Real-time Landslide Monitoring Method and System Based on Multi-Sensor Fusion and Intelligent Control

Abstract

This invention relates to the field of landslide control technology, and discloses a real-time landslide monitoring method and system based on multi-sensor fusion and intelligent control. The method includes placing at least one sensor inside the metal reinforcement structure and at key locations on the landslide body to monitor landslide data in real time; acquiring real-time water composition imaging data of the slope under rainfall conditions; deriving rainfall volume from the landslide water composition imaging data; automatically adjusting the valve openings of drainage and diversion channels based on the rainfall volume; calculating the stability parameter distribution from the rainfall volume; obtaining a safety factor based on the stability parameter distribution; obtaining the instability probability of the slope through multiple grid cell safety factors; and optimizing the structural design of the metal reinforcement piles and reinforcement frame through finite element analysis. This invention improves the accuracy of landslide detection and early warning capabilities through real-time monitoring, data analysis, and automatic control of multi-source data.

Description

Technical Field

[0001] This invention relates to the field of landslide control technology, and in particular to a real-time landslide monitoring method and system based on multi-sensor fusion and intelligent control. Background Technology

[0002] A landslide is a natural phenomenon in which soil or rock masses on a slope slide downhill, either as a whole or in parts, under the influence of gravity, due to factors such as river erosion, groundwater activity, rainwater soaking, earthquakes, and artificial slope cutting. The moving rock (soil) mass is called the displaced body or sliding body, and the underlying rock (soil) mass that has not moved is called the sliding bed. Currently, in emergency landslide prevention and control, multiple pile positions are usually arranged on the landslide surface to stabilize the landslide. However, these multiple pile positions are usually independent, lacking measures to connect and reinforce each other, which affects the effectiveness of landslide prevention and control, and thus the effectiveness of landslide prevention and control needs to be improved.

[0003] Existing technology 1, Chinese patent, patent number: 202411585519.0, relates to the fields of geological engineering and civil engineering, and particularly to a testing method for the impact of loess moisture content changes on slope stability. The method includes: collecting and preparing loess samples, measuring the physical parameters of the loess samples, and constructing a multidimensional physical property matrix of the loess samples; real-time monitoring of pore water pressure, calculating the shear stress of the loess samples based on the dynamic changes in pore water pressure; and dynamically updating the dynamic stability parameters of the slope based on the shear stress and the multidimensional physical property matrix of the loess samples, thus comprehensively assessing the slope instability risk. While this addresses the problems of traditional slope monitoring techniques lacking systematic research on the mechanical properties of loess, resulting in poor accuracy and reliability of prediction results; lacking real-time updates to slope stability, leading to the inability to form an accurate dynamic monitoring and recursive prediction mechanism; and failing to issue timely early warning signals, resulting in insufficient timeliness of landslide emergency response; however, it does not consider the impact on landslide prevention, leading to the slippage of metal reinforcement equipment.

[0004] Prior art two, Chinese patent, patent number: 202411108782.0, discloses an emergency road maintenance structure and method for interface-type landslide highways. The method includes the following steps: Step 1: Calculate the shear reinforcement of the pile foundation based on the thrust; Step 2: Close the highway traffic at the landslide section, mill and remove the road surface layer and base layer of the highway at the landslide section, and locate the positions of the consolidated pile foundation and the unconsolidated pile foundation; Step 3: Drill and construct the pile foundation, and construct the tie beam between the unconsolidated pile foundations; Step 4: Tie the steel reinforcement skeleton of the raft foundation and pour the raft foundation concrete to achieve the consolidation of the consolidated pile foundation and the raft foundation. The unconsolidated pile foundation and the tie beam are supported at the bottom of the raft foundation, completing the construction of the first treatment unit; Repeat the above steps to construct other treatment units; Step 5: Connect the highway in the non-landslide section to the raft foundation located at the end of the landslide section through a connecting plate, and carry out landslide treatment of the landslide section of the slope. Although the road at the landslide section could be separated from the slope, allowing the road to be reopened to traffic as soon as possible and greatly shortening the construction period, the lack of consideration for drainage channels resulted in the failure to properly manage rainwater diversion.

[0005] Prior art three, Chinese patent number 202410304622.7, relates to the field of landslide control technology, and particularly to a landslide control slope drainage and reinforcement system. It includes a slope body with an internal drainage structure and an external drainage structure. The internal drainage structure comprises seamless steel pipes and permeable foamed concrete columns. The exposed end of the internal drainage structure is connected to a slope intercepting ditch. The surface of the seamless steel pipe is provided with permeable holes and a filter screen. The slope intercepting ditch is connected to the slope drainage ditch, and the slope drainage ditch is connected to the slope toe drainage ditch. While it has the following effects: the internal drainage structure allows water inside the slope to enter the permeable foamed concrete through the permeable holes, automatically draining water using the pressure difference between the inside and outside of the pipe. The exposed end is connected to the slope surface intercepting ditch, and finally, the internal water is discharged from the slope through the external drainage structure. Simultaneously, the internal drainage structure reinforces the weak soil layer, effectively reducing the probability of landslides during the rainy season; however, it does not consider detection and early warning, and cannot respond promptly when a landslide occurs.

[0006] Currently, existing technologies 1, 2, and 3 lack consideration for the impact on landslide prevention, drainage channels, and detection and early warning, resulting in the need to further improve the effectiveness of landslide prevention. To solve the above problems, this invention provides a landslide emergency prevention technology. Summary of the Invention

[0007] The main objective of this invention is to provide a landslide emergency prevention and control technology to address the problem that existing technologies lack consideration of the impact on landslide prevention and control, drainage channels, and detection and early warning, which leads to the need to further improve the effectiveness of landslide prevention and control.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A landslide emergency prevention and control technology specifically includes the following steps:

[0010] At least one sensor is placed inside the metal reinforcement and at key locations on the landslide body to monitor landslide data in real time. The monitored landslide data is divided into multiple grid cells. The water composition imaging data of the slope is also acquired in real time under rainfall conditions.

[0011] The water composition data includes the water content of each grid cell; the sensors include displacement sensors, stress sensors, humidity sensors, and rainwater monitoring sensors, etc.

[0012] The water composition imaging data of the landslide body is compared with the preset water composition threshold to obtain the rainfall amount; the valve opening of the drainage channel and diversion channel is automatically adjusted according to the rainfall amount; the stability parameter distribution is calculated based on the rainfall amount.

[0013] The safety factor is obtained based on the distribution of stability parameters, and the instability probability of the slope is obtained through the safety factors of multiple grid units. If the instability probability reaches the start threshold, the grouting equipment is automatically started. When the grouting reaches the grouting threshold, the grouting equipment is shut down. The structural design of the metal reinforcement piles and reinforcement frame is optimized and adjusted through finite element analysis.

[0014] As a further improvement to the present invention, the process of acquiring real-time water composition imaging data of a slope under rainfall conditions specifically includes the following steps:

[0015] At least one sensor is placed inside the metal reinforcement and at key locations on the landslide body to monitor landslide data in real time. The landslide data is then divided into multiple grid units. Multiple electrodes are then arranged according to the multiple grid units.

[0016] The resistivity of multiple grids is measured; the resistivity at any location in the landslide body is acquired in real time; the resistivity is balanced by suction to obtain the corresponding resistance value, and the corresponding water content is obtained based on the resistance value;

[0017] Obtain the corresponding resistivity based on the resistance value, fit the moisture content to the resistivity to obtain the relationship between resistivity and moisture content; discretize the resistivity at different locations in the corresponding grid cells to obtain the resistivity of each grid cell; input the resistivity of each grid cell into the relationship between resistivity and moisture content to obtain the moisture content of each grid cell.

[0018] As a further improvement of the present invention, the process of obtaining the stability parameter distribution specifically includes the following steps:

[0019] The water composition imaging data of the landslide body is compared with the preset water composition threshold. Combined with the evolution law of the landslide body itself and the landslide deformation mechanism, the landslide body data is feature extracted to obtain the feature set of rainfall-induced landslide impact.

[0020] The rainfall-induced landslide impact feature set includes historical displacement data features, rainfall factor features, stability coefficient features, rainfall infiltration depth features, and time factor features.

[0021] Based on the characteristics of historical displacement data, a predicted displacement sequence is generated from the feature set of rainfall-induced landslide impacts. The predicted displacement sequence is used to determine the predicted landslide time. An early warning is issued based on the predicted landslide time and uploaded to the central control system.

[0022] The central control system analyzes the landslide time and automatically adjusts the valve opening of the drainage and diversion channels based on the analysis results; it also analyzes the predicted rainfall to obtain the distribution of stability parameters.

[0023] As a further improvement of the present invention, the process of obtaining the feature set of rainfall-induced landslide impacts specifically includes the following steps:

[0024] Collect historical displacement data characteristics; combine the landslide body's own evolution law and the mechanism of rainfall's effect on landslide deformation, and analyze rainfall factors based on historical average rainfall, previous effective rainfall, and number of rainfall days;

[0025] The importance of rainfall factors is determined to identify their characteristics; the stability coefficient and rainfall infiltration depth characteristics are determined based on the permeability coefficient, shear strength, and landslide slope; and the characteristics of time factors are determined based on the different monthly rainfall amounts.

[0026] The time-related features are classified based on the similarity between them; the feature set of rainfall-induced landslide impacts is determined based on historical displacement data features, rainfall factor features, stability coefficient features, rainfall infiltration depth features, and time-related features.

[0027] As a further improvement of the present invention, the process of predicting landslide time for early warning specifically includes the following steps:

[0028] The deformation acceleration point is calculated based on the curve between the reciprocal of the landslide velocity and time; the predicted landslide time is determined using the predicted displacement sequence based on the deformation acceleration point; and the difference between the predicted landslide time and the current time is used as an early warning indicator.

[0029] Construct a landslide monitoring and early warning mechanism based on early warning indicators and landslide movement rate; detect and issue landslide warnings based on the landslide monitoring and early warning mechanism; and upload the real-time monitored landslide early warning data to the central control system.

[0030] The central control system analyzes real-time landslide early warning data in conjunction with rainfall-induced landslide impact characteristic sets to predict the landslide phenomenon that will occur at the corresponding time.

[0031] As a further improvement of the present invention, the process of obtaining the stability parameter distribution specifically includes the following steps:

[0032] The landslide phenomenon occurring at the corresponding time is analyzed and calculated in conjunction with historical landslide phenomena to obtain valve opening and closing values; the valve opening of drainage channels and diversion channels is automatically adjusted based on the valve opening and closing values.

[0033] Based on the feature set of rainfall-induced landslides, the soil-water relationship is derived; based on the relationship between soil resistivity and volume water, the relationship between matrix suction and soil resistance is obtained; by introducing suction stress, the relationship between suction stress and water content is obtained.

[0034] The stress at any location on the landslide body is obtained, and the absorption stress is used to replace the pore pressure to obtain the relationship between the absorption stress and the effective stress. Based on the relationship between the absorption stress and the effective stress, combined with the rainfall, the distribution of stability parameters is obtained.

[0035] As a further improvement of the present invention, the process of optimizing the structural design of metal reinforced piles and reinforcement frames through finite element analysis specifically includes the following steps:

[0036] The safety factor of each grid cell is obtained based on the stability parameter distribution; the total number of grid cells is obtained, and the number of grid cells with a local safety factor less than 1 is counted; the percentage of grid cells with a local safety factor less than 1 to the total number of grid cells is obtained and their distribution is analyzed to obtain the instability probability of the slope.

[0037] The instability probability is compared with a preset instability probability threshold. When the instability probability reaches the activation threshold, the grouting equipment is automatically started. When the grouting reaches the grouting threshold, the grouting equipment is shut down, and the instability probability is recalculated. If it is greater than the activation threshold, the grouting equipment is automatically started. This process is repeated until the instability probability is lower than the activation threshold.

[0038] Features are extracted from the distribution of stability parameters. Based on the extracted features, the rainfall-landslide impact feature dataset is adjusted and fused to create a dynamic map in a high-dimensional data space. This yields quantitative results of the distribution of influencing factors at various locations of metal reinforcement piles and reinforcement frames.

[0039] As a further improvement of the present invention, the process of obtaining the quantitative results of the distribution of influencing factors at various locations of the metal reinforcement pile and the reinforcement frame specifically includes the following steps:

[0040] The adjusted rainfall impact feature datasets at each time point are combined into a high-dimensional vector, and standardized data are obtained through data preprocessing and normalization. The covariance matrix is ​​calculated based on the standardized data, and eigenvalues ​​and corresponding eigenvectors are obtained by eigenvalue decomposition of the covariance matrix.

[0041] Based on the magnitude of the eigenvalues, the largest eigenvalue and the corresponding eigenvector are selected to form the principal components. The standardized data are then projected onto the principal components to map the high-dimensional data to a low-dimensional space. Data points at each time point are plotted in the low-dimensional space to construct a dynamic spectrum.

[0042] Based on the distribution and trend of dynamic spectrum data points, quantitative analysis is performed to obtain a statistical summary of the impact at each time point or in each region; this is recorded as the quantitative results of the distribution of influencing factors for metal reinforced piles and reinforcement frames; the structure of metal reinforced piles and reinforcement frames is adjusted based on the quantitative results.

[0043] Quantitative analysis includes examining the distribution of data points in a low-dimensional space over time and identifying outliers.

[0044] As a further improvement to the present invention, the process of adjusting the metal reinforcement pile and reinforcement frame structure based on the quantification results specifically includes the following steps:

[0045] Based on dynamic spectrograms, analyze the distribution of data points at each time point or in each region; identify the changing patterns and potential outliers in the data over time; and quantify the data distribution in the dynamic spectrograms.

[0046] Calculate the cumulative contribution points of eigenvalues ​​and eigenvectors at each time point or region, evaluate the explanatory power of different principal components on data changes based on the cumulative contribution points, and determine the variance contribution value at each time point or region to determine the main direction of data change.

[0047] Based on the identification of key data and outliers in the dynamic spectrum, and combined with the contribution of principal components to the main direction of data change, the structural design of metal reinforced piles and reinforcement frames is adjusted; the adjusted data is then recalculated, and the calculation results are compared with the preset structural threshold. If the result is greater than the preset structural threshold, the adjustment is repeated; this process is iterated until the calculation result is less than the preset structural threshold.

[0048] To achieve the above objectives, the present invention also provides the following technical solution:

[0049] A landslide real-time monitoring system based on multi-sensor fusion and intelligent control includes:

[0050] Multiple metal reinforcement piles are driven into the slope of the landslide body in a matrix. The metal reinforcement piles are hollow inside, and multiple grouting ports are opened on the outer wall of the metal reinforcement piles. The multiple grouting ports are evenly distributed on the outer wall of the metal reinforcement piles. Then, concrete grout is injected into the multiple metal reinforcement piles. The concrete grout penetrates into the landslide body through the grouting ports. After the concrete grout solidifies, the metal reinforcement piles are fixed inside the slope of the landslide body, so as to grout and reinforce the landslide body.

[0051] Multiple metal reinforcement piles are arranged in four vertical rows. Each of the four rows of metal reinforcement piles is equipped with a reinforcement frame at the top. The bottom of the four reinforcement frames is fixed to the top of the four vertical rows of metal reinforcement piles by bolts.

[0052] A drainage channel is dug on one side of the landslide surface of the landslide body, and multiple diversion channels are dug on the landslide surface of the landslide body. The multiple diversion channels are inclined, and the downward inclined end of the drainage channel is connected to the interior of the drainage channel.

[0053] This invention involves driving multiple metal reinforcement piles in a matrix into the slope surface of a landslide, then injecting concrete grout into the piles. The grout penetrates into the landslide body through injection ports, fixing the metal reinforcement piles within the slope and reinforcing the landslide to resist sliding forces and stabilize it. Multiple reinforcement frames and connecting rods facilitate connecting the piles, increasing their stability and forming a unified structure, thus enhancing landslide prevention. The invention further strengthens the connection between the metal reinforcement piles and the surrounding structure through the use of connecting locks, reinforcement seats, and anchor bolts. The reinforcement frame is further tightened to prevent the metal reinforcement piles from sliding down, further improving the landslide prevention effect. By using anchor bolts, drainage channels, and filters, rainwater from the landslide surface is intercepted and diverted into multiple diversion channels during rainy weather. These channels then flow into drainage channels and are discharged out of the landslide body. This allows for the diversion and drainage of rainwater from the landslide surface, thus managing the landslide. At the same time, the filters remove leaves, dead branches, or other debris from the water, preventing them from entering the drainage and diversion channels and causing blockages. Attached Figure Description

[0054] Figure 1 This is a flowchart illustrating the steps of an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to the present invention.

[0055] Figure 2 This is a schematic diagram of the steps in an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control of the present invention, which acquires water composition imaging data of the slope in real time under rainfall conditions.

[0056] Figure 3 This is a schematic diagram of the steps for obtaining stability parameter scores in an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to the present invention.

[0057] Figure 4 This is a schematic diagram illustrating the steps of obtaining a rainfall-induced landslide impact feature set in an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to the present invention.

[0058] Figure 5 This is a schematic diagram of the steps for predicting landslide time and issuing early warnings in an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control of the present invention.

[0059] Figure 6 This is a schematic diagram illustrating the steps of obtaining the stability parameter distribution in an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to the present invention.

[0060] Figure 7 This is a schematic diagram illustrating the steps of optimizing the structural design of metal reinforcement piles and reinforcement frames through finite element analysis in an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control of the present invention.

[0061] Figure 8 This is a schematic flowchart of an embodiment of the present invention to obtain the quantitative results of the distribution of influencing factors at various locations of metal reinforced piles and reinforcement frames;

[0062] Figure 9 This is a schematic diagram of the steps for adjusting the metal reinforcement piles and reinforcement frame structure according to the quantification results in an embodiment of the landslide real-time monitoring method based on multi-sensor fusion and intelligent control of the present invention.

[0063] Figure 10 This is a schematic diagram of the functional modules of an embodiment of the landslide real-time monitoring system based on multi-sensor fusion and intelligent control of the present invention.

[0064] Figure 11 This is a functional module diagram of one embodiment of the pile connection reinforcement system of the present invention;

[0065] Figure 12 This is a schematic diagram of the functional modules of an embodiment of the drainage and protection system of the present invention;

[0066] Figure 13 This is a schematic diagram of the structure of an embodiment of the electronic device of the present invention;

[0067] Figure 14 This is a schematic diagram of the structure of one embodiment of the storage medium of the present invention. Detailed Implementation

[0068] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0069] The terms "first," "second," and "third" used in this invention are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of that feature. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified. All directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of this invention are only used to explain the relative positional relationships and movements between components in a specific orientation (as shown in the figures). If the specific orientation changes, the directional indications also change accordingly. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0070] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0071] like Figure 1 As shown, this embodiment provides an example of a landslide real-time monitoring method based on multi-sensor fusion and intelligent control. In this embodiment, the landslide real-time monitoring method based on multi-sensor fusion and intelligent control specifically includes the following steps:

[0072] Step S1: Place at least one sensor inside the metal reinforcement and at key locations on the landslide body to monitor landslide data in real time, divide the monitored landslide data into multiple grid cells, and acquire water composition imaging data of the slope body in real time under rainfall conditions.

[0073] The water composition data includes the water content of each grid cell; the sensors include displacement sensors, stress sensors, humidity sensors, and rainwater monitoring sensors, etc.

[0074] Step S2: Compare the water composition imaging data of the landslide body with the preset water composition threshold to obtain the rainfall amount; automatically adjust the valve opening of the drainage channel and diversion channel according to the rainfall amount; calculate the stability parameter distribution based on the rainfall amount;

[0075] Step S3: Obtain the safety factor based on the stability parameter distribution, and obtain the instability probability of the slope through the safety factors of multiple grid units; if the instability probability reaches the start threshold, automatically control the grouting equipment to start; when grouting reaches the grouting threshold, shut down the grouting equipment; optimize the structural design of the metal reinforcement piles and reinforcement frame through finite element analysis and make adjustments.

[0076] Preferably, in this embodiment, step S1 involves placing various sensors, including displacement sensors, stress sensors, humidity sensors, and rainwater monitoring sensors, inside the metal reinforcement and at key locations on the landslide body to monitor landslide data in real time. The monitored landslide data is divided into multiple grid cells for easier subsequent analysis and processing. Under rainfall conditions, real-time imaging data of the slope's water composition, including the water content of each grid cell, is acquired. This embodiment, through the deployment of multiple sensors, can acquire multi-dimensional data of the landslide body in real time, including displacement, stress, humidity, and rainfall, providing basic data for subsequent analysis. Dividing the landslide body into multiple grid cells helps to analyze the deformation characteristics and water composition distribution of the landslide body more precisely, improving the accuracy and reliability of monitoring. By acquiring water composition imaging data, the water content distribution inside the landslide body can be intuitively understood, providing an important basis for assessing landslide stability. In step S2, the water composition imaging data of the landslide body is compared with a preset water composition threshold to determine the rainfall amount. The opening of the valves in the drainage system and diversion channels is automatically adjusted according to the rainfall amount to control the water content of the landslide body. The stability parameter distribution is calculated based on the rainfall amount to assess the overall stability of the landslide body. This embodiment automatically adjusts the drainage system to promptly remove excess water from the landslide body, preventing landslide risks caused by water accumulation. The stability parameter distribution calculated using rainfall and water composition data allows for dynamic assessment of landslide stability, providing a scientific basis for early warning and prevention. Combined with real-time monitoring data and preset thresholds, the system can quickly respond to rainfall events and adjust drainage measures in a timely manner, reducing the likelihood of landslides. In step S3, a safety factor is calculated based on the stability parameter distribution to assess the instability probability of the landslide body. The instability probability of the slope body is comprehensively assessed using the safety factors of multiple grid units. When the instability probability reaches the activation threshold, the grouting equipment is automatically activated; otherwise, the equipment is shut down. This embodiment optimizes the structural design of the metal reinforcement piles and reinforcement frame through finite element analysis to improve overall stability. This embodiment calculates the safety factor and instability probability to provide early warning of the possibility of landslides, thus providing a time window for emergency response. When the instability probability reaches the threshold, the system automatically starts the grouting equipment for reinforcement, reducing manual intervention and improving response speed. This embodiment optimizes the design of the reinforcement structure through finite element analysis, improving the overall stability and durability of the landslide body and reducing the risk of future landslides.

[0077] Furthermore, such as Figure 2 As shown, step S1, which involves acquiring real-time water composition imaging data of the slope under rainfall conditions, specifically includes the following steps:

[0078] Step S11: Place at least one sensor inside the metal reinforcement and at key locations on the landslide body to monitor landslide data in real time, divide the landslide data into multiple grid units, and arrange multiple electrodes according to the multiple grid units.

[0079] Step S12: Measure the resistivity of multiple grids; obtain the resistivity at any location on the landslide body in real time; perform suction balance on the resistivity to obtain the corresponding resistance value, and obtain the corresponding water content based on the resistance value;

[0080] Step S13: Obtain the corresponding resistivity based on the resistance value, fit the moisture content and resistivity to obtain the correspondence between resistivity and moisture content; discretize the resistivity at different locations in the corresponding grid cells to obtain the resistivity corresponding to each grid cell; input the resistivity of each grid cell into the correspondence between resistivity and moisture content to obtain the moisture content of each grid cell.

[0081] Preferably, in step S11 of this embodiment, at least one sensor is placed inside the metal reinforcement and at a key location on the landslide body to monitor landslide data in real time. These sensors can include tilt sensors, displacement sensors, etc., capable of monitoring parameters such as deformation, displacement, and vibration of the landslide body; by deploying a group of sensors, monitoring points are formed, and data such as vibration frequency, tilt offset increment, tilt offset direction, and displacement increment of the landslide body are collected in real time; in step S12, the landslide body data is divided into multiple grid cells, each grid cell serving as the basic unit for resistivity measurement; multiple electrodes are arranged according to the grid cells, and the resistivity distribution inside the landslide body is measured using the high-density resistivity method; the resistivity of multiple grid cells is measured, and resistivity data at any location of the landslide body is acquired in real time; the resistivity is subjected to suction balance processing to obtain the corresponding resistance value, and the water content is calculated based on the resistance value; in step S13, the water content and resistivity are fitted to obtain the correspondence between resistivity and water content; the resistivity at different locations is discretized in the corresponding grid cells to obtain the resistivity corresponding to each grid cell; the resistivity of each grid cell is input into the correspondence between resistivity and water content to obtain the water content of each grid cell. In this embodiment, a sensor array and high-density resistivity method are used to monitor the deformation and moisture changes of landslide bodies in real time and accurately, improving the accuracy and real-time performance of landslide early warning. Real-time monitoring of resistivity changes in landslide bodies allows for rapid response to changes in the internal structure of the landslide body, providing reliable data support for landslide early warning. Utilizing resistivity imaging technology and multivariate data fusion methods, multidimensional imaging and analysis of the internal structure of landslide bodies can be achieved, accurately determining the location of potential slip surfaces and the stability of the landslide body. Combining data such as soil moisture content and rainfall, a multivariate data fusion model is constructed to improve the accuracy and real-time performance of landslide prediction. The high-density resistivity method can measure the resistivity distribution without disturbing the object being measured, improving the spatial resolution of the monitoring data. Through the design and optimization of the electrode network, the needs of large-area hazard point investigation and variable density and range monitoring in key areas can be met.

[0082] Furthermore, such as Figure 3 As shown, the process of obtaining the stability parameter distribution in step S2 specifically includes the following steps:

[0083] Step S21: Compare the water composition imaging data of the landslide body with the preset water composition threshold, and extract features from the landslide body data by combining the landslide body's own evolution law and landslide deformation mechanism to obtain the rainfall-induced landslide impact feature set;

[0084] The rainfall-induced landslide impact feature set includes historical displacement data features, rainfall factor features, stability coefficient features, rainfall infiltration depth features, and time factor features.

[0085] Step S22: Generate a predicted displacement sequence based on the characteristics of historical displacement data and the feature set of rainfall-induced landslide impacts; use the predicted displacement sequence to determine the predicted landslide time; issue an early warning based on the predicted landslide time and upload it to the central control system.

[0086] Step S23: The central control system analyzes the landslide time and automatically adjusts the valve opening of the drainage channel and diversion channel based on the analysis results; it also analyzes the predicted rainfall to obtain the distribution of stability parameters.

[0087] Preferably, in step S21 of this embodiment, the water composition imaging data of the landslide body is compared with a preset water composition threshold, and features of the landslide body data are extracted by combining the landslide body's own evolution law and landslide deformation mechanism. A rainfall-induced landslide impact feature set is extracted from the landslide body data, including historical displacement data features, rainfall factor features, stability coefficient features, rainfall infiltration depth features, and time factor features. Step S21, by combining the landslide body's own evolution law and deformation mechanism, can more accurately identify potential landslide risks, providing a scientific basis for subsequent early warning. Furthermore, by extracting multiple features, the stability of the landslide can be comprehensively assessed, thereby improving the overall performance of the early warning system. In step S22, based on historical displacement data features, a predicted displacement sequence is generated from the rainfall-induced landslide impact feature set. The predicted displacement sequence is used to determine the possible time of the landslide, and an early warning is issued based on the predicted time. The early warning information is uploaded to the central control system for further processing and response. Step S22, through the predicted displacement sequence, can dynamically predict the time of landslide occurrence, thereby achieving early warning and reducing disaster losses. Uploading the early warning information to the central control system enables multi-departmental collaborative response and improves emergency response efficiency. Step S23: The central control system analyzes the landslide time and automatically adjusts the valve opening of the drainage and diversion channels based on the analysis results; it analyzes the predicted rainfall to obtain the distribution of stability parameters of the landslide body, and through the automatic adjustment function of the central control system, it can adjust the drainage and diversion measures in a timely manner to reduce the risk of landslide; and through the analysis of rainfall and stability parameters, it can achieve refined management of the landslide body and improve the treatment effect.

[0088] Furthermore, such as Figure 4 As shown, step S21, which obtains the feature set of rainfall-induced landslide impacts, specifically includes the following steps:

[0089] Step S211: Collect historical displacement data characteristics; combine the landslide body's own evolution law and the mechanism of rainfall's effect on landslide deformation, and analyze rainfall factors based on historical average rainfall, previous effective rainfall, and number of rainfall days;

[0090] Step S212: Determine the importance of rainfall factors and their characteristics; determine the stability coefficient and rainfall infiltration depth characteristics based on the permeability coefficient, shear strength, and landslide slope; conduct characteristic analysis based on different monthly rainfall amounts to determine the characteristics of time factors;

[0091] Step S213: Classify time features based on the similarity between them; determine the feature set of rainfall-induced landslide impact based on historical displacement data features, rainfall factor features, stability coefficient features, rainfall infiltration depth features, and time factor features.

[0092] Preferably, in step S211, the historical displacement data of the landslide body is monitored to extract its evolution pattern; combined with the landslide body's own evolution pattern, the mechanism of rainfall's effect on landslide deformation is analyzed, including historical average rainfall, previous effective rainfall, and number of rainfall days; by combining historical displacement data with rainfall factors, the landslide deformation trend can be predicted more accurately, reducing medium- and long-term prediction errors; considering the impact of rainfall on landslide deformation makes the model more adaptable and practical. Step S212 performs an importance analysis of rainfall factors and extracts key features; based on the permeability coefficient, shear strength, and landslide slope, the stability coefficient and rainfall infiltration depth characteristics are determined; based on rainfall in different months, feature analysis is performed to determine time factor characteristics; in step S212, the model parameters are optimized through the importance analysis of rainfall factors to improve the model's interpretability and prediction accuracy; and through the stability coefficient and infiltration depth characteristics, the stability of the landslide is better assessed, providing a scientific basis for prevention and control measures. Step S213 classifies time features based on the similarity between them; combines historical displacement data features, rainfall factor features, stability coefficient features, rainfall infiltration depth features, and time factor features to determine the rainfall-induced landslide impact feature set; by determining the comprehensive feature set, the risk level of landslides under different conditions can be comprehensively assessed, providing more comprehensive data support for landslide prevention and control; combining the feature analysis of multiple factors improves the accuracy and timeliness of landslide early warning.

[0093] Furthermore, such as Figure 5 As shown, step S22, which predicts the landslide time for early warning, specifically includes the following steps:

[0094] Step S221: Calculate the deformation acceleration point based on the curve between the reciprocal of the landslide velocity and time; determine the predicted landslide time based on the deformation acceleration point and the predicted displacement sequence; use the difference between the predicted landslide time and the current time as an early warning indicator.

[0095] Step S222: Construct a landslide monitoring and early warning mechanism based on early warning indicators and landslide movement rate; detect and issue landslide warnings according to the landslide monitoring and early warning mechanism; and upload the real-time monitored landslide early warning data to the central control system.

[0096] Step S223: The central control system analyzes the real-time monitored landslide early warning data in conjunction with the rainfall-induced landslide impact feature set to predict the landslide phenomenon that will occur at the corresponding time.

[0097] Preferably, in step S221 of this embodiment, by analyzing the relationship between the reciprocal of the landslide velocity and time, the deformation acceleration point is identified, and the landslide time is predicted based on this point; using the displacement data after the deformation acceleration point, combined with the landslide creep theory, a landslide time prediction model is established to improve the prediction accuracy; the difference between the predicted landslide time and the current time is used as an early warning indicator to assess the urgency of the landslide; in step S221, by using the velocity reciprocal method and identifying the deformation acceleration point, the landslide occurrence time can be predicted more accurately, reducing the possibility of false alarms and missed alarms; the calculation of the early warning indicator provides a scientific basis for real-time monitoring and early warning, enabling timely detection of landslide risks and providing a time window for emergency response. Step S222 establishes a comprehensive early warning system that combines early warning indicators and landslide movement rates to dynamically monitor and issue early warnings for landslides. Real-time landslide early warning data is uploaded to the central control system to ensure the timeliness and effectiveness of the information. By combining early warning indicators and landslide movement rates, the early warning system in step S222 can more comprehensively assess landslide risks, reducing false alarms or missed alarms caused by a single indicator. The uploading of real-time data allows the central control system to obtain the latest landslide information in a timely manner, providing reliable data support for emergency response. In step S223, the central control system analyzes the real-time monitoring data using a rainfall-induced landslide impact characteristic set to predict the specific time of landslide occurrence. Furthermore, by analyzing the impact of rainfall and other environmental factors on landslides, it predicts specific phenomena during landslide occurrence, such as crack expansion and accelerated displacement. Combining the impact characteristics of rainfall and other environmental factors allows for more accurate prediction of the time and specific phenomena of landslide occurrence, providing a scientific basis for disaster prevention and mitigation. This embodiment, by predicting the specific phenomena of landslide occurrence, allows relevant departments to take targeted emergency measures in advance, such as personnel evacuation and facility reinforcement, to minimize disaster losses.

[0098] Furthermore, such as Figure 6 As shown, step S23, which obtains the stability parameter distribution, specifically includes the following steps:

[0099] Step S231: Analyze and calculate the landslide phenomenon that occurred at the corresponding time of the landslide in conjunction with historical landslide phenomena to obtain the valve opening value; automatically adjust the valve opening of the drainage channel and the diversion channel according to the valve opening value;

[0100] Step S232: Based on the feature set of rainfall-induced landslide impacts, the soil-water characteristic relationship is derived; based on the relationship between soil resistivity and volumetric water, the relationship between matrix suction and soil resistance is obtained; by introducing suction stress, the relationship between suction stress and water content is obtained.

[0101] Step S233: Obtain the stress at any location on the landslide body, replace the pore pressure with absorbed stress, and obtain the relationship between absorbed stress and effective stress; analyze the relationship between absorbed stress and effective stress in conjunction with rainfall to obtain the distribution of stability parameters.

[0102] In step S231, the calculation of the valve switch value and the adjustment of the valve opening are based on historical data, which is usually presented in time series form, such as annual rainfall or monthly displacement. To eliminate dimensional differences, the historical data is normalized.

[0103]

[0104] In the formula, T ref It is a reference time used to convert historical data into dimensionless values; T norm,i T represents the normalized value of the i-th historical data point; history,i This represents the i-th historical data (such as rainfall, displacement, etc.).

[0105] The normalized historical data is combined with the current feature parameters and assigned different weighting coefficients k1 and k2:

[0106]

[0107] In the formula, k1 and k2 are determined through experiments or experience, reflecting the degree of influence of historical and current data on the valve opening and closing values; B switch This represents the valve on / off value, a weighted result combining historical data and current characteristic parameters; P current The characteristic parameters of the current landslide phenomenon (such as current rainfall, current displacement, etc.) are represented; n represents the number of historical data.

[0108] The valve opening is adjusted based on an exponential function to ensure a smooth transition in opening degree with changes in the on / off value.

[0109]

[0110] When V switch When V approaches 0, the valve opening θ also approaches 0; when V... switch Much greater than V ref At this time, the valve opening θ approaches the maximum opening θ. max ;

[0111] Step S232 Soil-water characteristic relationship and suction stress calculation: The soil-water characteristic relationship describes the relationship between soil moisture content and matrix suction.

[0112]

[0113] In the formula, the soil moisture content θ w The relationship between the soil and the matrix suction ψ is nonlinear; parameters α, n, and m are determined through experimental data fitting, reflecting the soil's pore distribution and water-holding capacity; residual water content θ r and saturated water content θ s These are the basic physical properties of soil;

[0114] The relationship between matrix suction and soil resistivity, where matrix suction ψ passes through soil resistivity ρ. e Indirect measurement:

[0115] ψ=a·ln(ρ e )+b

[0116] In the formula, the soil resistivity ρ e With moisture content θ w Closely related; by fitting experimental data, parameters a and b were determined, and the logarithmic relationship between matrix suction and resistivity was established;

[0117] The relationship between absorbent stress and moisture content, absorbent stress σ suction It is a function of soil moisture content:

[0118]

[0119] The absorption stress decreases with increasing moisture content; by fitting experimental data, parameters c, d, and e are determined, and a power function relationship between absorption stress and moisture content is established.

[0120] Step S233: Relationship between absorbed stress and effective stress, and distribution of stability parameters. In unsaturated soil, the absorbed stress σ suction It replaced the role of pore water pressure:

[0121] σ′=σ-u a +χ·σ suction

[0122] In the formula, the total stress σ is determined by the soil's self-weight and the external load; the pore air pressure u a Typically close to atmospheric pressure; the weighting coefficient χ reflects the contribution of absorbed stress to effective stress, and its value is related to the soil saturation; σ′ represents effective stress, which is the actual stress borne between particles in the soil.

[0123] Stability parameters (safety factor F) s The calculation is based on the Mohr-Coulomb criterion:

[0124]

[0125] In the formula, the shear strength τ f Determined by the effective cohesion c′ and the effective internal friction angle φ′:

[0126] τ f =c′+σ′·tan(φ′)

[0127] The shear stress τ is determined by the soil's self-weight and the landslide surface dip angle:

[0128] τ=γ·H·sin(α)

[0129] Safety factor F s It is the ratio of shear strength to shear stress, reflecting the stability of the landslide; it quantitatively analyzes the distribution of landslide stability parameters and provides a scientific basis for landslide prevention and control; γ represents the unit weight of the soil, which is the weight of a unit volume of soil; H represents the height of the landslide body; α represents the dip angle of the landslide surface.

[0130] Preferably, in step S231 of this embodiment, by analyzing historical landslide phenomena corresponding to the landslide at the corresponding time and combining them with real-time monitoring data, the valve opening value is calculated. Based on the valve opening value, the valve opening of the drainage channel and the diversion channel is automatically adjusted to achieve dynamic management of drainage in the landslide area. Step S231, by adjusting the valve opening in real time, can more effectively remove accumulated water in the landslide area and reduce the possibility of landslides. Combining historical data and real-time monitoring, landslide risks can be predicted more accurately, allowing for early intervention and reducing the impact of landslides on the environment and infrastructure. In step S232, based on the impact feature set of rainfall-induced landslides, a soil-water characteristic relationship is established. By introducing the relationship between soil resistivity and volumetric water, the relationship between absorbent stress and water content is obtained. Absorbent stress is used to replace pore pressure to obtain the relationship between absorbent stress and effective stress. By establishing the soil-water characteristic relationship and the absorbent stress model, step S232 can more accurately assess the possibility of landslides and the distribution of their stability parameters. Combining parameters such as rainfall and soil suction, landslide risks can be identified earlier, providing a scientific basis for disaster prevention and mitigation. Step S233 obtains the stress at any location on the landslide body, replaces the pore pressure with absorbed stress, and obtains the relationship between absorbed stress and effective stress; analyzes the data in conjunction with rainfall to obtain the distribution of stability parameters; by monitoring the stress changes of the landslide body in real time and combining them with rainfall, step S233 can dynamically assess the stability of the landslide and adjust drainage and support measures in a timely manner; based on the distribution of stability parameters, a more scientific landslide treatment plan can be formulated to improve the treatment effect.

[0131] Furthermore, such as Figure 7 As shown, step S3, which optimizes the structural design of the metal reinforced piles and reinforcement frame through finite element analysis, specifically includes the following steps:

[0132] Step S31: Obtain the safety factor of each grid cell based on the stability parameter distribution; obtain the total number of grid cells and count the number of grid cells with a local safety factor less than 1; obtain the percentage of the number of grid cells with a local safety factor less than 1 to the total number of grid cells and analyze its distribution to obtain the instability probability of the slope.

[0133] Step S32: Compare the instability probability with the preset instability probability threshold. When the instability probability reaches the start threshold, the grouting equipment is automatically started. When the grouting reaches the grouting threshold, the grouting equipment is shut down, and the instability probability is recalculated. If it is greater than the start threshold, the grouting equipment is automatically started. This process is repeated until the instability probability is lower than the start threshold.

[0134] Step S33: Extract features from the stability parameter distribution, adjust the rainfall-landslide impact feature dataset based on the extracted features, and perform data fusion so that the adjusted rainfall-landslide impact feature dataset at different time points is in a high-dimensional data space. Establish a dynamic map to obtain the quantitative results of the distribution of influencing factors at each location of the metal reinforcement piles and reinforcement frames.

[0135] Preferably, in step S31 of this embodiment, the slope is divided into multiple grid units, and the safety factor of each grid unit is calculated through the distribution of stability parameters. Step S31, through gridded analysis, can more accurately assess the slope's instability probability, providing a basis for subsequent early warning and control. Real-time acquisition of the instability probability distribution helps to promptly identify potential landslide risks and improve the accuracy of early warnings. In step S32, the instability probability is compared with a preset threshold. When the instability probability reaches the activation threshold, the grouting equipment is automatically activated; when the grouting reaches the grouting threshold, the grouting equipment is shut down. The instability probability is recalculated; if it is still greater than the activation threshold, the grouting equipment continues to be activated until the instability probability is lower than the threshold. This embodiment reduces manual intervention and improves response speed and efficiency through automated control of the grouting equipment. Dynamic adjustment of the grouting volume based on the real-time instability probability ensures slope stability while avoiding resource waste. In step S33, features of stability parameter distribution are extracted, and the rainfall-induced landslide impact feature dataset is adjusted based on these features. A dynamic map is established in a high-dimensional data space to quantify the distribution of influencing factors at various locations of metal reinforcement piles and reinforcement frames. Through the high-dimensional dynamic map, the distribution of influencing factors at different locations can be understood more intuitively, providing a scientific basis for engineering design and construction. Reinforcement measures are dynamically adjusted based on real-time data to improve the accuracy and effectiveness of reinforcement.

[0136] Furthermore, such as Figure 8 As shown, step S33, which involves obtaining the quantitative results of the distribution of influencing factors at various locations of the metal reinforced piles and reinforcement frames, specifically includes the following steps:

[0137] Step S331: Combine the adjusted rainfall impact feature datasets at each time point into a high-dimensional vector, and obtain the preprocessed and normalized standardized data through data preprocessing and normalization; calculate the covariance matrix based on the standardized data, and perform eigenvalue decomposition on the covariance matrix to obtain eigenvalues ​​and corresponding eigenvectors.

[0138] Assuming there are m′ time points, and each time point has n′ features (such as rainfall, temperature, humidity, etc.), the dataset can be represented as a matrix X with dimensions m′×n′. To eliminate the differences in the units of measurement between features, the data needs to be standardized. The standardization formula is:

[0139]

[0140] In the formula, μ is the mean vector of each feature; σ″ is the standard deviation vector of each feature; and the standardized data X 标准化 It satisfies the condition that the mean is 0 and the variance is 1, which facilitates the calculation of the covariance matrix;

[0141] Calculation of the covariance matrix, standardized data X 标准化 It is directly used to calculate the covariance matrix C. The formula for the covariance matrix is:

[0142]

[0143] In the formula: X 标准化 It is a standardized data matrix; X 标准化 T It is X 标准化 The transpose of the matrix has dimensions (n′×m′); m′-1 is the sample size minus one, used for unbiased estimation of the covariance matrix; the covariance matrix describes the linear correlation between the features.

[0144] Eigenvalue decomposition: The covariance matrix C is decomposed using the following formula:

[0145] Cv l =λ l v l

[0146] In the formula, C is the covariance matrix with dimensions (n′×m′); λ l It is the l-th eigenvalue, representing the magnitude of the variance of the data along the corresponding eigenvector direction; v L It is the l-th feature vector, representing the change pattern of data in the corresponding direction;

[0147] The goal of eigenvalue decomposition is to find the eigenvalues ​​and eigenvectors of the covariance matrix C such that:

[0148] C = VΛV T

[0149] In the formula, V is the eigenvector matrix, and each column is an eigenvector v. i The dimension is (n′×n′); Λ is a diagonal matrix, and the elements on the diagonal are the eigenvalues ​​λ. iThe dimension is (n′×n′).

[0150] Step S332: Based on the magnitude of the eigenvalues, select the largest eigenvalue and the corresponding eigenvector to form the principal components; project the standardized data based on the principal components to map the high-dimensional data to the low-dimensional space, and draw the data points at each time point in the low-dimensional space to construct a dynamic spectrum.

[0151] Step S333: Based on the distribution and trend of the dynamic spectrum data points, perform quantitative analysis to obtain a statistical summary of the impact at each time point or in each region; record this as the quantitative result of the distribution of influencing factors of the metal reinforced piles and reinforcement frames; adjust the structure of the metal reinforced piles and reinforcement frames based on the quantitative results.

[0152] Quantitative analysis includes examining the distribution of data points in a low-dimensional space over time and identifying outliers.

[0153] Preferably, in this embodiment, the adjusted rainfall impact feature datasets for each time point are combined into a high-dimensional vector, and standardized data are obtained through data preprocessing and normalization. This step ensures that all features are on the same scale, which helps improve the accuracy of subsequent analysis. The standardization method typically includes subtracting the mean and dividing by the standard deviation, so that the mean of the data is 0 and the variance is 1. The covariance matrix is ​​calculated based on the standardized data. The covariance matrix describes the linear relationship between the data and is a symmetric matrix. The covariance matrix is ​​decomposed into eigenvalues ​​to obtain eigenvalues ​​and corresponding eigenvectors. The eigenvalues ​​represent the variance of the data in the feature space, while the eigenvectors represent the direction of the data in the feature space. Based on the magnitude of the eigenvalues, the largest eigenvalue and the corresponding eigenvector are selected to form principal components. These principal components are used to project the high-dimensional data into a low-dimensional space, thereby achieving dimensionality reduction. The standardized data is projected into a new space composed of the selected principal components to obtain the dimensionality-reduced data. The data points for each time point are plotted in the low-dimensional space to construct a dynamic spectrum. Dynamic spectra can visually display the distribution of data points over time; quantitative analysis can examine the distribution and trends of data points in low-dimensional space, identifying outliers. Principal component analysis projects high-dimensional data into a low-dimensional space while preserving as much of the original data's key information as possible. This helps reduce computational complexity and improve data analysis efficiency. By selecting the direction of maximum variance as the principal component, noise and redundant information can be effectively removed, thereby improving model performance and accuracy. Dynamic spectra can visually display the distribution of data points over time, helping to identify outliers or trend changes; they are of great significance for monitoring and adjusting metal reinforced piles and reinforcement structures. Quantitative analysis can provide a statistical summary of the impact at each time point or region, providing a basis for subsequent structural adjustments.

[0154] Furthermore, such as Figure 9 As shown, step S333, which involves adjusting the metal reinforcement piles and reinforcement frame structure based on the quantification results, specifically includes the following steps:

[0155] Step S3331: Based on the dynamic spectrum, analyze the distribution of data points at each time point or in each region; identify the change patterns and potential outliers in the data over time; quantify the data distribution in the dynamic spectrum.

[0156] Step S3332: Calculate the cumulative contribution points of the eigenvalues ​​and eigenvectors at each time point or each region, evaluate the explanatory power of different principal components on data changes based on the cumulative contribution points, and the variance contribution value at each time point or each region, and determine the contribution of principal components to the main direction of data change.

[0157] Step S3333: Identify key data and outliers based on the dynamic spectrum, and adjust the structural design of the metal reinforcement piles and reinforcement frames based on the contribution of principal components to the main direction of data change; recalculate the adjusted data, compare the calculation results with the preset structural threshold, and readjust if the result is greater; iterate in this way until the calculation result is less than the preset structural threshold.

[0158] Step S3331 represents dynamic spectral analysis and data quantization, with the dynamic spectral data distribution quantization formula as follows:

[0159]

[0160] In the formula, Q dyn (t,r) represents the dynamic spectrum quantization value at time point t and region r; S ij (t,r) represents the data point in the i-th row and j-th column at time t and region r; μ ij (t,r) represents the mean of the data at time point t and in region r; σ ij (t,r) represents the standard deviation of the data at time point t and region r; t0 represents the reference time point; τ represents the time decay coefficient, which controls the time weight; n″,m″ represent the number of rows and columns of the dynamic spectrum;

[0161] Anomaly identification formula:

[0162]

[0163] In the formula, A(t,r) represents the quantized value of the outlier at time point t and region r; S i (t,r) represents the i-th data point at time t and in region r; μ i (t,r) represents the mean data at time point t and region r; σ i(t,r) represents the standard deviation of the data at time point t and region r; r0 represents the reference region; ρ represents the spatial attenuation coefficient, which controls the weight of the region; k represents the total number of data points;

[0164] Step S3332 represents principal component analysis and contribution value calculation, with the formula for cumulative contribution points of eigenvalues:

[0165]

[0166] In the formula, C cum (t,r) represents the cumulative contribution point of the eigenvalues ​​at time point t and region r; λ p (t,r) represents the p-th feature value at time t and region r; α′ represents the attenuation coefficient, which controls the weight of the feature value.

[0167] Formula for variance contribution value:

[0168]

[0169] In the formula, V cont (t,r) represents the variance contribution values ​​at time point t and region r; λ p (t,r) represents the p-th feature value at time t and region r; φ p (t,r) represents the p-th eigenvector at time t and region r; P represents the total number of principal components;

[0170] Step S3333 represents structural design and iterative adjustment, with the structural adjustment formula as follows:

[0171]

[0172] In the formula, D adj (t,r) represents the structural adjustment value at time point t and region r; λ p (t,r) represents the p-th feature value at time t and region r; φ p (t,r) represents the p-th feature vector at time t and region r;

[0173] Iterative adjustment formula:

[0174]

[0175] In the formula, I iter (t,r) represents the iterative adjustment value for time point t and region r; D adj (t,r) represents the structural adjustment value; the above formulas achieve quantitative optimization of metal reinforced piles and reinforced frame structures through dynamic spectral analysis, principal component analysis and iterative adjustment; each formula includes mathematical operations and weight control to ensure the accuracy and reliability of the calculation results.

[0176] Preferably, in step S3331 of this embodiment, the distribution of data points at each time point or region is analyzed through dynamic spectral analysis to identify the change patterns and potential outliers in the data over time; the data distribution in the dynamic spectral analysis is quantified to more accurately describe the distribution characteristics of the data; this embodiment, through dynamic spectral analysis, can promptly discover change patterns and potential outliers in the data, providing a basis for subsequent data processing; quantifying the data distribution helps to more accurately describe the characteristics of the data, thereby improving the accuracy of the overall data analysis; in step S3332, the eigenvalues ​​and eigenvectors of each time point or region are calculated, and their cumulative contribution points are evaluated; based on the cumulative contribution points, the explanatory power of different principal components on data changes, as well as the variance contribution value of each time point or region, are evaluated to determine the contribution of principal components to the main direction of data change; this embodiment, through principal component analysis, projects high-dimensional data into a low-dimensional space while retaining key information, reducing noise, and improving data processing efficiency; evaluating the explanatory power of principal components helps to understand the main direction of data change, thereby providing a scientific basis for subsequent adjustments. In step S3333, the structural design of the metal reinforced piles and reinforcement frame is adjusted based on the key data and outliers identified in the dynamic spectrum and the results of principal component analysis. The adjusted data is recalculated and compared with a preset structural threshold. If the result is greater than the threshold, it is readjusted, and this process is iterated until the result is less than the threshold. In this embodiment, the structural design of the metal reinforced piles and reinforcement frame is optimized through the results of dynamic spectrum and principal component analysis, thereby improving the safety and stability of the structure. Through iterative optimization, the final design result is ensured to meet the preset safety standards, reducing potential risks.

[0177] like Figures 10-12As shown, this embodiment also provides an embodiment of a landslide real-time monitoring system based on multi-sensor fusion and intelligent control. In this embodiment, the landslide real-time monitoring system based on multi-sensor fusion and intelligent control is applied to the landslide real-time monitoring method based on multi-sensor fusion and intelligent control as described in the above embodiment. The landslide real-time monitoring system based on multi-sensor fusion and intelligent control includes driving the bottom ends of multiple metal reinforcement piles 2 into the slope surface of the landslide body 1 in a matrix manner. The metal reinforcement piles 2 have a hollow internal structure, and multiple grouting ports 21 are opened on the outer wall of the metal reinforcement piles 2. The multiple grouting ports 21 are evenly distributed on the outer wall of the metal reinforcement piles 2. Then, concrete grout is injected into the multiple metal reinforcement piles 2. The concrete grout penetrates into the landslide body 1 through the grouting ports 21, waiting for the concrete to be poured into the soil. After the grout solidifies, the metal reinforcement piles 2 can be fixed inside the slope of the landslide body 1, thus reinforcing the landslide body 1 through grouting. The multiple metal reinforcement piles 2 are divided into four vertical rows, and each of the four rows of metal reinforcement piles 2 is equipped with a reinforcement frame 3. There are four reinforcement frames 3 in total, and the bottom of the four reinforcement frames 3 can be fixed to the top of the four vertical rows of metal reinforcement piles 2 by bolts. A drainage channel 8 is dug on one side of the landslide surface of the landslide body 1, and multiple diversion channels 9 are dug on the landslide surface of the landslide body 1. The multiple diversion channels 9 are inclined, and the downward inclined end of the drainage channel 8 is connected to the interior of the drainage channel 8. Four reinforcement seats 6 are evenly provided on the top surface of the landslide body 1. A connecting locking ring 5 is fixedly installed on the top of each of the four reinforcement frames 3, and the other end of the connecting locking ring 5 is installed on one side of the reinforcement seat 6.

[0178] Each of the four reinforcing seats 6 has an anchor rod 7 at its bottom. By driving the anchor rod 7 into the interior of the top surface of the landslide body 1, the reinforcing seat 6 can be fixed to the top surface of the landslide body 1. Multiple connecting rods 4 are provided between each of the four reinforcing frames 3. The two ends of the connecting rod 4 are fixedly connected between two reinforcing frames 3 by bolts. The arrangement of multiple reinforcing frames 3 connects the four reinforcing frames 3 together, increasing the stability between them. Multiple diversion channels 9 are used to intercept and divert rainwater from the landslide surface of the landslide body 1. When rain falls on the landslide surface of the landslide body 1... Rainwater flows into multiple diversion channels 9 and is discharged into drainage channels 8 through the downward sloping ends of the diversion channels 9. It is then discharged through the drainage channels 8, which allows for the diversion and drainage of rainwater from the landslide surface of the landslide body 1, thereby managing the rainwater flow from the landslide surface of the landslide body 1. Filter screens are fixedly installed at the top of both the drainage channels 8 and the diversion channels 9 to filter leaves, dead branches, or other debris from the water, preventing debris from entering the drainage channels 8 and the diversion channels 9 and causing blockages.

[0179] Multiple metal reinforcement piles 2 are fixedly connected to both sides of the top of the piles 2. The connecting plates 22 are provided with threaded openings 23. The bottom of the reinforcement frame 3 is provided with threaded installation openings corresponding to the threaded openings 23. When installing the metal reinforcement piles 2 and the reinforcement frame 3, bolts can be used to fix the connecting plates 22 and the top of the metal reinforcement piles 2 to the bottom of the reinforcement frame 3 through the threaded openings 23. The depth of the bottom of the multiple metal reinforcement piles 2 is 1.3 to 1.5 times the thickness of the landslide body 1.

[0180] Preferably, in use, the bottom ends of multiple metal reinforcement piles 2 are driven into the slope surface of the landslide body 1 in a matrix manner, so that the depth of the bottom ends of the multiple metal reinforcement piles 2 is 1.3 to 1.5 times the thickness of the landslide body 1. Then, concrete grout is injected into the multiple metal reinforcement piles 2. The concrete grout penetrates into the landslide body 1 through the grouting port 21. After the concrete grout solidifies, the metal reinforcement piles 2 can be fixed inside the slope of the landslide body 1, thereby grouting and reinforcing the landslide body 1. Then, the bottom parts of the multiple reinforcement frames 3 are... Do not place it on top of multiple vertical metal reinforcement piles 2. Then, use bolts to fix the connecting plate 22 and the top of the metal reinforcement pile 2 to the bottom of the reinforcement frame 3 through the threaded joint 23. Then, install connecting rods 4 between multiple reinforcement frames 3 with bolts, so that multiple reinforcement frames 3 are connected together through multiple connecting rods 4, increasing the stability between multiple reinforcement frames 3. Then, drive the anchor rod 7 at the bottom of the reinforcement seat 6 into the top surface of the landslide body 1 to fix the reinforcement seat 6. The reinforcement seat 6 is connected to the metal reinforcement piles 2 and the reinforcement base 3 through the connecting locking ring 5. The reinforcement frame 3 is further tightened to prevent the metal reinforcement piles 2 from sliding down, further improving the landslide prevention effect. During rainy weather, rainwater from the landslide surface is intercepted and diverted into multiple diversion channels 9, and then discharged into drainage channels 8, allowing for drainage of the landslide surface of the landslide body 1, thereby controlling the rainwater flow on the landslide surface. Multiple metal reinforcement piles are driven into the slope of the landslide body in a matrix arrangement, and then... Concrete grout is injected and penetrates into the landslide body through the injection port, reinforcing the landslide body to support the sliding force of the landslide and stabilize the landslide. Multiple reinforcement frames and connecting rods are used to connect multiple metal reinforcement piles together, increasing the stability between the piles. The metal reinforcement piles and frames are further secured by connecting locking rings, reinforcement seats, and anchor rods to prevent the piles from sliding down, thus further improving the landslide prevention effect.

[0181] In this embodiment, during rainy weather, rainwater from the landslide surface is intercepted and diverted into multiple diversion channels 9. The water is then discharged into a drainage channel 8 through one end of each channel, and finally discharged through the drainage channel 8. This allows for drainage of the landslide surface of the landslide body 1, effectively controlling the rainwater flow. Simultaneously, the inclusion of filters removes leaves, branches, or other debris from the water, preventing them from entering the drainage and diversion channels 8 and causing blockages. The anchor rods 7 at the bottom of the reinforcement base 6 are driven into the top surface of the landslide body 1 to secure the reinforcement base 6. The reinforcement base 6 is further tightened to the reinforcement frame 3 and the metal reinforcement pile 2 via connecting locking rings 5, preventing the metal reinforcement pile 2 from sliding down and further improving the landslide prevention effect. A drainage channel 8 is dug on one side of the landslide surface of landslide body 1, and multiple diversion channels 9 are dug on the landslide surface of landslide body 1. The multiple diversion channels 9 are inclined, and the downward inclined end of the drainage channel 8 is connected to the interior of the drainage channel 8. The multiple diversion channels 9 are used to intercept and divert rainwater on the landslide surface of landslide body 1. When it rains, the rainwater falling on the landslide surface of landslide body 1 flows into the multiple diversion channels 9 and is discharged into the drainage channel 8 through the downward inclined end of the diversion channel 9, and then discharged through the drainage channel 8. This allows for the diversion and drainage of rainwater on the landslide surface of landslide body 1, thereby managing the rainwater on the landslide surface of landslide body 1. The top of the drainage channel 8 and the diversion channels 9 are fixedly installed with filter screens to filter leaves, dead branches or other garbage in the water, preventing garbage from entering the drainage channel 8 and the diversion channels 9 and causing blockage.

[0182] In this embodiment, multiple metal reinforcement piles 2 are driven into the slope surface of the landslide body 1 in a matrix manner, with the depth of the bottom of the multiple metal reinforcement piles 2 being 1.3 to 1.5 times the thickness of the landslide body 1, ensuring the stability of the metal reinforcement piles 2. Then, concrete grout is injected into the multiple metal reinforcement piles 2, and through the setting of grouting port 21, some of the concrete grout inside the metal reinforcement piles 2 can penetrate into the landslide body 1. After the concrete grout solidifies, the metal reinforcement piles 2 can be fixed inside the slope of the landslide body 1, thereby grouting and reinforcing the landslide body 1. Then, the bottom of multiple reinforcement frames 3 is placed on top of multiple vertical metal reinforcement piles 2, and bolts are used to fix the metal reinforcement piles 2 to the reinforcement frames 3, thereby increasing the stability between the vertical metal reinforcement piles 2. Then, connecting rods 4 are installed between the multiple reinforcement frames 3 with bolts, so that the multiple reinforcement frames 3 are connected together through multiple connecting rods 4, further increasing the stability between the multiple metal reinforcement piles 2.

[0183] like Figure 13As shown, this embodiment provides an embodiment of an electronic device. In this embodiment, the electronic device 10 includes a processor 81 and a memory 82 coupled to the processor 81.

[0184] The memory 82 stores program instructions for implementing the landslide real-time monitoring method based on multi-sensor fusion and intelligent control in any of the above embodiments.

[0185] The processor 81 is used to execute program instructions stored in the memory 82 to perform real-time monitoring of landslides based on multi-sensor fusion and intelligent control.

[0186] The processor 81 can also be referred to as a CPU (Central Processing Unit). The processor 81 may be an integrated circuit chip with signal processing capabilities. The processor 81 can also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. A general-purpose processor can be a microprocessor or any conventional processor.

[0187] Furthermore, Figure 14 This is a schematic diagram of the structure of a storage medium according to an embodiment of this application. The storage medium 9 of this embodiment stores program instructions 91 capable of implementing all the methods described above. These program instructions 91 can be stored in the storage medium in the form of a software product, including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets.

[0188] In the several embodiments provided by this invention, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between apparatuses or units through some interfaces, and may be electrical, mechanical, or other forms.

[0189] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated units described above can be implemented in hardware or as software functional units. The above are merely embodiments of the present invention and do not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

[0190] The specific embodiments of the invention have been described in detail above, but these are merely examples, and the invention is not limited to the specific embodiments described above. For those skilled in the art, any equivalent modifications or substitutions to the invention are also within the scope of this invention. Therefore, all equivalent transformations, modifications, and improvements made without departing from the spirit and principles of this invention should be included within the scope of this invention.

Claims

1. A landslide real-time monitoring method based on multi-sensor fusion and intelligent control, comprising a landslide real-time monitoring system including multiple metal reinforcement piles. The bottom ends of the multiple metal reinforcement piles are driven into the slope surface of the landslide body in a matrix manner. The metal reinforcement piles have a hollow internal structure, and multiple grouting ports are evenly distributed on the outer wall of the metal reinforcement piles. Concrete grout is then injected into the multiple metal reinforcement piles. The concrete grout penetrates into the landslide body through the grouting ports, and the process is completed. The process involves solidifying and fixing metal reinforcement piles inside the slope of the landslide body, thereby reinforcing the landslide body with grout. Multiple metal reinforcement piles are arranged in four vertical rows, each row topped with a reinforcement frame. The bottoms of the four reinforcement frames are bolted to the tops of the four vertical rows of metal reinforcement piles. A drainage channel is excavated on one side of the landslide surface, and multiple diversion channels are excavated on the landslide surface. These diversion channels are inclined, with their downward-sloping ends connected to the interior of the drainage channel. The process is characterized by... The real-time landslide monitoring method based on multi-sensor fusion and intelligent control specifically includes the following steps: At least one sensor is placed inside the metal reinforcement pile and at key locations on the landslide body to monitor landslide data in real time. The landslide data is divided into multiple grid cells. The water composition imaging data of the landslide body is also acquired in real time under rainfall conditions. The water composition imaging data includes the water content of each grid cell; the sensors include displacement sensors, stress sensors, humidity sensors, and rainwater monitoring sensors. The water composition imaging data of the landslide body is compared with the preset water composition threshold to obtain the rainfall amount; the valve opening of the drainage channel and diversion channel is automatically adjusted according to the rainfall amount; the stability parameter distribution is calculated based on the rainfall amount. The safety factor is obtained based on the distribution of stability parameters, and the instability probability of the landslide body is obtained through the safety factors of multiple grid units; if the instability probability reaches the activation threshold, the grouting equipment is automatically activated; when the grouting reaches the grouting threshold, the grouting equipment is shut down; the structural design of the metal reinforcement piles and reinforcement frame is optimized and adjusted through finite element analysis. The process of acquiring real-time water composition imaging data of a landslide body under rainfall conditions includes the following steps: At least one sensor is placed inside the metal reinforcement pile and at key locations on the landslide body to monitor landslide data in real time. The landslide data is then divided into multiple grid units. Multiple electrodes are then arranged according to the multiple grid units. The resistivity of multiple grid cells is measured; the resistivity at any location in the landslide body is acquired in real time; the resistivity is balanced by suction to obtain the corresponding resistance value, and the corresponding water content is obtained based on the resistance value; The moisture content and resistivity are fitted to obtain the corresponding relationship between resistivity and moisture content; the resistivity at different locations is discretized into the corresponding grid cells to obtain the resistivity of each grid cell; the resistivity of each grid cell is input into the resistivity-moisture content correspondence to obtain the moisture content of each grid cell. The process of obtaining the stability parameter distribution includes the following steps: The water composition imaging data of the landslide body is compared with the preset water composition threshold. Combined with the evolution law of the landslide body itself and the landslide deformation mechanism, the landslide body data is feature extracted to obtain the feature set of rainfall-induced landslide impact. The rainfall-induced landslide impact feature set includes historical displacement data features, rainfall factor features, stability coefficient features, rainfall infiltration depth features, and time factor features. Based on the characteristics of historical displacement data, a predicted displacement sequence is generated from the feature set of rainfall-induced landslide impacts. The predicted displacement sequence is used to determine the predicted landslide time. An early warning is issued based on the predicted landslide time and uploaded to the central control system. The central control system analyzes the predicted landslide time and automatically adjusts the valve openings of drainage and diversion channels based on the analysis results; it also analyzes the predicted rainfall to obtain the distribution of stability parameters. The process of optimizing the structural design of metal reinforced piles and reinforcement frames through finite element analysis includes the following steps: The safety factor of each grid cell is obtained based on the stability parameter distribution; the total number of grid cells is obtained, and the number of grid cells with a local safety factor less than 1 is counted; the percentage of grid cells with a local safety factor less than 1 to the total number of grid cells is obtained and their distribution is analyzed to obtain the instability probability of the landslide body. The instability probability is compared with a preset instability probability threshold. When the instability probability reaches the activation threshold, the grouting equipment is automatically started. When the grouting reaches the grouting threshold, the grouting equipment is shut down, and the instability probability is recalculated. If the instability probability is greater than the activation threshold, the grouting equipment is automatically started. This process is iterated until the instability probability is lower than the activation threshold. Features are extracted from the distribution of stability parameters. Based on the extracted features, the rainfall-landslide impact feature set is adjusted and data is fused to establish a dynamic map in a high-dimensional data space, so that the adjusted rainfall-landslide impact feature set at different time points is in a high-dimensional data space. This yields the quantitative results of the distribution of influencing factors at each location of the metal reinforcement piles and reinforcement frames.

2. The landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to claim 1, characterized in that, The process of obtaining the feature set of rainfall-induced landslide impacts includes the following steps: Collect historical displacement data characteristics; combine the landslide body's own evolution law and the mechanism of rainfall's effect on landslide deformation, and analyze rainfall factors based on historical average rainfall, previous effective rainfall, and number of rainfall days; The importance of rainfall factors is determined to identify their characteristics; the stability coefficient and rainfall infiltration depth are determined based on permeability coefficient, shear strength, and landslide slope; and the time factor characteristics are determined based on different monthly rainfall amounts. The time factor characteristics are classified according to the similarity between them; the rainfall landslide impact feature set is determined based on historical displacement data characteristics, rainfall factor characteristics, stability coefficient characteristics, rainfall infiltration depth characteristics, and time factor characteristics.

3. The landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to claim 1, characterized in that, The process of issuing early warnings based on predicted landslide times includes the following steps: The deformation acceleration point is calculated based on the curve between the reciprocal of the landslide velocity and time; the predicted landslide time is determined using the predicted displacement sequence based on the deformation acceleration point; and the difference between the predicted landslide time and the current time is used as an early warning indicator. Construct a landslide monitoring and early warning mechanism based on early warning indicators and landslide displacement rate; monitor and issue landslide warnings according to the landslide monitoring and early warning mechanism; and upload the real-time monitored landslide early warning data to the central control system. The central control system analyzes real-time landslide early warning data in conjunction with rainfall-induced landslide impact characteristic sets to predict the landslide phenomenon that will occur at the corresponding time.

4. The landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to claim 1, characterized in that, The process of obtaining the stability parameter distribution includes the following steps: The landslide phenomenon occurring at the corresponding time is analyzed and calculated in conjunction with historical landslide phenomena to obtain valve opening and closing values; the valve opening of drainage channels and diversion channels is automatically adjusted based on the valve opening and closing values. Based on the feature set of rainfall-induced landslides, the soil-water relationship is derived; based on the relationship between soil resistivity and volume water, the relationship between matrix suction and soil resistance is obtained; by introducing suction stress, the relationship between suction stress and water content is obtained. The stress at any location on the landslide body is obtained, and the absorption stress is used to replace the pore pressure to obtain the relationship between the absorption stress and the effective stress. Based on the relationship between the absorption stress and the effective stress, combined with the rainfall, the distribution of stability parameters is obtained.

5. The landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to claim 1, characterized in that, The process of obtaining the quantitative results of the distribution of influencing factors at various locations of metal reinforced piles and reinforcement frames specifically includes the following steps: The adjusted rainfall-landslide impact feature sets at each time point are combined into a high-dimensional vector. The data is then preprocessed and normalized to obtain standardized data. The covariance matrix is ​​calculated based on the standardized data, and eigenvalues ​​and corresponding eigenvectors are obtained by eigenvalue decomposition. Based on the magnitude of the eigenvalues, the largest eigenvalue and its corresponding eigenvector are selected to form the principal components. Based on the principal components, the standardized data is projected to map the high-dimensional data to a low-dimensional space, and the data points at each time point are plotted in the low-dimensional space to construct a dynamic graph. Based on the distribution and trend of dynamic graph data points, quantitative analysis is performed to obtain a statistical summary of the impact at each time point or in each region; this is recorded as the quantitative result of the distribution of influencing factors of metal reinforced piles and reinforcement frames; the structure of metal reinforced piles and reinforcement frames is adjusted based on the quantitative results. Quantitative analysis includes examining the changes in the distribution of data points in low-dimensional space over time and identifying outliers.

6. The landslide real-time monitoring method based on multi-sensor fusion and intelligent control according to claim 5, characterized in that, The process of adjusting the metal reinforcement piles and reinforcement frame structure based on the quantitative results includes the following steps: Based on dynamic graphs, analyze the distribution of data points at each time point or in each region; identify the changing patterns and potential outliers in the data over time; and quantify the data distribution in the dynamic graphs. Calculate the cumulative contribution points of eigenvalues ​​and eigenvectors at each time point or region, evaluate the explanatory power of different principal components on data changes based on the cumulative contribution points, and determine the variance contribution value at each time point or region to determine the main direction of data change. Based on the identification of key data and outliers in the dynamic graph, and combined with the contribution of principal components to the main direction of data change, the structural design of metal reinforcement piles and reinforcement frames is adjusted; the adjusted data is then recalculated, and the calculation results are compared with the preset structural threshold. If the calculation result is greater than the preset structural threshold, the adjustment is repeated; this process is iterated until the calculation result is less than the preset structural threshold.

Images