Research On Stability Analysis And Prediction Of Tailings Ponds Based On Multi-source Heterogeneous Dat

Tailings dams are at risk of collapse,and once they fail,they are prone to cause serious accidents and seriously damage the safety of people’s lives and property.Therefore,the study of tailing pond deformation evolution law,tailing pond stability analysis,and prediction is of great significance for tailing dam disaster prevention and control.In this paper,we analyzed the evolution law of tailing pond deformation in all directions based on GNSS monitoring data,analyzed the specific subsidence of the tailing pond based on SBAS-In SAR technology,simulated the stability of Chenkeng tailing pond dam using Geo-Studio software,constructed a real-time analysis module for tailing dam stability based on Swedish arc method,and finally constructed a PSO-SVR-KF based phreatic line The prediction model was constructed to predict the stability of the tailings dam by the Swedish arc method,which provides a reference for tailings pond disaster prevention and control.The research results are summarized as follows.(1)Based on the GNSS monitoring data of the tailings pond,the evolution law of deformation in each direction of the tailings pond was studied,and the specific subsidence of the tailings pond was analyzed based on SBAS-In SAR technology using Sentinel-1A image data.The results show that the displacement of Chenkeng tailings pond is stable in the X direction,the displacement of the south bank is dominant in the Y direction,and there is an obvious settlement in the Z direction,which is consistent with the general deformation law of tailings dams;the main settlement areas of the tailings pond are the tailings pond accumulation dam and the dry beach area at the top of the tailings pond,with a maximum settlement rate of 88.6 mm/a and a surface deformation rate of-88.6～30.7 mm/a;from June 28,2022,onward,The settlement of the top part of the accumulation dam in the tailings pond area keeps increasing,and the settlement range begins to spread outward from the center and eventually covers the whole tailings pond;there are periodic fluctuations in the tailings dam type variable,and the fluctuation of the shape variable is positively correlated with the amount of precipitation.(2)Geo-Studio software was used to simulate the stability of tailings dams in the Chenkeng tailings pond under normal conditions,flood conditions,and special conditions.The results show that the stability coefficients of the three working conditions are 1.708182,1.704205,and 1.347142 for the Swedish arc method and 1.772625,1.765512,and 1.408168 for the simplified Bishop method respectively;the stability coefficients of the three working conditions solved by the Swedish arc method and the simplified Bishop method meet the safety requirements.The overall stability of the Chenkeng tailing pond dam is safe;the stability settlement results of the simplified Bishop method are on the high side and should be considered in conjunction with other calculation methods in the actual engineering application.(3)Based on the Swedish arc method,a real-time analysis module of tailing dam stability was constructed,and the real-time information on the tailing pond infiltration line was determined by the actual measurement data of the pore water pressure meter,and the stability analysis was carried out for the calculation case and the Chenkeng tailing pond normal working condition and flood working condition.The results show that the minimum safety coefficients of the two working conditions of the case are 1.189 and 1.185,which increase by 1.9%and 3.8%respectively compared with the SLOPE solution;the minimum safety coefficients of the two working conditions of the Chenkeng tailings pond are 1.655224 and 1.654277,which decrease by 3.1%and 2.9%respectively compared with the SLOPE solution;the error between this module solution and the Geo-The error between this module solution and the SLOPE calculation results is less than 5%,and the results are in good agreement.(4)Based on the multi-source monitoring data of tailings ponds,a PSO-SVR-KF-based phreatic line prediction model was constructed,and the phreatic line prediction values were used to predict the stability of tailings dams by the Swedish circular method.The results show that compared with the conventional PSO-SVR,the accuracy of the PSO-SVR-KF-based phreatic line prediction model is better,with a 3.84%increase in R~2 and an 85.83%decrease in MSE for JR10-1;a 5.13%increase in R~2 and a 64.52%decrease in MSE for JR2-1;a 3.29%increase in R~2 and a decrease in MSE for JR6-1 The prediction accuracy was significantly improved;Compared with the actual results of stability analysis on October 9,2022,October 24,2022,and November 9,2022,the errors of the stability analysis results of the predicted values are 0.0002%,-0.0004%,and-0.0005%,respectively,which are in high agreement with the stability analysis results of the measured values.