Font Size: a A A

Research On Detection And Deduction Of Disaster And Linkage Rescue System Of Long Tunnel

Posted on:2024-01-06Degree:MasterType:Thesis
Country:ChinaCandidate:M L MaoFull Text:PDF
GTID:2542307157972799Subject:(degree of mechanical engineering)
Abstract/Summary:PDF Full Text Request
With the rapid development of highway long tunnels in China,the corresponding increase in the number of sudden disasters during the operation of long tunnels has become a key factor in reducing the harm of sudden disasters in long tunnels At present,there are problems in tunnel system,such as low automation of disaster detection,poor mastery of the development trend of disaster after disaster,and low linkage of rescue system during emergency rescue after disaster.Therefore,with the support of the scientific research project "Intelligent Expressway Construction and Operation Management Information Technology Research(21-04X)" by the Shaanxi Provincial Department of Transportation,a design scheme for a long tunnel system that combines disaster detection technology,disaster deduction technology,and linkage rescue technology has been designed,achieving comprehensive judgment and linkage rescue of long tunnel disasters from occurrence to end,providing strong support for China’s highway long tunnels to respond to sudden disasters.Firstly,based on the analysis of disaster detection needs,disaster deduction needs and linkage rescue needs of long and long tunnels,the overall system scheme including disaster detection module,field control module,tunnel management center and linkage rescue module has been designed.Secondly,the hardware selection and network design of tunnel disaster detection module have been carried out,which realizes the real-time collection function of disaster information during tunnel operation.A model for tunnel fire detection and vehicle type identification based on YOLO V5 s is designed for automatic identification of tunnel disaster accidents.The training results show that the model can identify tunnel fire with an identification rate of 89.74% and different types of vehicles with an identification rate of 92.95%.Then,the disaster deduction model of long tunnels has been designed,the evolution path of disaster scenario of long tunnels has constructed by PSR model,and disaster deduction model of long tunnels has been established by Bayesian network method.The model divides disaster situation of long tunnels into scenario input layer,scenario analysis layer and scenario output layer.Disaster information collected by disaster detection module is input to scenario input layer.Scenario output layer will output disaster consequences and the next development trend of disaster;Finally,the disaster linkage rescue module of long and long tunnel has been designed.Combining the disaster deduction results of long tunnel with different disaster locations,the corresponding linkage rescue plans for different locations and different disaster levels are formulated.On this basis,the linkage rescue management platform is designed to realize the linkage control of rescue facilities in the tunnel after a sudden disaster.The proposed detection and deduction of disaster and linkage rescue system of long tunnel realizes the three steps of disaster detection,prediction,and rescue in long tunnels,realizes the entire process of handling disasters in long tunnels from occurrence to rescue,provides basis for disaster disposal during operation period of tunnel system in China,and has certain theoretical significance and Engineering application value.
Keywords/Search Tags:Long tunnel, Disaster detection, Deep learning, Disaster deduction, Bayesian network, Linkage rescue
PDF Full Text Request
Related items