Slope structure stability safety monitoring method

Abstract

The invention relates to a slope structure stability safety monitoring method, which comprises the following steps of: acquiring characteristic data of geological conditions, stress states and potential instability modes of a slope, partitioning the slope according to the characteristic data, and acquiring a plurality of monitoring partitions; according to the weathering degree and structural surface development characteristics of the slope material in the monitoring subarea, obtaining a dominant deformation mode and key influence factors of the monitoring subarea; based on the dominant deformation mode and the key influence factors, obtaining a sensor optimization layout scheme in the monitoring subarea, and obtaining monitoring data of the monitoring subarea; and performing spatialization processing on the monitoring data, obtaining a spatial distribution diagram of the monitoring data, analyzing the spatial distribution diagram, identifying an abnormal region, evaluating the stability of the abnormal region, and obtaining the stable state of the slope. According to the invention, through zone monitoring and machine learning, the accuracy and pertinence of slope stability evaluation are improved, and a scientific basis is provided for making prevention and control measures and emergency plans.

Description

Technical Field

[0001] The invention relates to the technical field of slope monitoring, and in particular to a method for safely monitoring the stability of a slope structure. Background Art

[0002] In the safety monitoring of slope structure stability, due to the differences in geological conditions, stress states and potential instability modes in different areas of the slope, it is difficult to fully reflect the true stability of the slope by using a unified monitoring scheme. For example, for the strongly weathered layer on the upper part of the slope, its deformation is mainly manifested as creep and local collapse of the soil, and it is necessary to focus on monitoring the surface displacement and changes in soil moisture content; while for the bedrock layer at the lower part of the slope, its deformation is mainly controlled by structural surfaces such as faults and joints, and it is necessary to focus on monitoring the dislocation of the structural surface and changes in stress. In addition, due to the spatial heterogeneity of the geological conditions and external environment of the slope, it is difficult to form a complete stability evaluation. Therefore, it is urgent to establish a safety monitoring method for slope structure stability. Summary of the invention

[0003] The purpose of the present invention is to provide a method for safely monitoring the stability of a slope structure. According to the geological conditions and instability characteristics of the slope, it is divided into several monitoring zones, the sensor layout scheme is optimized according to the characteristics of each zone, and an assessment of the slope stability based on the monitoring data is established, thereby realizing a comprehensive dynamic evaluation of the slope stability.

[0004] To achieve the above object, the present invention provides the following solutions:

[0005] A slope structure stability safety monitoring method, comprising:

[0006] Acquire characteristic data of geological conditions, stress state and potential instability mode of the slope, divide the slope into zones according to the characteristic data, and obtain a plurality of monitoring zones;

[0007] According to the weathering degree and structural surface development characteristics of the slope materials in the monitoring zone, the dominant deformation mode and key influencing factors of the monitoring zone are obtained;

[0008] Based on the dominant deformation mode and the key influencing factors, an optimal arrangement scheme of sensors in the monitoring partition is obtained, and monitoring data of the monitoring partition is obtained;

[0009] The monitoring data is spatially processed to obtain a spatial distribution map of the monitoring data, the spatial distribution map is analyzed to identify abnormal areas, and the stability of the abnormal areas is evaluated to obtain the stability state of the slope.

[0010] Optionally, partitioning the slope according to the characteristic data to obtain a plurality of monitoring partitions includes:

[0011] Preprocessing the characteristic data to obtain a standardized characteristic data set, wherein the characteristic data includes lithology, formation occurrence, fault distribution, groundwater level, stress distribution, and historical instability cases;

[0012] The K-means clustering algorithm is used to perform cluster analysis on the standardized feature data set to obtain a number of monitoring partitions.

[0013] Optionally, obtaining the dominant deformation mode and key influencing factors of the monitoring partition includes:

[0014] According to the weathering degree of the slope material and the evaluation index data of the structural surface development characteristics, the dominant deformation mode and the membership degree of the key influencing factors of the monitoring partition are obtained through fuzzy comprehensive operation;

[0015] According to the membership degree of the dominant deformation mode and the key influencing factor, the dominant deformation mode and the key influencing factor of the monitoring area are determined.

[0016] Optionally, based on the dominant deformation mode and the key influencing factors, obtaining the optimal sensor deployment plan within the monitoring area includes: inputting the dominant deformation model and the key influencing factors into a C5 decision tree to obtain the optimal sensor deployment plan, wherein the C5 decision tree is trained based on a training set, and the training set includes historical dominant deformation modes and historical key influencing factors and corresponding sensor deployment patterns.

[0017] Optionally, obtaining the sensor deployment pattern includes: using a fuzzy C-means clustering algorithm to classify existing sensor optimization configuration schemes to obtain the sensor deployment pattern, wherein the sensor deployment pattern includes sensor deployment patterns suitable for soil slopes, rock slopes and composite slopes.

[0018] Optionally, obtaining the monitoring data of the monitoring partition includes:

[0019] Soil moisture content, rainfall, and creep duration are input into the support vector machine to obtain the creep deformation rate of the weathering layer in the future period of time;

[0020] Input rainfall, evaporation, and groundwater level into the BP neural network to obtain the change range of water content in the future period;

[0021] Input the deformation displacement and pore water pressure into the first ARIMA time series model to obtain the stress change in the future period;

[0022] The displacement rate and rainfall are input into the second ARIMA time series model to obtain the displacement of the structural surface in the future period of time.

[0023] Optionally, analyzing the spatial distribution map to identify abnormal areas includes:

[0024] The geographical location information of the monitoring area is extracted by GIS software, and the monitoring data and the creep deformation rate of the weathering layer, the change amplitude of the water content, the stress change, and the displacement of the structural surface in the future period are compared with the preset threshold value to obtain the spatial position of the abnormal area;

[0025] The abnormal range is determined according to the number of pixels in the abnormal area.

[0026] Optionally, evaluating the stability of the abnormal area to obtain the stable state of the slope includes: inputting the monitoring data in the abnormal area and the creep deformation rate of the weathering layer, the moisture content change amplitude, the stress change amount, and the structural surface displacement amount in the future period into a slope stability evaluation model, and outputting the slope stability grade, wherein the slope stability evaluation model is a support vector machine;

[0027] According to the spatial position, abnormal range and stability level of the abnormal area, the hierarchical analysis method is used to construct a judgment matrix and perform consistency test to obtain a comprehensive evaluation result of the overall stability of the slope.

[0028] The beneficial effects of the present invention are as follows: the present invention discloses a slope stability assessment method based on monitoring zoning. The method first uses a clustering algorithm to partition the slope, and then uses fuzzy comprehensive evaluation to determine the dominant deformation mode and key influencing factors for each partition, and optimizes the sensor layout accordingly. Corresponding sensors are deployed in different partitions, such as surface displacement meters and soil moisture sensors in the strongly weathered layer, and stress meters and displacement meters in the bedrock layer. The monitoring data is processed and visualized by spatial statistical methods to reveal the spatial distribution pattern. Finally, a machine learning algorithm is used to train a slope stability evaluation model to achieve dynamic assessment and early warning. The present invention improves the accuracy and pertinence of slope stability assessment through zoning monitoring and machine learning, and provides a scientific basis for the formulation of prevention and control measures and emergency plans. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0030] Figure 1 The present invention is a flowchart of a slope structure stability safety monitoring method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0032] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0033] like Figure 1 As shown, this embodiment provides a slope structure stability safety monitoring method, including:

[0034] Obtain characteristic data of the geological conditions, stress state and potential instability mode of the slope, divide the slope into zones according to the characteristic data, and obtain several monitoring zones;

[0035] According to the weathering degree and structural surface development characteristics of the slope materials in the monitoring area, the dominant deformation mode and key influencing factors of the monitoring area are obtained;

[0036] Based on the dominant deformation mode and key influencing factors, the optimal layout plan of sensors in the monitoring area is obtained, and the monitoring data of the monitoring area is obtained;

[0037] The monitoring data is spatially processed to obtain the spatial distribution map of the monitoring data, the spatial distribution map is analyzed to identify abnormal areas, and the stability of the abnormal areas is evaluated to obtain the stability state of the slope.

[0038] Furthermore, the slope is partitioned according to the characteristic data to obtain several monitoring partitions including:

[0039] Preprocess the characteristic data to obtain a standardized characteristic data set, where the characteristic data includes lithology, formation occurrence, fault distribution, groundwater level, stress distribution, and historical instability cases;

[0040] The K-means clustering algorithm is used to perform cluster analysis on the standardized feature data set to obtain several monitoring partitions.

[0041] Specifically, we first obtain the characteristic data of the geological conditions, stress state and potential instability mode of the slope, including lithology, formation occurrence, fault distribution, groundwater level, stress distribution, historical instability cases, etc., and perform normalization and denoising on these characteristic data to obtain a standardized characteristic data set. Then, we use the K-means clustering algorithm to perform cluster analysis on the standardized characteristic data set, setting the number of clusters k to 5, the number of iterations to 100, and the convergence threshold to 0.01, and finally obtain 5 monitoring partitions.

[0042] Furthermore, the dominant deformation mode and key influencing factors of the monitored partition are obtained, including:

[0043] According to the evaluation index data of the weathering degree of slope materials and the development characteristics of structural surfaces, the dominant deformation mode of the monitoring area and the membership degree of key influencing factors are obtained through fuzzy comprehensive operation;

[0044] According to the membership degree of the dominant deformation mode and key influencing factors, the dominant deformation mode and key influencing factors of the monitoring area are determined.

[0045] Specifically, in a certain slope project, through field investigation and indoor tests, the evaluation index data such as the weathering degree of slope materials and the development characteristics of structural surfaces were obtained. According to expert experience, a fuzzy comprehensive evaluation model including four indicators such as weathering degree, structural surface inclination, structural surface spacing, and structural surface filling hardness was established. The analytic hierarchy process was used to determine the weights of each indicator as 4, 3, 2, and 1, respectively, and the trapezoidal membership function was selected to describe the indicator membership. Taking a certain monitoring zone as an example, through fuzzy comprehensive operation, the memberships of the sliding deformation and collapse deformation of the zone were 8 and 4, respectively, and the memberships of rainfall infiltration and unloading excavation were 7 and 5, respectively, indicating that the dominant deformation mode of the zone is sliding deformation, and the key influencing factor is rainfall infiltration.

[0046] Furthermore, based on the dominant deformation mode and key influencing factors, obtaining the optimal sensor deployment plan within the monitoring area includes: inputting the dominant deformation model and key influencing factors into the C5 decision tree to obtain the optimal sensor deployment plan, wherein the C5 decision tree is trained based on a training set, and the training set includes historical dominant deformation modes and historical key influencing factors and corresponding sensor deployment patterns.

[0047] Specifically, the C5 decision tree algorithm is used, with the dominant deformation mode and key influencing factors as input attributes, and the optimal sensor configuration scheme under different engineering geological conditions as output attributes. The optimal partition attribute is selected through the information gain ratio to generate a sensor optimization configuration decision tree. If the dominant deformation mode of a partition is sliding deformation, displacement sensors are preferentially deployed; if the dominant deformation mode is collapse deformation, stress and strain sensors are preferentially deployed. The key areas for sensor deployment are determined according to the key influencing factors. If the key influencing factor is rainfall infiltration, piezometers are deployed at the top and foot of the slope; if the key influencing factor is unloading excavation, displacement sensors are deployed near the excavation surface. The sensor optimization configuration scheme is classified by the fuzzy clustering algorithm, and the sensor deployment mode suitable for different engineering geological conditions is obtained to guide engineering practice.

[0048] Furthermore, obtaining the sensor deployment pattern includes: using a fuzzy C-means clustering algorithm to classify existing sensor optimization configuration schemes to obtain the sensor deployment pattern, wherein the sensor deployment pattern includes sensor deployment patterns suitable for soil slopes, rock slopes and composite slopes.

[0049] Specifically, the fuzzy C-means clustering algorithm was used to classify the existing sensor optimization configuration schemes. By setting the number of cluster centers to 3 and the maximum number of iterations to 100, three sensor deployment modes suitable for soil slopes, rock slopes and composite slopes were obtained, providing a reference for engineering practice.

[0050] Furthermore, obtaining the monitoring data of the monitoring partition includes:

[0051] Soil moisture content, rainfall, and creep duration are input into the support vector machine to obtain the creep deformation rate of the weathering layer in the future period of time;

[0052] Input rainfall, evaporation, and groundwater level into the BP neural network to obtain the change range of water content in the future period;

[0053] Input the deformation displacement and pore water pressure into the first ARIMA time series model to obtain the stress change in the future period;

[0054] The displacement rate and rainfall are input into the second ARIMA time series model to obtain the displacement of the structural surface in the future period of time.

[0055] Specifically, the geological characteristic information of the slope is obtained through engineering geological survey, including the thickness of the strongly weathered layer, the degree of weathering, the lithology, the occurrence of the structural surface, etc., as well as the lithology, integrity, and degree of development of the structural surface of the bedrock layer. Areas with a thickness of more than 5m and a degree of weathering of fully weathered or strongly weathered are divided into strongly weathered layer monitoring areas; areas with good integrity of the bedrock layer and developed structural surfaces are divided into bedrock layer monitoring areas. In the strongly weathered layer monitoring area, a surface displacement sensor and a soil moisture sensor are arranged every 50m to obtain surface displacement data and soil moisture data in real time, with a sampling frequency of 1 time / hour. In the bedrock layer monitoring area, a stress sensor and a displacement sensor are arranged every 30m along the dominant structural surface to obtain structural surface dislocation data and stress change data, with a sampling frequency of 1 time / hour. The support vector machine algorithm was used to establish the weathering layer creep deformation prediction model, and the historical soil moisture content, rainfall, creep duration and historical surface displacement data were used for training. The BP neural network algorithm was used to establish the moisture content change prediction model, and the historical rainfall, evaporation, groundwater level and soil moisture content data were used for training. The ARIMA time series model was used to establish the bedrock layer structural surface dislocation prediction model and stress change prediction model, and the historical deformation displacement, pore water pressure and stress change data, historical displacement rate, rainfall and structural surface dislocation data were used to train the first ARIMA time series model and the second ARIMA time series model respectively. The real-time monitoring data of each monitoring area was input into the corresponding prediction model to obtain the weathering layer creep deformation rate, moisture content change amplitude, bedrock layer structural surface dislocation and stress change in the next 7 days.

[0056] Furthermore, the spatial distribution map is analyzed to identify abnormal areas including:

[0057] The geographical location information of the monitoring area is extracted through GIS software, and the monitoring data is compared with the creep deformation rate, moisture content change amplitude, stress change, and structural surface displacement of the weathering layer in the future period of time with the preset threshold to obtain the spatial location of the abnormal area;

[0058] The abnormal range is determined based on the number of pixels in the abnormal area.

[0059] Specifically, the monitoring data of different partitions of the slope are obtained. According to the spatial heterogeneity of the slope, spatial statistical methods such as Kriging interpolation are used to perform spatial processing on the monitoring data. Based on the monitoring data after spatial processing, the spatial distribution map of the slope monitoring data is generated through visualization technology to intuitively display the spatial distribution characteristics of the monitoring data of different partitions. For each sub-area, its geographical location information is extracted through GIS software, and the spatial distribution map is analyzed to identify the abnormal areas of the monitoring data, and the location and range of the abnormal areas are determined. Taking a certain slope as an example, a total of 50 monitoring points were set up, and monitoring data such as slope displacement and soil moisture content were obtained. Using the Kriging interpolation tool in ArcGIS software, the interpolation parameters were set, and the monitoring data were spatially interpolated and analyzed to generate a spatial distribution map of the slope monitoring data. Through the analysis of the spatial distribution map, three abnormal monitoring data areas were identified. The monitoring data features of each sub-region are compared with the preset abnormal threshold; if any monitoring data feature of the sub-region exceeds the abnormal threshold, the sub-region is judged to be an abnormal region; for the sub-region judged to be abnormal, the spatial position of the abnormal region is determined according to its position coordinates; and the abnormal range is determined according to the number of pixels in the abnormal sub-region.

[0060] Furthermore, the stability of the abnormal area is evaluated to obtain the stability state of the slope, including: inputting the monitoring data in the abnormal area and the creep deformation rate of the weathering layer, the change amplitude of the water content, the stress change, and the displacement of the structural surface in the future into the slope stability evaluation model, and outputting the slope stability grade, wherein the slope stability evaluation model is a support vector machine;

[0061] According to the spatial location, abnormal range and stability level of the abnormal area, the hierarchical analysis method is used to construct a judgment matrix and conduct consistency test to obtain the comprehensive evaluation result of the overall stability of the slope.

[0062] Specifically, a monitoring data set in a unified format is obtained through data cleaning and feature extraction, and a slope stability evaluation model is established using the support vector machine algorithm. In this embodiment, through the training and verification of the model, the stability level of the abnormal area is obtained as level four, and there is a certain risk of instability. Taking into account the location, scope and stability evaluation results of the abnormal area, the hierarchical analysis method is adopted. By constructing a judgment matrix and performing a consistency test, the comprehensive evaluation result of the overall stability of the slope is finally obtained as level three. The slope is in a basically stable state as a whole, but there is a risk of instability in local areas. According to the comprehensive research and judgment results, it is determined that the dangerous area of ​​the slope is located near the abnormal area, and the potential instability mode may be local collapse or slippage, which provides an important basis for the formulation of slope management measures. And according to the warning level, the corresponding emergency response mechanism is automatically triggered.

[0063] According to the evaluation and early warning results, the scenario analysis method is used to simulate the potential instability risk of the slope under different scenarios, and the impact range of slope instability is calculated through numerical simulation technology. According to the results of scenario analysis and numerical simulation, corresponding slope prevention and control measures are formulated, including engineering control measures and non-engineering control measures, and a slope disaster emergency plan is established. Obtain slope monitoring data, predict the trend of slope monitoring data, and judge the stability state of the slope through machine learning algorithms such as support vector machines based on the prediction results and monitoring data. If the risk of slope instability increases, dynamically adjust the monitoring plan, increase the monitoring frequency and monitoring parameters, and realize dynamic control of slope disaster risks. By establishing a full-process management platform for slope disasters, integrating evaluation and early warning, scenario analysis, numerical simulation, monitoring and early warning, prevention and control measures, emergency plans and other functions, the full life cycle management of slope disaster risks can be realized. If the slope is unstable, the emergency plan is activated, and rescue work is carried out according to the procedures and measures specified in the plan to minimize casualties and economic losses caused by slope instability. In the whole process management of slope disasters, we will continue to optimize slope assessment and early warning models, scenario analysis methods, numerical simulation technologies, monitoring and early warning algorithms, etc., continuously improve the intelligence level of slope disaster risk management, and provide scientific and technological support for slope disaster prevention and control.

[0064] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A slope structure stability safety monitoring method, characterized in that: include: Acquire characteristic data of geological conditions, stress state and potential instability mode of the slope, divide the slope into zones according to the characteristic data, and obtain a plurality of monitoring zones; According to the weathering degree and structural surface development characteristics of the slope materials in the monitoring zone, the dominant deformation mode and key influencing factors of the monitoring zone are obtained; Based on the dominant deformation mode and the key influencing factors, an optimal arrangement scheme of sensors in the monitoring partition is obtained, and monitoring data of the monitoring partition is obtained; The monitoring data is spatially processed to obtain a spatial distribution map of the monitoring data, the spatial distribution map is analyzed to identify abnormal areas, and the stability of the abnormal areas is evaluated to obtain the stability state of the slope.

2. The slope structure stability safety monitoring method according to claim 1 is characterized in that: Partitioning the slope according to the characteristic data to obtain a number of monitoring partitions includes: Preprocessing the characteristic data to obtain a standardized characteristic data set, wherein the characteristic data includes lithology, formation occurrence, fault distribution, groundwater level, stress distribution, and historical instability cases; The K-means clustering algorithm is used to perform cluster analysis on the standardized feature data set to obtain a number of monitoring partitions.

3. The slope structure stability safety monitoring method according to claim 1 is characterized in that: Obtaining the dominant deformation mode and key influencing factors of the monitoring area includes: According to the weathering degree of the slope material and the evaluation index data of the structural surface development characteristics, the dominant deformation mode and the membership degree of the key influencing factors of the monitoring partition are obtained through fuzzy comprehensive operation; According to the membership degree of the dominant deformation mode and the key influencing factor, the dominant deformation mode and the key influencing factor of the monitoring area are determined.

4. The slope structure stability safety monitoring method according to claim 1 is characterized in that: Based on the dominant deformation mode and the key influencing factors, obtaining the optimal sensor deployment plan in the monitoring area includes: inputting the dominant deformation model and the key influencing factors into a C5 decision tree to obtain the optimal sensor deployment plan, wherein the C5 decision tree is trained based on a training set, and the training set includes historical dominant deformation modes and historical key influencing factors and corresponding sensor deployment patterns.

5. The slope structure stability safety monitoring method according to claim 4 is characterized in that: Acquiring the sensor deployment pattern includes: using a fuzzy C-means clustering algorithm to classify existing sensor optimization configuration schemes to acquire the sensor deployment pattern, wherein the sensor deployment pattern includes sensor deployment patterns suitable for soil slopes, rock slopes and composite slopes.

6. The slope structure stability safety monitoring method according to claim 1 is characterized in that: After acquiring the monitoring data of the monitoring partition, the following steps are included: Soil moisture content, rainfall, and creep duration are input into the support vector machine to obtain the creep deformation rate of the weathering layer in the future period of time; Input rainfall, evaporation, and groundwater level into the BP neural network to obtain the change range of water content in the future period; Input the deformation displacement and pore water pressure into the first ARIMA time series model to obtain the stress change in the future period; The displacement rate and rainfall are input into the second ARIMA time series model to obtain the displacement of the structural surface in the future period of time.

7. The slope structure stability safety monitoring method according to claim 6 is characterized in that: The spatial distribution map is analyzed to identify abnormal areas including: The geographical location information of the monitoring area is extracted by GIS software, and the monitoring data and the creep deformation rate of the weathering layer, the change amplitude of the water content, the stress change, and the displacement of the structural surface in the future period are compared with the preset threshold value to obtain the spatial position of the abnormal area; The abnormal range is determined according to the number of pixels in the abnormal area.

8. The slope structure stability safety monitoring method according to claim 7 is characterized in that: Evaluating the stability of the abnormal area to obtain the stability state of the slope includes: inputting the monitoring data in the abnormal area and the creep deformation rate of the weathering layer, the moisture content change amplitude, the stress change amount, and the structural surface displacement amount in the future period into a slope stability evaluation model, and outputting the slope stability grade, wherein the slope stability evaluation model is a support vector machine; According to the spatial position, abnormal range and stability level of the abnormal area, the hierarchical analysis method is used to construct a judgment matrix and perform consistency test to obtain a comprehensive evaluation result of the overall stability of the slope.

Images