Research On The Synthetical Prediction Method Of Slope Deformation

Slope instability is one of common geological disasters, a key method to solve the problem of geological disasters is to predict and forecast the disaster accurately so as to take timely measures to avoid personnel and property losses. Landslide prediction can be divided into two aspects, time and space prediction. Based on the slope deformation monitoring data analysis and calculation can predict the trend of slope deformation and provide the basis for geological hazard analysis; through the analysis of the geological exploration data and the slope stability evaluation, can provide the basis for spatial prediction.Based on the Zijingshan gold-copper mine as engineering background, with GPS slope deformation monitoring data and on the basis of engineering geological exploration data for slope deformation forecast early warning research, the main contents are as follows:(1) Collect all kinds of Zijingshan gold-copper mine geological exploration data and the two-dimension data, extracting the relevant information in CAD file and integrating information;based on ArcGis platform, establish Zijingshan’s three-dimension geological mode respectively in GIS, sketchUP, lumion; At the same time, using spatial analysis module of ArcGis, space for the Zijingshan gold-copper mine is calculated and extract the relevant information, establish the geographic information system database, implementation of mining information management system of 3-dimension digital visualization and big data information management.(2) Took influence factors of slope stability into consideration, such as formation lithology, geological structure, weathering and alteration, slop, aspect, terrain curvature, hydrogeological condition, surface runoff, existing large project, road engineering, vegetation cover and temperature differences. Then the three kinds weights (the fuzzy-floor, the most likely value and the fuzzy-ceiling) are calculated by the improved fuzzy analytic hierarchy process. In order to use the spatial analysis module so as can weight stacking various factors of raster layers, geographic database of one mine is built based on ArcGis. Different kinds of slope instability have been evaluated by adjusting fuzzy weight. It provides a new thought of slope stability evaluation.(3) Predict the deformation trend of displacement monitoring data in unstable area. Based on the understanding that every residual error is the amendment for the predicted value, according to the time series method, establish the ARMA (n, n) mathematical model for the slope deformation monitoring data on the platform of Matlab. This model setted the order with the advanced Pandit-Wu method, and calculated parameters with Quasi-Newton(DFP、BFGS) of the Engineering optimization method, and predicted the deformation trend of the slope displacement dynamically.(4) Treat the slope system as a chaotic system, on the basis of the fractal theory, selected different delay time and reset phase space to the original monitoring data sequence to obtain a new sequence;it took the linear optimization method, which on the condition of the minimum sum of squares of prediction residue error, to calculate the weights of different methods’ predictions, such as ARMA (n,n-1) time series、GM(1,1) grey theory and Quasi-Newt on-based ARMA (n,n),and made first synthetical prediction;then made second synthetical prediction on the basis of different delay times’ synthetical results.