Research On Comprehensive Early Warning Method Of Slope Stability Based On Beidou And Microseismic Monitoring System

The instable risky slopes are extremely harmful,so it is particularly important to use the intelligent online monitoring system to monitor the risk status of the slope in real time.However,it is often difficult to accurately reflect the trend of the slope by a single safety monitoring method,and it is necessary to monitor the surface changes and internal activities of the slope simultaneously.Therefore,a “ Beidou + microseismic ” slope safety comprehensive monitoring program has been established.It is necessary to improve the existing research methods of analyzing and extracting effective information of slope monitoring signals in order to avoid danger before landslide disaster occur and hierarchical control of risk slopes.The mine slope is taken as the main research object,on the one hand,based on the slope displacement data obtained by the Beidou monitoring system,the accuracy of slope displacement prediction was improved by further studying of slope displacement prediction method.On the other hand,the accuracy of picking up primary microseismic signals is improved by studying the denoising method and the primary picking method of microseismic signals.The processing results of the above improved methods provide accurate data sources for the establishment and application of slope safety early warning criteria.The main research contents and achievements are as follows:(1)For the difficulty in predicting slope displacement induced by deformation,the EEMD-PSO-ELM model for slope displacement prediction has been established by combining ensemble empirical mode decomposition(EEMD)and extreme learning machine optimized with particle swarm optimization(PSO-ELM).Taking the No.5 monitoring point of Pangang Group limestone mine as an example,the original data are denoised by wavelet in the first step.The denoising data are decomposed into the fluctuating displacement and the trending displacement by EEMD method,and then the PSO-ELM optimization model is used to predict the displacement in the next period.To obtain the cumulative displacement prediction results of the slope,the two displacement prediction results are overlayed.The prediction results are compared and analyzed.The mean relative error,the root mean square error and the goodness of fit are 0.15%,2.98%and 0.9999 respectively by the displacement prediction of the EEMD-PSO-ELM model.In addition,the model is used to predict the displacement of monitoring points 8 and 9.The above results show the accuracy and applicability of the model,which has certain guiding significance for the early warning of landslide disaster.(2)In the process of microseismic signal acquisition,there are a large number of interference signals with different frequencies,which makes it difficult to pick up the primary arrival of the signal.A method based on empirical wavelet transform(EWT)combined with EWT component reconstruction threshold rule and singular value decomposition(SVD)technology for signal denoising is proposed.The results show that the reconstruction of inherent modal components with a correlation coefficient greater than 0.3and a variance contribution rate greater than 15% has a better denoising effect for high signal-to-noise ratio signals.For the low signal-to-noise ratio signal,based on the high signal-to-noise ratio denoising method,a new denoising method using SVD to denoise the high-frequency components and reconstructing with low-frequency effective components decomposed by EWT is proposed.The signal-to-noise ratio of the structural signals with different signal-to-noise ratio and the actual microseismic signals is obviously improved,which verifies the effectiveness of the denoising method.(3)Based on the EWT signal denoising,a method of picking up primary arrival of microseismic signal is proposed.The common primary arrival picking algorithms are analyzed,the modified energy ratio(MER)and fractal-dimension(FD)with obvious advantages are used to pick up the primary arrival time for the denoising signal finally.The signal denoising method based on EWT is combined with the MER method and the FD method to pick up the primary arrival time of the actual microseismic signals respectively.The test results show that the model is highly accurate in picking up primary arrival time of effective signals and has a strong applicability to signals of different data quantities.(4)Based on the three-phase law of slope deformation and the improved displacement tangent angle method,the corresponding relationship among displacement deformation stage,tangent angle threshold,displacement rate and early warning level,and the warning criterion of displacement in Panzhihua limestone mine are established.The number of microseismic events and the ratio of the event rate,the microseismic event density and the b value calculated from the magnitude-frequency relationship are used as the microseismic early warning criteria.Finally,combining the displacement early warning criterion and the microseismic early warning criterion,the four-grade early warning criterion for the slope is established comprehensively.The established warning criteria were used to analyze and verify the monitoring points 1,5,8,and 9 displacement data of the limestone mine slope.The results show that the slope is in a safety status,which is consistent with the actual situation.