A slope stability analysis method and system based on dynamic monitoring data

Abstract

The present invention relates to the field of slope safety monitoring, and particularly to a method and system for analyzing slope stability based on dynamic monitoring data. A slope stability analysis system based on dynamic monitoring data includes: a dynamic monitoring data acquisition module, a slope state discrimination module, and a slope collapse risk prediction module. The present invention adjusts the warning thresholds of some monitoring parameters in each data group specifically by statistically calculating the maximum temperature difference and rainfall in the target slope area at the beginning of each environmental monitoring cycle; this dynamic adjustment mechanism fully considers the influence of environmental factors on slope stability monitoring parameters, makes the warning threshold more in line with the actual working conditions, effectively improves the accuracy and reliability of slope stability monitoring, avoids false alarms or missed alarms caused by fixed thresholds, and provides a more accurate monitoring basis for slope safety protection.

Description

Technical Field

[0001] The present invention relates to the field of slope safety monitoring, and particularly to a slope stability analysis method and system based on dynamic monitoring data. Background Art

[0002] Slope stability analysis is a key link in geological disaster prevention and engineering construction, and its results are directly related to the safety of people's lives and property and the stable operation of infrastructure. Traditional slope stability analysis methods mainly rely on static geological exploration data and limited monitoring information. This method has many limitations when facing complex and changing natural environments and engineering conditions.

[0003] With the continuous progress of technology, slope monitoring technology has been significantly improved, and a variety of monitoring means have emerged, which can obtain multi-faceted data such as displacement, stress, and hydrology of slopes in real time. However, these monitoring data are often scattered and complex, lacking effective integration and analysis methods, and it is difficult to fully utilize their value in slope stability assessment. At the same time, environmental factors (such as temperature changes and rainfall) have an undeniable impact on slope stability, but traditional methods are difficult to fully incorporate these factors into the analysis system, resulting in key parameters such as early warning thresholds being unable to be dynamically adjusted according to actual environmental conditions, reducing the accuracy and reliability of monitoring results.

[0004] In addition, slope stability analysis not only needs to focus on the current state, but also needs to predict future risks. Existing technologies have deficiencies in multi-source data fusion and in-depth analysis, making it difficult to accurately grasp the complex relationships between each monitoring data group and slope collapse, and unable to provide strong support for risk prediction. To solve the above problems, the present invention proposes a slope stability analysis method and system based on dynamic monitoring data. Summary of the Invention

[0005] The present invention adjusts the early warning thresholds of some monitoring parameters in each data group specifically by statistically calculating the maximum temperature difference and rainfall in the target slope area at the beginning of each environmental monitoring cycle. This dynamic adjustment mechanism fully considers the influence of environmental factors on slope stability monitoring parameters, makes the early warning threshold more in line with the actual working conditions, effectively improves the accuracy and reliability of slope stability monitoring, avoids false alarms or missed alarms caused by fixed thresholds, and provides a more accurate monitoring basis for slope safety protection.

[0006] A slope stability analysis method and system based on dynamic monitoring data, comprising:

[0007] Collect the first slope data set, the second slope data set, the third slope data set, and the fourth slope data set of the target slope area at the current monitoring time point; the first slope data set includes surface horizontal displacement, surface vertical displacement, and deep stratified displacement; the second slope data set includes acoustic emission energy and microseismic wave propagation time difference; the third slope data set includes rock mass surface load stress and stress distribution gradient; the fourth slope data set includes pore water pressure and seepage rate;

[0008] Set the environmental monitoring period. At the beginning of the current environmental monitoring period, calculate the maximum temperature difference of the target slope area in the most recent 12 hours, and obtain the rainfall in the previous environmental monitoring period. Subsequently, adjust the warning thresholds of some monitoring parameters in each data set; then, during the current environmental monitoring period, determine whether all the obtained monitoring parameters meet the corresponding adjusted warning thresholds. If so, no operation is performed; if not, execute the warning response measures;

[0009] Calculate the correlation degree of each data set with slope collapse, and then calculate and obtain the influence weight of each data set; at the beginning of the current environmental monitoring period, apply the slope collapse risk prediction model to obtain the slope collapse risk prediction result; apply the slope collapse risk prediction result and execute the corresponding slope collapse prevention measures.

[0010] Preferably, the operation of adjusting the warning thresholds of some monitoring parameters in each data set is as follows:

[0011] For all monitoring parameters of each data set , i = 1, 2,..., 9; to correspond to surface horizontal displacement, surface vertical displacement, deep stratified displacement, acoustic emission energy, microseismic wave propagation time difference, rock mass surface load stress, stress distribution gradient, pore water pressure, and seepage rate in sequence; set the conventional warning threshold of each monitoring parameter ;

[0012] Based on the maximum temperature difference and rainfall obtained in the current environmental monitoring period, for surface horizontal displacement , use the formula to calculate and obtain the adjusted warning threshold of surface horizontal displacement ; where is the surface displacement thermal expansion dynamic factor;

[0013] For rock mass surface load stress , use the formula to calculate and obtain the adjusted warning threshold Adjusted warning threshold ; where is the surface stress temperature correction coefficient;

[0014] For the stress distribution gradient , use the formula to calculate and obtain the stress distribution gradient Adjusted warning threshold ; where is the temperature gradient correction coefficient;

[0015] For the deep stratified displacement , use the formula to calculate and obtain the deep stratified displacement Adjusted warning threshold ; where is the deep displacement rainfall accumulation coefficient;

[0016] For the acoustic emission energy , use the formula to calculate and obtain the acoustic emission energy Adjusted warning threshold ; where is the acoustic emission energy level water pressure coupling coefficient;

[0017] For the pore water pressure , use the formula to calculate and obtain the pore water pressure Adjusted warning threshold ; where is the pore water pressure rainfall response coefficient;

[0018] For the seepage rate , use the formula to calculate and obtain the seepage rate Adjusted warning threshold ; where is the seepage rate rainfall multiplication coefficient.

[0019] Preferably, calculate the correlation degree of each data group with the slope collapse, and the specific operation is as follows:

[0020] Obtain several slope reference samples, based on the monitoring parameters of all slope reference samples and the slope collapse state , = 1, 2,..., ; represents the number of slope reference samples;

[0021] Based on the monitoring parameters included in the slope reference samples , calculate and obtain the standardized monitoring parameters ;

[0022] Set multiple grid combinations , traverse any one of the grid combinations , calculate and obtain the monitoring parameters when applying the current grid combination and the mutual information coefficient with slope collapse ;

[0023] Based on the mutual information coefficients corresponding to all the obtained grid combinations, select the maximum mutual information coefficient as the monitoring parameter and the maximum information coefficient with slope collapse ;

[0024] Take to the maximum value in it as the correlation degree between the first slope data group and slope collapse ;

[0025] Take to the maximum value in it as the correlation degree between the second slope data group and slope collapse ;

[0026] Take to the maximum value in it as the correlation degree between the third slope data group and slope collapse ;

[0027] Take to the maximum value in it as the correlation degree between the fourth slope data group and slope collapse .

[0028] Preferably, calculate and obtain the influence weights of each data group, and the specific operation is as follows:

[0029] Use the formula to calculate and obtain the influence weights of each data group , = 1, 2, 3, 4.

[0030] Preferably, apply the slope collapse risk prediction model to obtain the slope collapse risk prediction result, and the specific operation is as follows:

[0031] At the beginning of the current environmental monitoring period, the first slope data set, the second slope data set, the third slope data set, and the fourth slope data set obtained in the previous environmental monitoring period are sorted by time to form the first time series set, the second time series set, the third time series set, and the fourth time series set respectively; the obtained time series sets and the influence weights of each data set are used as the input of the slope collapse risk prediction model, and the slope collapse risk prediction result is output.

[0032] Preferably, the slope collapse risk prediction model is established based on LSTM and includes an input layer, an LSTM layer, a feature weighted fusion layer, a fully connected layer, and an output layer.

[0033] Preferably, for the training of the slope collapse risk prediction model, the specific operations are as follows:

[0034] Obtain a number of model training samples with labeled slope collapse risk prediction results; divide all the obtained model training samples into a training set and a validation set, use the training set to train the slope collapse risk prediction model with initialized parameters, and then verify the slope collapse risk prediction model through the validation set to obtain a verification result; set training conditions, and judge whether the obtained verification result meets the training conditions. If so, output the trained slope collapse risk prediction model.

[0035] A slope stability analysis system based on dynamic monitoring data includes:

[0036] A dynamic monitoring data acquisition module for collecting the first slope data set, the second slope data set, the third slope data set, and the fourth slope data set of the target slope area at any monitoring time point;

[0037] A slope state discrimination module includes an early warning threshold adjustment unit and a discrimination unit; the early warning threshold adjustment unit is used to adjust the early warning thresholds of some monitoring parameters in each data set at the beginning of any environmental monitoring period; the discrimination unit is used to judge whether all the obtained monitoring parameters meet the corresponding adjusted early warning thresholds during any environmental monitoring period. If so, no operation is performed; if not, an early warning response measure is executed;

[0038] A slope collapse risk prediction module applies the slope collapse risk prediction model to obtain the slope collapse risk prediction result; and applies the slope collapse risk prediction result to execute corresponding slope collapse prevention measures.

[0039] The present invention has the following advantages:

[0040] 1. The present invention adjusts the warning thresholds of some monitoring parameters in each data group specifically at the beginning of each environmental monitoring cycle by statistically analyzing the maximum temperature difference and rainfall in the target slope area. This dynamic adjustment mechanism fully considers the influence of environmental factors on slope stability monitoring parameters, making the warning thresholds more in line with the actual working conditions, effectively improving the accuracy and reliability of slope stability monitoring, avoiding false alarms or missed alarms caused by fixed thresholds, and providing a more accurate monitoring basis for slope safety protection.

[0041] 2. The present invention collects multiple groups of data and further calculates the correlation degree between these data groups and slope collapses to determine the influence weights of each data group, enabling the reasonable integration of information from each data group in the slope collapse risk prediction model and highlighting the role of key factors. This multi-source data fusion analysis method makes full use of the effective information in various monitoring data, improves the accuracy and reliability of slope stability prediction, and provides strong technical support for taking preventive measures in advance and ensuring slope safety. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 It is a schematic structural diagram of a slope stability analysis system based on dynamic monitoring data adopted in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0043] In order to enable those skilled in the art of the present technology to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0044] Embodiment 1, a slope stability analysis method and system based on dynamic monitoring data, comprising:

[0045] Collect the first slope data set, the second slope data set, the third slope data set, and the fourth slope data set of the target slope area at the current monitoring time point; the first slope data set includes surface horizontal displacement, surface vertical displacement, and deep stratified displacement. These displacement data can directly reflect the deformation of the slope at different depths and directions. Monitoring of surface horizontal displacement and vertical displacement can promptly detect macroscopic changes on the slope surface, such as early signs of landslides. Deep stratified displacement monitoring can further explore the displacement characteristics of each layer inside the slope, helping to understand the stability state of the internal structure of the slope and providing key basis for evaluating the overall stability of the slope; the second slope data set includes acoustic emission energy and microseismic wave propagation time difference. Acoustic emission energy monitoring can capture elastic wave signals generated by microcracks inside the rock mass, thereby realizing real-time monitoring of the internal damage evolution of the rock mass and early warning potential rock mass instability areas. Monitoring of the microseismic wave propagation time difference can analyze the changes in the propagation characteristics of seismic waves in the rock mass, and judge the integrity of the internal structure of the rock mass through the changes in the time difference, providing important reference for evaluating the dynamic response and stability of the slope; the third slope data set includes the load stress on the rock mass surface and the stress distribution gradient. Monitoring of the load stress on the rock mass surface can directly reflect the actual pressure on the slope surface and understand the mechanical state of the slope under various loads. Stress distribution gradient monitoring further reveals the uneven distribution of internal stress in the slope, helping to identify stress concentration areas, which are often high-risk areas for slope instability. By monitoring these two data, the mechanical stability of the slope can be grasped more accurately; the fourth slope data set includes pore water pressure and seepage rate. Changes in pore water pressure will directly affect the effective stress of the rock and soil mass, and thus affect the stability of the slope. Seepage rate monitoring can reflect the flow of groundwater in the slope. The seepage of groundwater will not only change the pore water pressure, but may also cause problems such as softening and strength reduction of the rock and soil mass. Therefore, monitoring of these two data is crucial for evaluating the impact of hydrogeological conditions on slope stability, especially in cases of frequent hydrogeological activities such as rainfall and snowmelt, which can help to promptly detect slope instability risks caused by water action;

[0046] Set the environmental monitoring period. At the start of the current environmental monitoring period, calculate the maximum temperature difference of the target slope area within the last 12 hours, and obtain the rainfall amount during the previous environmental monitoring period. The calculation of this temperature difference is crucial because temperature changes can have significant physical and mechanical effects on the rock and soil mass of the slope. For example, drastic temperature changes can cause the thermal expansion and contraction of the rock and soil mass, thereby triggering changes in displacement and stress. At the same time, it is also essential to obtain the rainfall amount during the previous environmental monitoring period. Rainfall is one of the key factors affecting slope stability. The infiltration of rainwater will increase the pore water pressure and reduce the shear strength of the rock and soil mass, thus increasing the risk of slope instability. Subsequently, adjust the warning thresholds of some monitoring parameters in each data group. The purpose is to consider the influence of environmental factors on the monitoring parameters, make the warning thresholds more in line with the actual working conditions, and improve the accuracy and reliability of monitoring. Subsequently, during the current environmental monitoring period, determine whether each obtained monitoring parameter meets the corresponding adjusted warning threshold. If so, no operation is performed. If not, execute warning response measures, which may include increasing the monitoring frequency, on-site investigation, expert consultation, etc., to further evaluate the stability of the slope, timely detect potential instability risks, and take corresponding preventive measures to ensure the safety and stability of the slope.

[0047] Calculate the correlation degree of each data group with slope collapse, and then calculate and obtain the influence weight of each data group. The influence weight reflects the relative importance of different data groups in the risk prediction of slope collapse. For example, if the correlation degree of the first slope data group (including surface horizontal displacement, surface vertical displacement, and deep stratified displacement) is relatively high, then its influence weight in the risk prediction model will be larger, indicating that these displacement data contribute more to the prediction of slope collapse risk. At the start of the current environmental monitoring period, apply the slope collapse risk prediction model to obtain the slope collapse risk prediction result. Apply the slope collapse risk prediction result and execute corresponding slope collapse prevention measures, such as increasing the monitoring frequency, setting warning signs, evacuating surrounding personnel, etc., to reduce the losses that may be caused by the disaster. Through this data-driven and model-predicted method, slope collapse prevention can be carried out more scientifically and effectively, ensuring the safety of people's lives and property and the stable operation of infrastructure.

[0048] Adjust the warning thresholds of some monitoring parameters in each data group. The specific operations are as follows:

[0049] For all monitoring parameters of each data group , = 1, 2, …, 9; to correspond to surface horizontal displacement, surface vertical displacement, deep stratified displacement, acoustic emission energy, microseismic wave propagation time difference, rock mass surface load stress, stress distribution gradient, pore water pressure, and seepage rate in sequence; Set each monitoring parameter Conventional warning threshold ;

[0050] Based on the maximum temperature difference obtained in the current environmental monitoring period and rainfall , for surface horizontal displacement , using the formula Calculate and obtain the surface horizontal displacement Adjusted warning threshold ; where is the surface displacement thermal expansion dynamic factor, which is used to reflect the non-linear growth of the displacement rate caused by the thermal expansion effect. Its value is dynamically optimized based on the correlation analysis of thermal cycle tests and displacement during setting, and can take values within 0.0015 - 0.0035. For sandstone, it takes 0.002, and for jointed rock masses, it is recommended to increase to 0.003;

[0051] For the load stress on the rock mass surface , using the formula Calculate and obtain the load stress on the rock mass surface Adjusted warning threshold ; where is the surface stress temperature correction coefficient, which is used to quantify the influence of temperature change on the load-bearing capacity of the rock mass. Its value is adjusted according to the measured value of the thermal expansion coefficient of the minerals composing the rock mass during setting, and can take values within 0.001 - 0.003. The reference value for granite is 0.002, the lower limit for rocks with high quartz content (such as sandstone), and for rock masses containing clay minerals (such as shale), it can reach 0.004;

[0052] For the stress distribution gradient , using the formula Calculate and obtain the stress distribution gradient Adjusted warning threshold ; where is the temperature gradient correction coefficient, which is used to quantify the change in the stress spatial distribution gradient caused by temperature difference. Its value is corrected according to the thermal conductivity of the rock formation during setting, and can take values within 0.04 - 0.08. The empirical value for limestone is 0.05, and the lower limit for rock masses with high thermal conductivity (such as basalt);

[0053] For the deep stratified displacement , using the formula Calculate and obtain the deep stratified displacement Adjusted warning threshold ; where is the deep displacement rainfall accumulation coefficient, which is used to describe the cumulative increase in the displacement threshold of deep rock masses caused by softening or dilation per unit rainfall. Its value is calibrated by linking with the rock mass permeability during setting, and can take values within 0.002 - 0.008. For every 1×10⁻ increase in permeability6 m / s, Increase by 0.0015; the typical value for clay slopes is 0.004, and for gravelly soil layers, it can be taken as 0.006;

[0054] For acoustic emission energy , use the formula to calculate and obtain the acoustic emission energy The adjusted warning threshold ; where is the water pressure coupling coefficient of acoustic emission energy level, which is used to exponentially characterize the amplification effect of fissure water pressure on the energy of micro - fractures, reflecting the non - linear increase contribution of water - rock interaction (such as hydraulic fracturing) to the energy threshold of rock mass instability precursor. Its value is adjusted according to the fissure water pressure fracturing test during setting and can be taken within the range of 0.003 - 0.008. For every 1 MPa decrease in the critical value of water pressure fracturing, increase by 0.001, take 0.005 for brittle limestone, and it can be as low as 0.002 for mudstone;

[0055] For pore water pressure , use the formula to calculate and obtain the pore water pressure The adjusted warning threshold ; where is the rainfall response coefficient of pore water pressure, which is used to linearly reflect the pressure transfer efficiency of rainfall infiltration in pores. Its value is adjusted according to the connectivity of pore networks during setting and can be taken within the range of 0.015 - 0.03. Initially recommend 0.02 for weathered rock masses;

[0056] For seepage rate , use the formula to calculate and obtain the seepage rate The adjusted warning threshold ; where is the rainfall multiplication coefficient of seepage rate, which is used to quantify the multiple effect of rainfall - induced expansion of soil seepage channels and control the dynamic amplification ratio of the seepage velocity threshold. Its value is corrected by combining seepage tests during setting and can be taken within the range of 0.005 to 0.02. For every 1 - fold increase in the ratio of saturated to unsaturated permeability coefficients, increase by 0.005, recommend 0.01 for sandy soil, and suggest reducing to 0.003 for fractured rock masses.

[0057] Calculate the correlation degree between each data group and slope collapse. The specific operation is as follows:

[0058] Obtain a number of stable slope samples. Each stable slope sample contains a first slope data set, a second slope data set, a third slope data set, and a fourth slope data set obtained from a slope area where slope collapse has not occurred for a long time. At the same time, obtain a number of collapsed slope samples. Each collapsed slope sample contains a first slope data set, a second slope data set, a third slope data set, and a fourth slope data set obtained most recently before the slope collapse in a slope area.

[0059] Take all the obtained stable slope samples and collapsed slope samples as slope reference samples, and based on the monitoring parameters of all slope reference samples and the slope collapse status , i = 1, 2, …, ; n represents the number of slope reference samples;

[0060] Based on the monitoring parameters contained in n slope reference samples , and the slope collapse status , use the formula to calculate and obtain the standardized monitoring parameters of n slope reference samples , where and respectively represent the maximum and minimum values among the m monitoring parameters ;

[0061] Set multiple grid combinations , traverse any one grid combination , for each of the slope reference samples, determine the position of the value combination of this slope reference sample in the current grid combination ; use the formula to calculate and obtain the joint probability distribution ; n_i represents the number of samples falling into the grid i in the n slope reference samples; then use the formulas and respectively to calculate and obtain the marginal probability distributions and , where x_i represents the x value of the current grid combination; finally, use the formula to calculate and obtain the mutual information coefficient between the monitoring parameter and slope collapse when applying the current grid combination ; ;

[0062] Based on the mutual information coefficients corresponding to all the obtained grid combinations, select the maximum mutual information coefficient as the monitoring parameter and the maximum information coefficient of slope collapse ;

[0063] Take to the maximum value among them as the correlation degree between the first slope data group and slope collapse ;

[0064] Take to the maximum value among them as the correlation degree between the second slope data group and slope collapse ;

[0065] Take to the maximum value among them as the correlation degree between the third slope data group and slope collapse ;

[0066] Take to the maximum value among them as the correlation degree between the fourth slope data group and slope collapse .

[0067] Calculate the influence weights of each data group. The specific operations are as follows:

[0068] Use the formula to calculate the influence weights of each data group , where \(i = 1, 2, 3, 4\).

[0069] Apply the slope collapse risk prediction model to obtain the slope collapse risk prediction result. The specific operations are as follows:

[0070] At the beginning of the current environmental monitoring period, sort the first slope data group, the second slope data group, the third slope data group, and the fourth slope data group obtained in the previous environmental monitoring period by time to form the first time series set, the second time series set, the third time series set, and the fourth time series set respectively; use each obtained time series set and the influence weights of each data group as the input of the slope collapse risk prediction model, and output the slope collapse risk prediction result.

[0071] The slope collapse risk prediction model is established based on LSTM and includes an input layer, an LSTM layer, a feature weighted fusion layer, a fully connected layer, and an output layer;

[0072] The input layer is used to receive the first time series set, the second time series set, the third time series set, the fourth time series set, and the influence weights of each data group;

[0073] The LSTM layer is used to extract the temporal feature vectors of the first temporal set, the second temporal set, the third temporal set, and the fourth temporal set respectively;

[0074] The feature weighted fusion layer is used to perform weighted fusion on the temporal feature vectors corresponding to each temporal set by applying the influence weights of each data group to obtain the weighted fusion temporal feature vectors;

[0075] The fully connected layer is used to perform non - linear transformation and high - order abstraction on the weighted fusion temporal feature vectors and output high - dimensional feature representations;

[0076] The output layer is used to output the prediction result of the slope collapse risk.

[0077] For the training of the slope collapse risk prediction model, the specific operations are as follows:

[0078] Obtain a number of model training samples with labeled slope collapse risk prediction results. Each model training sample stores the first slope data group, the second slope data group, the third slope data group, and the fourth slope data group sorted by time within an environmental monitoring period for a certain slope area; divide all the obtained model training samples into a training set and a validation set, use the training set to train the slope collapse risk prediction model with initialized parameters, and then verify the slope collapse risk prediction model through the validation set to obtain the verification result; set the training conditions, and judge whether the obtained verification result meets the training conditions. If so, output the trained slope collapse risk prediction model.

[0079] Embodiment 2, a slope stability analysis system based on dynamic monitoring data, as Figure 1 shown, includes:

[0080] The dynamic monitoring data acquisition module is used to collect the first slope data group, the second slope data group, the third slope data group, and the fourth slope data group of the target slope area at any monitoring time point;

[0081] The slope state discrimination module includes an early warning threshold adjustment unit and a discrimination unit; the early warning threshold adjustment unit is used to calculate the maximum temperature difference of the target slope area within the most recent 12 hours at the beginning of any environmental monitoring period, and obtain the rainfall amount within the previous environmental monitoring period, and then adjust the early warning thresholds of some monitoring parameters in each data group; the discrimination unit is used to judge whether all the obtained monitoring parameters meet the corresponding adjusted early warning thresholds within any environmental monitoring period. If so, no operation is performed; if not, warning response measures are executed;

[0082] The slope collapse risk prediction module is used to calculate the correlation degree between each data group and slope collapse, and further calculate and obtain the influence weight of each data group; at the beginning of any environmental monitoring period, apply the slope collapse risk prediction model to obtain the slope collapse risk prediction result; apply the slope collapse risk prediction result to execute corresponding slope collapse prevention measures.

[0083] It should be understood that those of ordinary skill in the art can make improvements or transformations according to the above description, and all such improvements and transformations shall fall within the protection scope of the appended claims of the present invention. The parts not described in detail in this specification belong to the well-known prior art of those skilled in the art.

Claims

1. A slope stability analysis method based on dynamic monitoring data, characterized in that, Including: Collecting a first slope data set, a second slope data set, a third slope data set, and a fourth slope data set of the target slope area at the current monitoring time point; the first slope data set includes surface horizontal displacement, surface vertical displacement, and deep stratified displacement; the second slope data set includes acoustic emission energy and microseismic wave propagation time difference; the third slope data set includes rock mass surface load stress and stress distribution gradient; the fourth slope data set includes pore water pressure and seepage rate; Setting an environmental monitoring period, at the beginning of the current environmental monitoring period, calculating the maximum temperature difference of the target slope area within the last 12 hours, and obtaining the rainfall amount within the previous environmental monitoring period, and then adjusting the warning thresholds of some monitoring parameters in each data set; Subsequently, within the current environmental monitoring period, determining whether each obtained monitoring parameter meets the corresponding adjusted warning threshold. If so, no operation is performed; if not, a warning response measure is executed; Calculating the correlation degree between each data set and slope collapse, and further calculating and obtaining the influence weight of each data set; at the beginning of the current environmental monitoring period, applying a slope collapse risk prediction model to obtain a slope collapse risk prediction result; applying the slope collapse risk prediction result to execute corresponding slope collapse prevention measures; Adjusting the warning thresholds of some monitoring parameters in each data set, and the specific operation is as follows: All monitoring parameters for each data group , = 1, 2, …, 9; to correspond to surface horizontal displacement, surface vertical displacement, deep stratified displacement, acoustic emission energy, microseismic wave propagation time difference, rock mass surface load stress, stress distribution gradient, pore water pressure, and seepage rate in sequence; set the conventional warning thresholds for each monitoring parameter ; The maximum temperature difference obtained based on the current environmental monitoring period and rainfall , for the surface horizontal displacement , using the formula to calculate and obtain the surface horizontal displacement The adjusted warning threshold ; where is the surface displacement thermal expansion dynamic factor; For the load stress on the rock mass surface , use the formula to calculate and obtain the load stress on the rock mass surface The adjusted warning threshold ; among them, is the surface stress temperature correction coefficient; For the stress distribution gradient , use the formula to calculate and obtain the stress distribution gradient The adjusted warning threshold ; where is the temperature gradient correction coefficient; For deep stratified displacement , use the formula to calculate and obtain the deep stratified displacement The adjusted warning threshold ; where is the cumulative coefficient of rainfall for deep displacement For acoustic emission energy , use the formula to calculate and obtain the acoustic emission energy The adjusted warning threshold ; where is the acoustic emission energy level water pressure coupling coefficient; For pore water pressure , use the formula to calculate and obtain the pore water pressure The adjusted warning threshold ; where is the pore water pressure rainfall response coefficient; For the seepage rate , use the formula to calculate and obtain the seepage rate The adjusted warning threshold ; where is the rainfall multiplication factor of the seepage rate.

2. The slope stability analysis method based on dynamic monitoring data according to claim 1, wherein Calculating the correlation degree between each data set and slope collapse, and the specific operation is as follows: Obtain a number of slope reference samples, based on the monitoring parameters of all slope reference samples and the slope collapse state , = 1, 2, …, ; represents the number of slope reference samples; Based on the monitoring parameters included in a slope reference sample , calculate and obtain the standardized monitoring parameters ; Set multiple grid combinations and traverse any one of the grid combinations to calculate and obtain the monitoring parameters when applying the current grid combination and the mutual information coefficient of slope collapse ; Based on the mutual information coefficients corresponding to all the obtained grid combinations, select the maximum mutual information coefficient as the monitoring parameter and the maximum information coefficient of slope collapse ; Take to the maximum value as the correlation degree between the first slope data group and slope collapse ; Take the maximum value among as the correlation degree between the second slope data group and slope collapse ; Take to the maximum value among them as the correlation degree between the third slope data group and slope collapse ; Take to the maximum value as the correlation degree between the fourth slope data group and slope collapse .

3. The slope stability analysis method based on dynamic monitoring data according to claim 2, wherein Calculating and obtaining the influence weight of each data set, and the specific operation is as follows: Using the formula calculate and obtain the influence weights of each data group , = 1, 2, 3, 4.

4. A method for analyzing slope stability based on dynamic monitoring data according to claim 3, characterized in that, Applying a slope collapse risk prediction model to obtain a slope collapse risk prediction result, and the specific operation is as follows: At the beginning of the current environmental monitoring period, respectively sorting the first slope data set, the second slope data set, the third slope data set, and the fourth slope data set obtained within the previous environmental monitoring period according to time to form a first time series set, a second time series set, a third time series set, and a fourth time series set; taking each obtained time series set and the influence weight of each data set as the input of the slope collapse risk prediction model, and outputting a slope collapse risk prediction result.

5. A method for analyzing slope stability based on dynamic monitoring data according to claim 4, characterized in that, The slope collapse risk prediction model is established based on LSTM and includes an input layer, an LSTM layer, a feature weighted fusion layer, a fully connected layer, and an output layer.

6. The slope stability analysis method based on dynamic monitoring data according to claim 5, characterized in that, Regarding the training of the slope collapse risk prediction model, the specific operation is as follows: Obtaining several model training samples with labeled slope collapse risk prediction results; dividing all obtained model training samples into a training set and a validation set, using the training set to train the slope collapse risk prediction model with initialized parameters, and then verifying the slope collapse risk prediction model through the validation set to obtain a verification result; Setting training conditions, and determining whether the obtained verification result meets the training conditions. If so, outputting the trained slope collapse risk prediction model.

7. A slope stability analysis system based on dynamic monitoring data, characterized in that The system is applied to a slope stability analysis method based on dynamic monitoring data according to any one of claims 1-6 above, including: A dynamic monitoring data acquisition module for collecting a first slope data set, a second slope data set, a third slope data set, and a fourth slope data set of the target slope area at any monitoring time point; The slope state discrimination module includes an early warning threshold adjustment unit and a discrimination unit; the early warning threshold adjustment unit is used to adjust the early warning thresholds of some monitoring parameters in each data group at the beginning of any environmental monitoring period; the discrimination unit is used to determine whether each obtained monitoring parameter meets the corresponding adjusted early warning threshold within any environmental monitoring period. If so, no operation is performed; if not, an early warning response measure is executed. The slope collapse risk prediction module applies the slope collapse risk prediction model to obtain the slope collapse risk prediction result; and executes the corresponding slope collapse prevention measures based on the slope collapse risk prediction result.

Images