Time Series Analysis Of Landslide Displacement Based On Grey Wolf Optimization Algorithm

The stability of the open-pit mine slope directly affects whether the mine can carry out safe production work,therefore,it is important to study the factors affecting the landslide displacement.In this paper,the landslide in Gongchangling Open-pit Mine is used as the research object.The method of combining on-site survey,landslide deformation monitoring and numerical simulation is used to compare the actual and predicted values of the slope displacement using artificial intelligence algorithms.Through a variety of optimization algorithms compared with the classic intelligent algorithm,the obtained data and images are compared and verified.The main research work and results are as follows:(1)On the basis of geological investigation and slope slip analysis,the monitoring network,main monitoring line and auxiliary monitoring line of the southern slope of Gongchangling are constructed.The time-displacement curve of point H29 Y axis in Hejia stope is obtained,and the characteristics of the curve and the landslide mechanism are analyzed.(2)For the acquired data,there are still some abnormal data and some missing data.The abnormal data are eliminated by t-test method,and the missing data are supplemented by piecewise linear interpolation method,which improves the preliminary acquired slope displacement monitoring data.Finally,the tangent angle method is used to analyze the improved data,and four different prediction methods are obtained: the whole data,the last 3/4 data,the last 1/2 data and the last 1/4 data.Using time series analysis method and formula X(t)=alpha(t)+beta(t)+gamma(t),the monitoring data of point H29 Y axis in Hojia mining area of Gongchangling Open-pit Mine are divided into trend term and periodic term.The trend term is predicted by particle swarm optimization and the period term is predicted by grey wolf optimization algorithm.By analyzing the SD value in the calculation process,the error values of the two algorithms are obtained,and a more accurate conclusion is drawn by comparing the grey wolf optimization algorithm.Finally,the sum of trend and periodic results is obtained by the two algorithms,and the total S-t curve is obtained.For the prediction model,the feasibility of the prediction is verified by field investigation.