Reinforcement Design And Visualization Of Micro-pile For Accumulation Layer Landslide

With the continuous development of information technology,the modernization of the engineering industry has also received great attention.Not only does it break through the two-dimensional limitation of traditional CAD drawing,the BIM platform,but also provides a reference for actual engineering and actual natural disaster management by creating a visual model.Likewise,analyzing data to provide guidance for future engineering practice.As a safe,reliable and low-cost landslide support structure,micro-piles are widely used in the treatment of accumulation layer landslides in southern Shaanxi due to their convenience and flexibility.It is not mature enough to use neural network to predict the law of slope sliding.The main research contents of this paper are as follows:(1)Taking the landslide in Yuandun Town,Mian County,Southern Shaanxi as an example,research and obtain various parameters to realize project information management,and realize that the entire project components and related family components can be called,and determined by the method of grabbing the elevation and point number positioning and mapping Contour topographic map,and build a site model based on this link Revit,by summarizing the previous scholars' design experience for landslide micro-pile,carry out the relevant model parametric design,and finally establish three types of two different micro-pile groups.To realize the reinforcement design of the landslide model.(2)Build a BP neural network prediction model under the MATLAB environment.The related theories of neural network were analyzed,and a neural network prediction model of landslide deformation was created by taking(Zhu Baolong)the actual monitoring displacement data of the K103 expressway landslide reinforced with micro piles as the reference group and training samples,and the current step comparison and analysis were carried out.Prediction and analysis of the future step,and obtain the deformation displacement prediction data of the BP network based on this.(3)The finite element numerical simulation of the K103 landslide is carried out through the ABAQUS platform,and the safety factor is obtained through the finite element strength reduction method to determine the necessity of using micro-pile to reinforce the landslide.At the same time,the finite element simulation calculation data is obtained.Compared with the predicted data of BP neural network,it can be concluded that the predicted value of BP neural network has high accuracy and good coincidence with the actual displacement curve,while ABAQUS has the characteristics of low accuracy but strong predictability.Therefore,ABAQUS simulation can make forward-looking predictions before the start of the project,and can prepare and control the possible accidents in advance.The prediction results of MATLAB neural network can better provide data support for landslide slip early warning under the condition that the project has preliminary data.