Dynamic Prediction Of Horizontal Displacement Of The Slope Top Of The Rock Deep Foundation Pit And Security Control Of The Pit

With the development of urban construction, high-rise buildings and municipalworks are great emerging and limited land resources has been in the city can not meetpeople’s needs, followed by large-scale construction of high-rise buildings andunderground engineering; Meanwhile,development and utilization of undergroundspace in general in the city center, the construction process has great influence on thesurrounding buildings and structures, so it must be real-time monitoring of theexcavation process of displacement and groundwater changes. This particular case ofdeep foundation pit has increasingly become a widespread research topic of greatsignificance.More focused on the deformation of deep excavation in the selection of thesupporting structure, calculation methods and the excavation process control, however,it is lack of the research on uncertainty of parameter of rock and soil and real-timeforecasting on pit deformation, especially in the rock pit. Based on the deep foundationpit engineering in the Jiefangbei Guotai Square of Yuzhong District, Chongqing City,from the construction process of dynamic forecasting and security control of the rock pitdeformation, combined with the BP neural network and the cusp catastrophe theory, asuggestion has been made for rock pit of dynamic construction and safety evaluationunder every working condition, providing a reference for future similar excavationengineering. Main job is as follows:①For the actual geological conditions, using finite difference software FLAC3D,Stratified excavation of the rock pit to simulate the whole process of foundation pitsupport and monitoring the top of the pile deformation.②According to the geological conditions of Chongqing, to collect field samples,using the BP neural network, training network to post calculation, through monitoringof slope top horizontal displacement, inversion to get the rock and soil parameters.Inversion parameters within the range of inversion of sample values in order to increasethe accuracy.Based on inversion parameter model to predict displacement value of theslope top, and compared with the actual values.③According to the actual date, establishing the pit horizontal deformation for theaging component of cusp catastrophe mechanical model, safety evaluation formonitoring data under every condition, and carry out next condition. In this paper, researches show that the aging component model adaptability, and high precision.④This paper uses MATLAB software, independent write BP neural network andleast square method analysis software, to accelerate the speed and accuracy of thecalculation and analysis. At the same time, These software combined with finitedifference software“FLAC3D” to analyze the specific content.In the study, the rule reflected by numerical and monitored results are insubstantial agreement, which is elaborated by theoretical calculation result. The thesiscould also provide reference for optimal design and construction.