Research On Integration Method Of Long-Span Spatial Steel Structure Health Monitoring System Based On BIM Platform

The large-scale and complex large-span spatial steel structure is often used in the station buildings and platform sheds of high-speed railway stations,resulting in the complexity of the monitoring system,the massive amount of monitoring data and so on.How to make full use of massive monitoring data and achieve the interaction,analysis and visualization of those data has become an urgent problem.According to the characteristics of long-span spatial steel structure and the advantages of BIM technology in information integration and three-dimensional visualization,an integrated method of health monitoring system for long-span spatial steel structure based on BIM platform is proposed in this paper.On this basis,the health monitoring function modules are developed in four aspects: monitoring database integration,monitoring information visualization,BIM model and finite element model interaction and monitoring data analysis.The main research works are as follows:(1)Taking the long-span spatial steel structure of Haikou East Station as the actual engineering background,the static and dynamic characteristics of the structure under 83 working conditions are analyzed by SAP2000,and a set of health monitoring system of the structure is studied and established.In addition,based on the theoretical response value of the most unfavorable working conditions and the limit value of the design code as the basis for the establishment of the early warning threshold,the method of early warning according to the type of members is put forward.(2)The database system for long-span spatial steel structure is established,the monitoring information of database is integrated with BIM model,and the functional modules of visualization of monitoring information are developed by RevitAPI.The modules are sensor point query,monitoring data query,monitoring curve display,real-time automatic early warning and background management module.(3)The data interaction function between finite element and BIM model is developed by using SAP2000 API and RevitAPI,and the finite element real-time analysis of the structure is carried out in combination with the integrated monitoring database.In addition,in order to solve the problem of weak generality of the above functions,The interface between BIM model and SAP2000 structural model is developed and the feasibility and accuracy of the transformation interface are verified based on Haikou East Station project.The automatic mapping between BIM model member and finite element model member is realized.(4)Taking the temperature and strain monitoring data of the steel grid roof of the Zhuhai Highway Port of the HongKong-Zhuhai-Macao Bridge as the actual engineering background,in order to predict the temperature effect and identify the wind load of long-span spatial steel structures,BP neural network is used to simulate the nonlinear relationship between temperature and strain.Finally,using interface of Matlab,the neural network program based on BIM platform is developed,which provides a custom neural network training operation interface,which can be used to determine the key members to be inspected in strong wind environment,and guide the monitors to repair the structure.