Analysis Of Monitoring Data And Stability Evaluation Of Wangjiapo Landslide

Landslide monitoring,analysis and evaluation have important practical significance to the effective prevention and control of landslide.Taking wangjiapo landslide in baqiao district of xi'an city as the research object.A monitoring network is designed and laid in the landslide area for dynamic monitoring of the landslide.Elman neural network algorithm and finite element numerical simulation were used to explore the characteristics and rules of landslide deformation,development trend and landslide stability.In this process,some knowledge and conclusions are obtained:(1)Based on beidou BDS technology,a monitoring network was established in wangjiapo landslide area.The Base station is built in a stable area on the outside of the landslide.The monitoring points of beidou no.2 and no.3 BDS are arranged at the edge and inside of the residential area in the middle section of the landslide area.Two displacement sensors are arranged on both sides of a ground crack in the middle residential area.Using uav monitoring dynamic and overallmonitoring of the development of the landslide;(2)According to the monitoring data,the following conclusions are drawn:Three stages of deformation activities of wangjiapo landslide are proposed(initial deformation stage,deformation acceleration stage and deceleration deformation stage).Expounding the basic characteristics of the three stages.Thinking wangjiapo landslide is in the third stage.Although the deformation activity of the wangjiapo landslide only decreases slightly,it is stable and safe.(3)Based on Elman neural network algorithm,the prediction model of sliding trend of wangjiapo landslide was established.In this model,MAE,MAPE and RMS were used for evaluation.Come to a conclusion:?with the prediction time length unchanged,the prediction error value can be reduced to 1mm with the increase of the training data;Under the condition that the amount of training data remains unchanged,the prediction error value decreases to 0.9mm as the predicted time length decreases;?The results show that the Elman neural network algorithm is suitable for the prediction of landslide development trend,and the prediction effect is ideal.(4)Nmerical simulation of wangjiapo landslide was conducted using ANSYS finite element software.Using two kinds of strength reduction methods that is equal proportion and unequal proportion to built the landslide model.the landslid estability coefficients under two different conditions are obtained.Provide technical support for subsequent landslide control projects.It is calculated that the stability coefficient of wangjiapo landslide is 1.24 under the condition of equal scale reduction and 1.25 under the condition of unequal scale reduction.