Study On Monitoring And Forecasting Method Of Surface Subsidence In Mining Area

Mining in mining areas is often carried out in the form of large area,high intensity and multiple working faces.Long-term and high-intensity mining makes the area of mined-out area larger and larger.With the passage of time,a large area of mined-out area will cause a certain degree of surface subsidence in the mining area.The subsidence of mining area often causes geological disasters such as ground collapse,road cracking and landslides,which seriously threatens the rational development of mining area and the safety and stability of mining environment.Therefore,it is of great significance for disaster prevention and environmental restoration of mining areas to use InSAR technology to monitor mining area subsidence and extract mining area subsidence characteristics.In this paper,Yaojie mining area is taken as the research object,using Sentinel-1A image and based on SBAS-InSAR technology,the subsidence characteristic information of mining area,such as annual average subsidence rate,cumulative subsidence and time series subsidence,is obtained.Based on the monitoring results of SBAS-InSAR,combined with meteorological data such as rainfall,sunshine duration,temperature and humidity,the influence of meteorological factors on the settlement of mining area is analyzed.Based on the time series settlement,four models are constructed to predict the settlement in the mining area,and various indexes are used to evaluate the prediction accuracy and determine the optimal prediction model.The main work and conclusions of this paper are as follows:(1)Using 47 Sentinel-1A images covering the study area from 2018 to 2021,SBAS-InSAR technology was used to obtain the annual average subsidence rate,cumulative subsidence,time series subsidence and other mining area subsidence characteristics information,and the mining area subsidence characteristics were analyzed from their respective perspectives;The settlement rate and settlement amount in the mining area are divided according to the threshold value,and the settlement degree in the area is systematically analyzed.The results show that most of the study area is in a stable state,and there are three obvious deformation areas.The average annual settlement rate shows that the maximum settlement rate of Yaojie No.3 Mine is 51.2mm/a,that of Jinhe Coal Mine is 124.9mm/a and that of Haishiwan Coal Mine is 157mm/a..The cumulative settlement shows that the cumulative settlement of Yaojie No.3 Mine is 310.82mm,that of Jinhe Coal Mine is470.25mm,and that of Haishiwan Coal Mine is 681.82 mm.Referring to the classification standard of settlement grade in related research in academic circles,the mining area settlement is divided into mild,severe and extremely severe with the cumulative settlement of300mm and 600mm as thresholds;According to the calculation,the area of light subsidence area is 4.115km~2,the area of heavy subsidence area is 2.811km~2,and the area of extremely heavy subsidence area is 0.106km~2.(2)Based on the monitoring data of mining subsidence,combined with meteorological data such as rainfall,sunshine duration,temperature and humidity,the relationship between meteorological data and mining subsidence is studied from the perspectives of single factor and multiple factors.The results show that the influence of rainfall on the settlement of mining area is different and lagging,and the lag period is about one month.The influence of sunshine duration on the settlement of mining area has seasonal characteristics.From December to January every year,the mining area is in the freezing period,and the settlement of mining area changes little.From February to March every year,the mining area is in the melting period,and the sunshine duration increases and the settlement of mining area increases.The temperature has a seasonal influence on the settlement of the mining area.From February to April every year,with the gradual increase of the temperature,the settlement of the mining area becomes larger.Humidity has little effect on the settlement of mining area;The influence degree of meteorological factors on mining subsidence from high to low is:rainfall,sunshine duration and temperature.(3)Based on the 24 time series settlement of deformation centers in each mining area from 2020 to 2021,four prediction models,GM(2,1),BP neural network,PSO-SVR and LSTM,are constructed to predict the short-term settlement trend in the mining area.The prediction accuracy of each model is evaluated by four indicators,such as MAE value,RMSE value,overall prediction accuracy of the model and prediction accuracy under different time lengths,and finally the best one is selected.Comparing the MAE value and RMSE value of each model,it is found that the MAE value and RMSE value of LSTM model are less than3mm,and the prediction accuracy is the highest.Comparing the overall prediction accuracy of the four models,it is found that the overall prediction accuracy of GM(2,1),BP neural network,PSO-SVR and LSTM are 50.51%,66.58%,78.82%and 90.49%respectively,and the overall prediction accuracy of LSTM model is higher than other models.When comparing the prediction accuracy of each model under different prediction time,it is found that when the prediction time is 4 months,the prediction accuracy of LSTM model is 88.02%,which is the highest among the four models,and it can be used as the first choice model for predicting the later settlement in this study area.