Deformation Prediction And Early Warning Of Jihe High Slope

Landslide is one of the most common natural disasters along the expressway and the occurrence frequency of landslide is higher.Its seriously threatening people's life safety.At present,the stability analysis and safety monitoring and early warning of the slope have attracted extensive attention.But it is urgent to study these problems that the existed single prediction method cannot effectively predict the continuous deformation displacement and the optimization process between the combinatorial model algorithm makes it difficult to accurately predict the slope landslide threshold.Therefore,for the high-precision prediction and early warning algorithm in specific application scenarios,this paper introduces the radial basis function(RBF)neural network model to predict the slope displacement based on the previous research.Taking the high slope of Linfen-Jihe expressway and selecting cohesion,internal friction angle,temperature,humidity,rainfall as input factors.The radial basis function neural network-particle swarm optimization(RBF-PSO)algorithm is used to predict the displacement of this high slope.Finally,the load response rate algorithm is introduced to analyze the slope early warning.The main contents of this article are:(1)The RBF neural network model algorithm is established for data training and analysis,and the structure of the network is shown as 5-14-1.Through sample training and prediction,it is found that the model has large errors and low accuracy in the displacement analysis and prediction of the Jihe expressway high slope.Therefore,the PSO algorithm is introduced to optimize the RBF neural network.(2)The RBF-PSO algorithm is proposed and a slope displacement prediction model based on this model is created.Combined with the deformation characteristics of this slope,a comprehensive analysis was carried out to analyze the displacement curve of the slope in three directions of X,Y,and Z under three-dimensional rectangular coordinates.Theexperimental results based on the sample data of monitoring station BP3-2-1show that the average relative errors of the RBF-PSO algorithm in the X,Y,and Z directions are 2.63%,0.24%,and 0.44%,respectively,which are less than that of the BP-PSO algorithm,GM(1,1)algorithm and gray Verhulst algorithm.At the same time,the experimental results show that after the RBF neural network is optimized by the PSO algorithm,the prediction accuracy in the three directions of X,Y,and Z increased by 25.3%,8.9%,and 3.29%,respectively.(3)Taking the high slope of the Linfen-Jihe expressway as the background,the load response rate algorithm is selected as the judgment condition for judging whether the slope is dangerous.And according to the actual situation,it is determined that when the slope is in a stable state,the threshold range of the load response rate of the slope in the X,Y,and Z directions are [0.616,0.959],[0.618,0.960] and [0.592,0.921],respectively.In order to predict the future load response rate,it's need to combined with the RBF-PSO algorithm to predict the future displacement of the slope monitoring station BP3-2-1.Finally,under the GIS environment,the comprehensive factor response rate of the entire slope is converted into grids,and the susceptibility of the slope to the landslide is calculated.So as to make an early warning analysis of the future stability of the slope.