Deformation Monitoring Model For 300m Level High Arch Dam In Initial Impoundment Period

With the continuous development of dam construction technology in my country,a number of high concrete arch dams represented by the Xiao wan,Xiluodu and Jinping I have been or will be built and used.Many of the technical indicators of these high dams and large reservoirs have exceeded the existing specifications and standards.The overall stability problem is very prominent,and the operation process may be affected by various adverse factors such as special loads and extreme weather.The safe operation of the dam is facing severe tests.Deformation is one of the more intuitive effect quantities that reflect the working behavior of high arch dams.Strengthening the research on the deformation behavior of high arch dams and its safety monitoring model is one of the important means to offer the safe operation to high arch dams.According to the existing engineering experience,engineering problems often appear in the initial stage of impoundment.The study of deformation monitoring models that are suitable for the initial stage of impoundment of high arch dams has important theoretical significance and practical value.The main work of this paper includes:(1)Sorting out the research results of the safety monitoring of 300m-level ultra-high arch dams in recent years,and introducing different analysis methods,such as multiple regression analysis,stepwise regression analysis,principal component analysis and radial basis neural network,including their principles and results evaluation criteria.What were summarized is that the modeling steps and implementation methods of the deformation monitoring model for a 300m-level ultra-high arch dam.(2)The paper analyzed the actual monitoring data of a super-high concrete double-curved arch dam at the initial stage of water storage,and used statistical regression analysis and radial basis neural network to establish the high arch dam deformation safety monitoring statistical model and the model of radial basis neural network at the initial stage of water storage.It focused on the aging components forms of expression and the impact of parameters of the aging components on the prediction performance of the model.And based on the factor set composed of the measured temperature,the principal component analysis is introduced to establish a dam deformation monitoring model based on the principal component analysis of the temperature component.In addition,the difference between Hydrostatic-Temperature-Time(HTT)model and the Hydrostatic-Season-Time(HST)model is compared.The results show that:(1)compared with the traditional statistical regression model,the radial basis network model has higher training and prediction accuracy in the deformation monitoring modeling of the super high arch dam at the initial stage of water storage.(2)The aging component occupies a large proportion of the dam deformation at the initial stage of water storage.What’s more,what were recommended is that the reasonable expression form of the aging deformation component of the high arch dam at the initial stage of water storage and the reasonable value range of related parameters.(3)The measured temperature can reflect the influence of the unstable temperature field on the deformation of the dam at the initial stage of impoundment of the dam,and can achieve better performance than the seasonal temperature model.