Research Of Safety Evaluation System On Affiliated Facilities Of Utility Tunnel Based On Bayesian Network

With the development of China's urban economy,the area of municipal facilities is very wide,and many phenomenons,like “sky spider webs” and “road zippers”,have appeared,which consume a lot of manpower,material and financial resources.Therefore,the construction of utility tunnel is particularly important.The utility tunnel contains multiple municipal transportation pipelines such as water supply,power supply,and gas pipelines.However,the utility tunnel has a large number of pipelines and a high layout density,so the pipelines may affect each other.Problems with any one pipeline will cause other pipelines,and the failure of the one system leads to the paralysis of the entire utility tunnel,and there may many risk factors such as fire,explosion,and flood.Therefore,the research on the safety of the affiliated facilities of the utility tunnel is what China needs to face and obtain in the subsequent utility tunnel construction process.On the basis of the construction and research status of affiliated facilities engineering systems of utility tunnel at home and abroad,this paper conducts a structural review of the subsystems of the affiliated facilities systems of utility tunnel,and uses the accident tree method to identify the risk sources in each system.Classified and sorted out,and finally established the accident tree model of seven subsidiary systems.A qualitative calculation of the minimum cut set for each system using Boolean algebra is performed to determine the importance of each factor of the corresponding system.Then based on the established accident tree model,a Bayesian network model of the ventilation system and drainage system is obtained by the conversion from the accident tree model to the Bayesian network,and a fuzzy probability method incorporating expert knowledge is proposed.And finally established a dynamic security evaluation method for the affiliated facilities system of utility tunnel based on fuzzy number and Bayesian network.Finally,taking Haikou's utility tunnel as an example,the fuzzy probability algorithm combining expert knowledge and Bayesian network is combined to analyze the actual application.The calculated fuzzy probabilities of the 23 basic events of the ventilation system and the 14 basic events of the drainage system are taken as prior probabilities,and they are input to the Bayesian network to update the probability to predict the failure probability of the two systems.Reveal the basic events of the ventilation system safety are the length of the fire prevention zone,the number of ventilations and the alarm interval.The basic events of the drainage system safety are the failure of the shut-off valve and the submersible sewage pump,and put forward targeted suggestions and measures.