred.The specific steps are: establishing the basic influencing factor set,collecting and sorting the evaluation set,calculating the membership matrix,completing the probability transformation of the membership matrix,calculating the fuzzy conditional probability,and reasoning and analysis.This paper established the basic factor set including 18 risk factors.Established the fire risk of the very large bridge bridge(? level),major fire accident risk(?),large bridge fire accident risk(?),general bridge fire accident risk(?)four level of evaluation set,the risk degree is described as a dangerous(" Critical "),High(" High "),Medium(" Medium "),Low(" Low ");The trapezoidal distribution function is used to calculate the fuzzy membership matrix.By transforming membership degree into probability and calculating fuzzy conditional probability,NETICA software is adopted for reasoning calculation of bridge fire risk results.Finally,the risk assessment model of taizhou Yangtze river bridge is applied.Through causal reasoning,diagnostic reasoning and sensitivity analysis of the calculation results,it can be known that the fire risk degree of taizhou Yangtze river bridge is "Medium";The order of importance of fire factors is as follows: disaster prevention and reduction factor,risk source factor,risk carrier factor;Based on the calculation of the sensitivity of risk factors,the key factors that should be paid attention to are: ecological impact,traffic congestion,safety distance between Bridges and major sources of fire and explosion,and the degree of difficulty in repairing.The risk sensitivity is 0.00641,0.00329,0.00311 and 0.00257,respectively.The results show that the fuzzy theory is introduced into the traditional bayesian network,can well solve the traditional bayesian network the continuous variables and the demand of a large amount of data that exist in the defects,at the same time the introduction of fuzzy theory to enhance the scientific nature and rationality of the bayesian networks,the evaluation results and taizhou changjiang river bridge in recent years the fire happen to coincide with the actual situation.The application of fuzzy bayesian network risk assessment model in taizhou Yangtze river bridge can provide reference for other bridge fire disaster risk management.