Research On Safety Monitoring And Damage Diagnosis Of Large Conveyance Structure For South-to-North Water Diversion Project

The first stage of South-to-North Water Diversion Project in east and middle route is the most important inter-basin water resources allocation project in China,which assure water supply safety in the Beijing,tianjin and north China.Especially in the middle route,due to insufficient storage capacity,requires a high level of safe operation.The first stage of the middle route main channel,with complex geological terrain conditions along the road,has built large number of water conveyance structures,such as channel,aqueduct,tunnel,pipe culvert(PCCP),siphon etc.The first stage of the east route,crossed by numerous river networks,is constructed a large and complex water conveyance system by pumping stations and storage reservoir.The security of engineering operation is always a main concern of the managers and social attention since the project operation,and it is extremely important to research the methods that how to diagnose,recognition and evaluate the engineering health state.For the safe operation of the South-North Water Diversion Project requirements,this paper proposes a structural damage diagnosis method for large-scale water conveyance structures based on bayesian method.The main content and achievement as follow:(1)For large-scale water conveyance structures,the manifestations of structural damage are complex in different building,but the main factors that cause structure damage and damage detection mechanism are basically the same.Through the damage types and mechanism of the aqueduct water conveyance structures analysis,this paper select the factors influencing the aqueduct greatly and establish the aqueduct damage correlation model,and then evaluate the influence of different factors on the damage degree by AHP-Entropy weight method,this provides a basis for subsequent structural diagnosis;(2)According to the project profile of the Caohe aqueduct in the Beijing-Shijiazhuang section of South-North Water Diversion Project,this paper analyze the arrangement of safety monitoring instruments of the Caohe aqueduct,describe the specific observation frequency and operation of the instrument,process and explore the date collected from all parts of the aqueduct,thus determine the safety and stability during operation;(3)According to the aqueduct damage correlation model,this paper construct the bayesian network model for damage diagnosis of the aqueduct,determine the network structure and model parameters,establish the HMM model by the relation between the observed variable and the aqueduct damage state,apply the K-means algorithm to classify actual monitoring data processing,use the HMM learning algorithm to obtain the damage state transfer parameters;(4)Through the Netica simulation software to realize damage diagnosis of the aqueduct,this paper use dynamic bayesian network and static bayesian network to diagnose aqueduct damage,the results show that the dynamic bayesian network is more accurate in damage diagnosis and estimation,and can provide auxiliary damage diagnosis decision in practical engineering.