Deformation Prediction And Stability Analysis Of Multi-level High Fill Slope In Loess Area

Our country is a country with many mountains and rivers.Especially in Lanzhou,the land area is far from enough to meet people's living needs.Therefore,there have been many cases of cutting mountains and reclaiming land.The construction of the side slope.Therefore,in order to prevent landslide disasters,the prevention and control of slopes is the top priority.This article is based on a multi-level high-fill slope project in Lanzhou City.The slope is a high-fill multi-level slope produced by cutting the mountain to create land.Although the slope has been supported and reinforced,However,due to the impact of the slope on the residents and buildings at the bottom of the slope,the monitoring of the slope has practical significance.Based on the theory and the monitoring data of the slope horizontal displacement,this paper conducts research and discussion on the stability analysis of the slope through the establishment of a model:1)In combination with the actual project,the slope of the project is monitored for deformation.By determining the monitoring items,the monitoring is carried out twice a day.In order to ensure the real-time monitoring of the displacement,the interval is12 hours each time.The monitoring software is then used for 24-hour unmanned real-time monitoring of the slope to realize automatic monitoring.2)Obtain the monitoring displacement data according to the on-site monitoring method,and based on the slope deformation data,through theoretical derivation,establish the high-fill slope gray prediction model,and use MATLAB to realize the program writing,substituting the data into the predicted data,and then correct The prepared prediction model is verified.3)The uncertainty and randomness of the slope rock and soil parameters are clarified,and it is necessary to inverse the rock and soil parameters.Through sensitivity analysis,calculate the gray correlation coefficients of various soil parameters,sort their sizes,and determine the soil parameters to be inverted;at the same time,with the help of the uniform design table,different parameter groups are combined to provide training for the BP network.Group,bring the training group into the BP neural network to train the prediction model and verify it.4)Finally,determine the slope stability evaluation method,and use FLAC3D software to model the slope,substitute the parameter groups that have been obtained,and obtain different displacement values corresponding to different parameter groups,debug the parameter inversion model used,and perform authenticating.Substitute the predicted displacement into the verified parameter inversion model to obtain the rock and soil parameters at the future time,and substitute the rock and soil parameters into the FLAC^3D model to study the stability of the slope at the moment,and compare and analyze the changes caused by time Safety factor and changes in displacement.Through the above research,the occurrence of slope disasters can be prevented more effectively.The research provides certain reference opinions for similar local geotechnical engineering,and has certain research significance at the same time.