Study On Analysis And Prediction Method Of Monitoring Data For Box Culvert Structure Of South-to-North Water Diversion Project

The South-to-North Water Diversion Project is an important means to optimize the space-time distribution of China’s water resources and solve the practical problems of water resources shortage.In order to ensure the safety of the project’s water delivery and normal operation,a large number of safety monitoring works have been carried out on the typical section along the project.In this thesis,the changing rules and influencing factors of each monitoring data was summarized and prediction models were established by processing and analyzing the safety monitoring data of the box culvert structure of Tianjin.The safe operation of the project was ensured by the box culvert structure monitoring information analysis module which was developed on the basis of the needs of project.The main research works and results of this thesis are as follows:(1)The changing rules and main influencing factors of each monitoring data was summarized by processing and analyzing the safety monitoring data.Each monitoring data changed like sinusoidal curve according to the annual cycle.The results showed that the opening-closing degrees’ main influencing factor is temperature and the reading of non-stressmeters was influenced by temperature and time,and steel stress was influenced by temperature,time and water level.(2)The forms of influencing factors were determined and statistical regression models of box culvert structure was established by referencing dam safety monitoring data analysis theory.The linear expansion coefficient and autogenous volume deformation process of concrete and the deformation features in horizontal direction of the box culvert were analyzed.(3)The evolutionary rule of fruit fly optimization algorithm(FOA)were improved to overcome the demerits of FOA,such as easily falling into local optimum,slow convergence rate and low convergence precision.The taste judging concentration of FOA was improved according to the feature of the support vector regression(SVR)model’s kernel function parameter.The SVR model of steel stress was established by using IFOA to optimize the optimum parameters of SVR.The finite element model of box culverts was established.The superiority of IFOA-SVR steel stress prediction model was verified by comparing the simulation results of the statistical regression model,FEM,and the IFOA-SVR model.(4)The box culvert structure monitoring information analysis module was designed and developed on the basis of ensuring the safe operation of the project and meeting the needs of users.A series of functions were realized in this system,such as visualization,processing,alarming and early-warning of monitoring data.