Landslide intelligent monitoring and early warning method and precise prevention and ecological collaborative restoration system

Abstract

The invention discloses an intelligent landslide monitoring and early warning method and a precise prevention and ecological collaborative restoration system, and relates to the technical field of geological disaster monitoring and early warning. The method comprises the following steps: acquiring InSAR area deformation rate data and GNSS point position displacement rate data, performing space-time fusion by adopting a Kalman filtering algorithm, and outputting fused earth surface deformation data; acquiring deep deformation data, a pore water pressure value and temperature data, and performing compensation processing on the deep deformation data; integrating the fused earth surface deformation data, the compensated deep deformation data, the pore water pressure value and the temperature data into multi-source synchronous time sequence data; the LSTM model is utilized to predict the multi-source synchronous time sequence data; and carrying out landslide early warning judgment according to the predicted multi-source synchronous time sequence data. Through combination of multi-source data fusion and prediction, the landslide risk rising trend can be identified in advance, the early warning time window is significantly prolonged, and time is won for disaster prevention and reduction.

Description

Technical Field

[0001] This invention belongs to the field of geological disaster prevention and control engineering technology, specifically involving a landslide intelligent monitoring and early warning method and a precise prevention and control and ecological collaborative restoration system. Background Technology

[0002] Landslides, as a typical geological hazard, are often accompanied by the instability and sliding of soil and rock masses. Currently, landslide monitoring and early warning technologies mainly rely on point displacement monitoring methods such as Global Navigation Satellite Systems (GNSS). However, existing technologies still have the following shortcomings:

[0003] First, achieving both high spatiotemporal resolution and high accuracy in monitoring data is difficult. Single monitoring methods have inherent limitations. For example, while GNSS monitoring can acquire high-precision absolute displacement data, it is "point-based monitoring" and cannot comprehensively reflect the overall deformation field characteristics of a landslide. While Synthetic Aperture Radar Interferometry (InSAR) technology can achieve "area" coverage and acquire regional continuous deformation fields, its monitoring results are susceptible to spatiotemporal decoherence and atmospheric delay. Currently, monitoring is merely a data listing, lacking a deep spatiotemporal fusion mechanism, leading to contradictions in the spatial continuity and temporal consistency of monitoring data, making it difficult to construct a high-precision, high-reliability surface deformation field.

[0004] Secondly, landslide early warning criteria are often simplistic, relying on fixed threshold triggering mechanisms. Traditional early warning models typically set fixed thresholds based on a single indicator (such as displacement rate or cumulative displacement), issuing an alarm only when monitored data exceeds this threshold. This "threshold-triggered" mode exhibits significant lag; by the time the deformation rate exceeds the limit, the landslide body has often already entered an accelerated deformation or even imminent landslide stage, leaving very little time for evacuation and emergency response. Furthermore, fixed thresholds are difficult to adapt to the geological evolution patterns of different landslide bodies, easily leading to false alarms or missed alarms, severely impacting the reliability and practicality of the early warning system.

[0005] Finally, existing technologies do not fully utilize multi-source monitoring data and lack forward-looking predictive capabilities. Landslide evolution is a complex process influenced by geological conditions, hydrological environment, and external stresses; single monitoring data is insufficient to comprehensively depict its evolutionary mechanism. Although some schemes attempt to integrate multi-source data, they mostly remain at the data-level fusion stage, failing to effectively integrate and synchronously analyze key inducing factors such as deep deformation, pore water pressure, and temperature with surface deformation data, resulting in inaccurate assessments of landslide evolution trends. More importantly, most existing technologies can only assess the current stable state and lack the ability to predict future trends, failing to identify rising risks before disasters occur. Summary of the Invention

[0006] This invention aims to overcome the problems of insufficient monitoring accuracy and coverage in existing landslide control technologies, and provides a landslide intelligent monitoring and early warning method and a precise prevention and ecological collaborative restoration system. It can accurately predict future multi-source synchronous time-series data, identify the rising trend of potential risks in advance before the macroscopic deformation of the landslide body accelerates, effectively improve the timeliness, predictability and practicality of landslide early warning, and achieve better disaster prevention results.

[0007] To achieve the above objectives, the present invention employs the following technical solution:

[0008] In a first aspect, the present invention provides a landslide intelligent monitoring and early warning method, comprising the following steps:

[0009] Acquire InSAR data, GNSS data, pore water pressure values, raw acceleration and angular velocity data from MEMS inclinometers, and temperature data; InSAR stands for Synthetic Aperture Radar Interferometry, and GNSS stands for Global Navigation Satellite System.

[0010] Differential interferometry is performed on InSAR data to obtain deformation rate data of the InSAR region;

[0011] Carrier phase differential processing is performed on GNSS data to obtain GNSS point displacement rate data;

[0012] Based on the InSAR regional deformation rate data and GNSS point displacement rate data, the Kalman filter algorithm is used to perform spatiotemporal fusion and output the fused surface deformation data.

[0013] Preliminary data on deep deformation of MEMS are obtained from the original acceleration and angular velocity data of MEMS inclinometer. Combined with temperature data and a pre-established temperature drift compensation model, the preliminary data on deep deformation of MEMS is compensated in real time to obtain the compensated deep deformation data; MEMS stands for micro-motor system.

[0014] The fused surface deformation data, compensated deep deformation data, pore water pressure values ​​and temperature data are integrated into multi-source synchronous time-series data.

[0015] Input multi-source synchronous time series data with a preset time window length into a trained time series prediction model to predict multi-source synchronous time series data for future time periods.

[0016] Based on the multi-source synchronous time-series data of the future time period, the fused surface deformation rate, compensated deep deformation rate, pore water pressure value and temperature data are obtained, and the landslide stability coefficient is calculated in combination with the pre-stored test parameters.

[0017] The technical effects achieved by the above settings are as follows: Through multi-level data processing and fusion, the accuracy of monitoring data and the reliability of early warning are significantly improved. The Kalman filter algorithm combines the wide-area coverage advantage of InSAR and the high-precision positioning advantage of GNSS to output spatiotemporally continuous fused surface deformation data, which not only ensures the monitoring range but also improves local accuracy. The temperature drift compensation model effectively eliminates the interference of temperature changes on deep deformation monitoring, making deep deformation data more accurate. The time series prediction model can automatically learn the importance weights of different features in multi-source time series data and accurately predict future deformation trends. By combining the predicted data to calculate the landslide stability coefficient and comparing it with multi-level early warning thresholds, a leap from "post-event alarm" to "pre-event prediction" is achieved, allowing sufficient time for disaster response and making early warning information more accurate and timely.

[0018] Simultaneously, Kalman filtering fuses sparse, high-precision GNSS point data with continuous but noisy InSAR surface data. This process is equivalent to "denoising" and "spatiotemporal registration" of the original data, generating "clean" time-series data that possesses both spatial continuity and temporal consistency. The fused multi-source synchronous time-series data eliminates noise interference such as atmospheric delay and spatiotemporal decoherence, enabling the LSTM model to more accurately learn the true physical trends and intrinsic laws of landslide evolution, rather than learning errors from the data acquisition process. This is a prerequisite for accurate prediction.

[0019] Furthermore, differential interferometry is performed on the InSAR data to obtain InSAR region deformation rate data, including:

[0020] By acquiring two synthetic aperture radar (SAR) images of the same area at different times, the surface deformation is inverted using the phase difference between the images, forming a time series of deformation:

[0021]

[0022] in, The total phase difference, The phase difference is caused by surface deformation. The phase difference is caused by topographic relief. The phase difference is caused by atmospheric delay. This is the noise phase.

[0023] Deformation The calculation is performed using the following formula:

[0024]

[0025] in, For deformable variables, For SAR image wavelength, The radar incident angle is π, where π is the mathematical constant pi.

[0026] Based on the time series of deformation, InSAR regional deformation rate data are obtained.

[0027] Furthermore, carrier phase differential processing is performed on the GNSS data to obtain GNSS point displacement rate data, including: receiving BeiDou + GPS satellite signals, using carrier phase differential technology to obtain the three-dimensional coordinates of the monitoring point, calculating the coordinate changes at different times as displacement values; and obtaining GNSS point displacement rate data based on the time series of displacement values.

[0028] Furthermore, based on the InSAR regional deformation rate data and GNSS point displacement rate data, a Kalman filter algorithm is used for spatiotemporal fusion to output fused surface deformation data, including:

[0029] Using InSAR regional deformation rate data as prior values ​​and GNSS point displacement rate data as observed values, the time series of spatiotemporally continuous fused surface deformation data is output. The fusion formula is as follows:

[0030]

[0031]

[0032] in, This is the fused deformation state vector. This is InSAR regional deformation rate data. For GNSS point displacement rate data, For Kalman gain, For the observation matrix, To predict the error covariance, To observe the noise covariance, the superscript T indicates transpose.

[0033] The technical effects achieved by the above settings are as follows: The Kalman filter fusion formula provides a specific calculation method for the fusion of InSAR and GNSS data, and the optimal estimation is achieved by dynamically adjusting the weights through Kalman gain.

[0034] Furthermore, the time series prediction model is an improved LSTM intelligent early warning model, which is a long short-term memory neural network based on an attention mechanism. Its input is multi-source synchronous time series data, and its output is multi-source synchronous time series data for future time periods.

[0035] Furthermore, the temperature drift compensation model is as follows:

[0036]

[0037] in, The data is the compensated deep deformation data. Preliminary data on deep deformation of MEMS. This is the temperature drift coefficient. This represents the change in temperature.

[0038] Furthermore, the method for establishing the temperature drift compensation model includes: acquiring reference data on the deep deformation of the fiber grating; establishing a temperature drift compensation model based on the acquired reference data on the deep deformation of the fiber grating and temperature data; calibrating the temperature drift coefficient of the MEMS using the high stability of the fiber grating data; performing real-time compensation on the preliminary data of the deep deformation of the MEMS; and outputting the compensated deep deformation data.

[0039] Furthermore, methods for calculating landslide stability coefficients include:

[0040] The Morganston-Price method from the limit equilibrium method is used to determine this, as shown in the following formula:

[0041]

[0042] in, This is the landslide stability coefficient.

[0043] For sandstone cohesion, The segment length of the sliding surface,

[0044] For segmented soil and rock gravity, The inclination angle of the sliding surface segment,

[0045] The internal friction angle of the rock and soil. This refers to the pore water pressure.

[0046] The technical effects achieved by the above settings are as follows: the temperature drift compensation model uses a simple and efficient linear formula to correct for the influence of temperature; the Morganston-Price method, as a classic limit equilibrium method, calculates the stability coefficient in real time based on monitoring data, providing a solid theoretical foundation for stability assessment. These formulas together constitute the core algorithm of the intelligent monitoring and early warning subsystem, ensuring the scientific nature and accuracy of data processing and early warning judgment.

[0047] Furthermore, the preset warning thresholds include:

[0048] Attention level: The surface deformation rate after integration is 0.3-0.5 mm / day, or the deep deformation rate after compensation is 0.1-0.3 mm / day, or the landslide stability coefficient is 1.15-1.25;

[0049] Preparation level: The surface deformation rate after fusion is 0.5-1 mm / day, or the deep deformation rate after compensation is 0.3-0.5 mm / day, or the landslide stability coefficient is 1.05-1.15;

[0050] Action level: The surface deformation rate after fusion is greater than 1 mm / day, or the deep deformation rate after compensation is greater than 0.5 mm / day, or the landslide stability coefficient is less than 1.05.

[0051] The technical effects achieved by the above settings are as follows: By setting three levels of early warning thresholds, refined hierarchical management of landslide risk is realized. The "Attention" level indicates "Attention," the "Preparation" level indicates "Preparation," and the "Action" level indicates "Immediate Action." Different levels correspond to different emergency response measures, avoiding false alarms or missed alarms caused by a single threshold. The threshold range comprehensively considers three dimensions of indicators: surface displacement, deep deformation, and stability coefficient, forming a multi-parameter joint judgment mechanism, significantly improving the accuracy and reliability of early warnings and providing a scientific basis for disaster prevention and mitigation.

[0052] Furthermore, the method also includes:

[0053] Real-time InSAR regional deformation rate data and GNSS point displacement rate data, compensated deep deformation rate, and landslide stability coefficient are acquired and compared with preset real-time warning thresholds to obtain the corresponding real-time warning level.

[0054] Attention level: InSAR area deformation rate data is 0.3-0.5 mm / day, or GNSS point displacement rate data is 0.5-1 mm / day, or compensated deep deformation rate is 0.1-0.3 mm / day, or stability coefficient meets 1.15-1.25;

[0055] Preparation level: InSAR area deformation rate data is 0.5-1 mm / day, or GNSS point displacement rate data is 1-3 mm / day, or compensated deep deformation rate is 0.3-0.5 mm / day, or stability coefficient meets 1.05-1.15;

[0056] Action level: InSAR area deformation rate data >1 mm / day, or GNSS point displacement rate data >3 mm / day, or compensated deep deformation rate >0.5 mm / day, or stability coefficient <1.05.

[0057] The technical effects achieved by the above settings are as follows: A multi-parameter joint early warning mechanism using real-time parameters further improves the accuracy of early warnings, preventing inaccurate or untimely prediction models. Different monitoring technologies (InSAR, GNSS, MEMS) reflect the slope deformation state from different dimensions; joint judgment effectively avoids false alarms that may occur with a single technology. The stability coefficient, as a comprehensive indicator, corroborates the deformation rate, forming a double guarantee. The setting of three threshold levels, combined with engineering practice, provides clear and operable decision-making basis for on-site management personnel.

[0058] Secondly, the present invention provides a precision prevention and control and ecological synergistic restoration system, comprising:

[0059] A solid waste-based low-carbon grouting reinforcement subsystem includes a grouting conduit and a solid waste-based low-carbon grouting material. The solid waste-based low-carbon grouting material is injected into the grouting conduit to form a deep reinforcement area. The grouting conduit is deployed within the landslide body and penetrates the potential sliding surface.

[0060] An ecological slope protection subsystem includes a lattice-structured ecological concrete frame, foundation anchors, and a vegetation system. The lattice-structured ecological concrete frame is laid on the slope surface of the landslide body. The foundation anchors anchor the lattice-structured ecological concrete frame to the landslide body. The vegetation system is installed within the lattice-structured ecological concrete frame.

[0061] The intelligent monitoring and early warning subsystem includes:

[0062] The surface displacement monitoring module is used to acquire InSAR and GNSS data of the landslide area;

[0063] The deep deformation monitoring module is used to acquire the raw acceleration and angular velocity data of the MEMS inclinometer in the landslide area;

[0064] A pore water pressure monitoring module is used to acquire pore water pressure values ​​in the landslide area.

[0065] The micro-meteorological monitoring module is used to acquire temperature data in the landslide area;

[0066] The data transmission system is used to connect the cloud platform server with the surface displacement monitoring module, the deep deformation monitoring module, the pore water pressure monitoring module, and the micro-meteorological monitoring module.

[0067] The cloud platform server is used to perform the method described in the first aspect to predict and warn of landslide deformation based on the data uploaded by the surface displacement monitoring module, deep deformation monitoring module, pore water pressure monitoring module and micro-meteorological monitoring module.

[0068] The technical effects achieved by the above settings are as follows: This invention realizes integrated landslide management through the structural coupling and functional synergy of the solid waste-based low-carbon grouting reinforcement subsystem, the ecological slope protection subsystem, and the intelligent monitoring and early warning subsystem. First, the solid waste-based low-carbon grouting reinforcement subsystem uses all-solid waste-based materials (steel slag, fly ash, etc.) to prepare grouting materials, replacing traditional cement-based materials and significantly reducing carbon emissions. Simultaneously, grouting pipes penetrate potential sliding surfaces and inject grout to form deep reinforcement zones, effectively improving the deep stability of the slope. Second, the ecological slope protection subsystem uses solid waste-based ecological concrete of the same origin as the grouting materials to prepare a lattice frame, which is anchored to the slope using foundation anchors. Combined with a vegetation system, it forms a shallow ecological slope protection structure, balancing structural stability and ecological compatibility, avoiding the ecological damage problems caused by traditional "gray hardened slopes." Third, the intelligent monitoring and early warning subsystem integrates multi-source monitoring technologies such as InSAR, GNSS, MEMS, and fiber Bragg gratings. Data fusion and intelligent early warning are achieved through a cloud platform server, realizing comprehensive and high-precision monitoring from the macroscopic surface to deep soil and rock masses, solving the problems of disconnect between treatment and monitoring and inaccurate early warning information in traditional technologies.

[0069] The three subsystems work together: deep grouting reinforcement provides a stable foundation for shallow ecological slope protection, shallow ecological slope protection provides an installation carrier for monitoring equipment, and intelligent monitoring and early warning provide decision-making basis for long-term operation and maintenance of the system, forming a deep collaborative governance model of "deep reinforcement - shallow ecology - intelligent monitoring", which effectively solves the technical defects of existing technologies such as high carbon emissions, large ecological damage, and insufficient monitoring accuracy.

[0070] Meanwhile, the intelligent monitoring and early warning subsystem acquires deformation data from both deep and shallow layers in real time and analyzes the slope condition through a cloud platform server. When the monitoring data approaches the early warning threshold, it can promptly guide the maintenance of the grouting reinforcement area or ecological slope protection structure (such as local grouting and vegetation replanting), realizing a shift from passive repair to proactive prevention, and effectively improving the safety and durability of the solid waste-based low-carbon grouting reinforcement subsystem and the ecological slope protection subsystem.

[0071] Furthermore, the surface displacement monitoring module includes an InSAR monitoring unit and a GNSS monitoring unit;

[0072] The InSAR monitoring unit is used to cover the landslide area and buffer zone, and to acquire InSAR data of the landslide area;

[0073] The GNSS monitoring units are deployed at key locations and in the anomalous areas of the InSAR monitoring units to acquire GNSS data of the landslide area.

[0074] The deep deformation monitoring module includes a MEMS inclinometer and a fiber Bragg grating inclinometer, which are staggered in the landslide body and work together to locate the sliding surface.

[0075] The pore water pressure monitoring module includes a vibrating wire pore water pressure gauge embedded in the landslide body.

[0076] The data transmission system adopts LoRa+4G / 5G dual-mode communication.

[0077] The technical effects achieved by the above setup are as follows: Through the coordinated deployment of InSAR and GNSS, a surface displacement monitoring network combining "area-point" coverage is realized. InSAR provides wide-area coverage to identify anomaly areas, while GNSS provides intensive monitoring at key points, thus controlling costs while ensuring monitoring accuracy. MEMS and fiber optic inclinometers are deployed alternately, utilizing the high sensitivity of MEMS to capture minute deformations and the high stability of fiber optic gratings to provide a reference. Together, they can accurately locate the sliding surface. Vibrating wire pore water pressure gauges monitor groundwater pressure changes in real time, providing key parameters for stability calculations. LoRa+4G / 5G dual-mode communication ensures stable data transmission in complex mountainous environments, avoiding data interruptions caused by signal blind spots.

[0078] Furthermore, the solid waste-based low-carbon grouting material comprises, by weight, 40-60 parts steel slag powder, 20-30 parts fly ash, 10-20 parts slag powder, 5-10 parts alkali activator, 0.5-1.5 parts thickener, and 20-30 parts water.

[0079] The alkali activator is a composite solution of water glass with a modulus of 1.2-1.5 and NaOH with a concentration of 8-12 mol / L mixed in a mass ratio of 3:1, and the thickener is hydroxypropyl methylcellulose.

[0080] The technical effects achieved by the above settings are as follows: By optimizing the proportions of solid waste-based grouting materials, a balance between material performance and environmental benefits is achieved. Steel slag powder, fly ash, and slag powder are all industrial solid wastes; their comprehensive utilization significantly reduces material costs and reduces solid waste stockpiling. Alkali activators activate the solid waste materials at room temperature, giving them gelling properties, replacing traditional cement clinker, and significantly reducing carbon emissions. The thickener hydroxypropyl methylcellulose improves the stability and pumpability of the grout, ensuring the quality of grouting construction. Under this proportion, the grouting material has a 28-day compressive strength ≥30MPa, a lower permeability coefficient, and a shrinkage rate ≤0.5%, with mechanical properties and durability meeting engineering requirements.

[0081] Furthermore, the lattice-type ecological concrete frame is made of solid waste-based low-carbon ecological concrete of the same origin as the solid waste-based low-carbon grouting material. Its lattice form is rectangular, with a grid size of 2m×2m, cross-sectional dimensions of horizontal and vertical beams of 300mm×400mm, 28-day compressive strength ≥20MPa, flexural strength ≥3MPa, porosity of 15-20%, and reinforcing ribs and monitoring holes are provided at the nodes.

[0082] The basic anchors are 32mm diameter threaded steel bars, which are implanted into the stable soil layer of the slope. The pull-out force of a single anchor is ≥80kN, and they are arranged one-to-one with the grid nodes.

[0083] The vegetation system is a native grass-shrub composite plant, whose root system forms a solidified body with the foundation anchor and the solid waste-based low-carbon grouting material.

[0084] The technical effects achieved by the above configuration are as follows: The lattice-structured ecological concrete frame is made from materials of the same origin as the grouting material, ensuring material compatibility and deformation coordination, and avoiding stress concentration at the interfaces of different materials. Through stability coefficient verification, the grid size, cross-sectional dimensions, and mechanical properties of the frame have been optimized to meet structural load-bearing requirements while reserving space for vegetation growth; the node reinforcing ribs improve the overall integrity of the frame, and the reserved monitoring holes facilitate the installation of monitoring equipment later. The foundation anchors correspond one-to-one with the lattice nodes, forming a reliable force transmission path and ensuring that the frame and the slope share the load collaboratively. The interception and drainage system effectively guides the slope runoff, reducing the adverse effects of rainwater infiltration on slope stability; porous concrete and permeable hoses enhance drainage efficiency and prevent water accumulation from damaging the structure and vegetation.

[0085] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0086] 1. This invention constructs a high-precision spatiotemporal continuous deformation field, achieving complementary advantages between area monitoring and point-based monitoring. This invention innovatively introduces a Kalman filter algorithm for spatiotemporal fusion of InSAR regional deformation data and GNSS point displacement data. This algorithm not only effectively utilizes the complementarity of the two, overcoming the noise caused by the spatiotemporal decoherence of InSAR data, but also compensates for the sparsity of GNSS points, outputting fused deformation data with both high spatial and temporal resolution. Compared to traditional single monitoring methods, this invention can more comprehensively and accurately depict the true deformation state of landslides from the surface to their deeper layers.

[0087] 2. This invention achieves spatiotemporal synchronization and multidimensional feature correlation of multi-source heterogeneous data, improving the accuracy of landslide evolution analysis. This invention integrates fused high-precision surface deformation data, compensated deep deformation data (such as underground displacement and strain), pore water pressure values, and temperature data into multi-source synchronous time-series data under a unified spatiotemporal benchmark. This multi-dimensional data integration constructs a comprehensive monitoring system encompassing "surface deformation" to "deep disturbances," and "direct displacement" to "inducing factors." Joint analysis of these synchronous data allows for a deeper understanding of the intrinsic connections and evolutionary patterns among various physical fields during landslide evolution, providing a more reliable basis for stability assessment.

[0088] 3. This invention possesses forward-looking early warning capabilities, transforming passive response into proactive prevention and control. This invention is not limited to evaluating current data, but further utilizes a Long Short-Term Memory (LSTM) artificial neural network to perform deep learning and mining on the integrated multi-source synchronous time-series data. The LSTM network can effectively capture long-term dependencies and evolutionary trends in time-series data, thereby enabling accurate prediction of multi-source synchronous time-series data for the next few days. By analyzing the future trends reflected in the predicted data, this invention can identify the rising trend of potential risks days or even weeks in advance, before the macroscopic deformation of the landslide accelerates.

[0089] 4. This invention significantly extends the early warning window, greatly improving the effectiveness of disaster prevention and mitigation. Addressing the lag inherent in traditional threshold-triggered early warning systems, this invention uses future data predicted by LSTM as the basis for judgment, achieving a leap from "post-event alerts" to "pre-event predictions" in the early warning mechanism. This predictive early warning is based on future trends.

[0090] 5. This invention provides reliable fault tolerance by introducing a parallel real-time early warning channel. When the LSTM prediction model becomes distorted due to significant shifts in the input data distribution or deviations in the model itself, the system can issue an early warning based on real-time InSAR deformation, GNSS displacement, deep deformation rate, and stability coefficient. This mechanism ensures that the landslide monitoring and early warning system will not be paralyzed due to the failure of a single model under any circumstances, providing a bottom-line safety guarantee for geological disaster prevention and control.

[0091] 6. The system of this invention integrates multi-source sensing technologies such as InSAR, GNSS, MEMS, and fiber optic inclinometers to collect real-time deformation, hydrological, and meteorological data of the slope's "macroscopic surface - microscopic positioning - deep soil and rock mass." Combined with machine learning models, it achieves refined assessment and early warning of landslide risk, providing sufficient time for disaster response. In terms of mechanical performance, the coupling mechanism of grouting reinforcement + lattice-type slope protection + various plant root systems forms a three-dimensional slope reinforcement system with a high landslide stability coefficient. The lattice framework transfers loads to stable strata, improving overall resistance to overturning and sliding. In terms of ecological effects, solid waste-based biochar enhances soil water and fertilizer retention capacity and adsorbs heavy metals. Plant roots penetrate the filling matrix and slope soil layers, forming a "biological reinforcement network," improving the ecological landscape of the slope and improving the regional microclimate. Attached Figure Description

[0092] Figure 1 This is a flowchart of the landslide intelligent monitoring and early warning method of the present invention;

[0093] Figure 2 This is a schematic diagram of the overall system for precise prevention and control and ecological synergistic restoration of the present invention;

[0094] Figure 3 This is a flowchart illustrating the installation and usage method of the precision prevention and ecological synergistic restoration system of the present invention. Detailed Implementation

[0095] The present invention will be further described below with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and should not be used to limit the scope of protection of the present invention.

[0096] Example 1:

[0097] like Figure 1 As shown in the figure, this embodiment provides a landslide intelligent monitoring and early warning method, including the following steps:

[0098] Step 1: Acquire InSAR data, GNSS data, pore water pressure values, raw wavelength data from fiber optic inclinometers, raw acceleration and angular velocity data from MEMS inclinometers, and temperature data;

[0099] InSAR data comes from InSAR technology, which acquires SAR images from remote sensing satellites covering the landslide area and a surrounding 1-2km buffer zone to achieve wide-area surface deformation monitoring.

[0100] GNSS data comes from GNSS monitoring stations, which are deployed at key locations and in the anomalous areas of InSAR monitoring units. They use BeiDou + GPS dual-mode reception and acquire high-precision point displacement data through carrier phase differential technology.

[0101] The pore water pressure value comes from a vibrating wire pore water pressure gauge embedded in the slope. The raw frequency data of the pore water pressure is collected in real time and converted to obtain the pore water pressure value, which provides key parameters for the calculation of the landslide stability coefficient.

[0102] The raw wavelength data from the fiber optic grating inclinometer and the raw acceleration and angular velocity data from the MEMS inclinometer are obtained from both instruments, which are interleaved in the monitoring holes within the landslide body. The MEMS inclinometer integrates a triaxial accelerometer and a gyroscope to acquire raw acceleration and angular velocity data; the fiber optic grating inclinometer acquires raw wavelength data. Working together, the MEMS inclinometer provides highly sensitive continuous deformation monitoring, while the fiber optic grating inclinometer provides highly stable reference data, jointly and accurately locating the position and deformation characteristics of the potential sliding surface.

[0103] Temperature data comes from micro-meteorological modules (such as temperature sensors) to acquire meteorological data such as temperature, humidity, and rainfall in the landslide area. The temperature data is used for temperature drift compensation in deep deformation monitoring.

[0104] Step 2: Perform differential interferometry on the InSAR data to remove terrain phase, atmospheric delay phase, and noise phase, obtaining the surface deformation phase difference. Then, calculate the InSAR area deformation rate data based on the radar wavelength and incident angle. Perform carrier phase differential processing (RTK) on the GNSS data to obtain the GNSS point displacement rate data. Based on the acquired InSAR area deformation rate data and GNSS point displacement rate data, perform spatiotemporal fusion using the Kalman filter algorithm. Use the InSAR data as the prior value for system state prediction and the GNSS data as the observation value for updating, outputting the fused surface deformation data.

[0105] Step 3: Based on the acquired preliminary data of MEMS deep deformation, reference data of fiber optic grating deep deformation, and temperature data, establish a temperature drift compensation model; use the high stability of fiber optic grating data to calibrate the temperature drift coefficient of MEMS, perform real-time compensation on the preliminary data of MEMS deep deformation, and output the compensated deep deformation data.

[0106] Step 4: Integrate the fused surface deformation data, compensated deep deformation data, pore water pressure values, and temperature data into multi-source synchronous time-series data according to time correlation, and store them in the database with a unified timestamp;

[0107] Step 5: Input the multi-source synchronous time series data of a preset time window length into the trained time series prediction model to predict the multi-source synchronous time series data for future time periods; the time series prediction model is an improved LSTM intelligent early warning model, which is a long short-term memory neural network based on an attention mechanism. Its input is multi-source synchronous time series data, and its output is multi-source synchronous time series data for future time periods.

[0108] Step 6: Based on the multi-source synchronous time-series data for future time periods, obtain the fused surface deformation rate, compensated deep deformation rate, pore water pressure value, and temperature data; and calculate the landslide stability coefficient by combining the pre-stored test parameters.

[0109] Step 7: Based on the fused surface deformation rate, the compensated deep deformation rate, and the landslide stability coefficient, and the preset warning threshold, conduct graded early warning.

[0110] In step 2, InSAR technology acquires two SAR images of the same area at different times and uses the phase difference between the images to invert the core formula of surface deformation, as follows:

[0111]

[0112] in, The total phase difference, The phase difference is caused by surface deformation. The phase difference is caused by topographic relief. The phase difference is caused by atmospheric delay. Noise phase;

[0113] Deformation The calculation is performed using the following formula:

[0114]

[0115] in, For deformable variables, For SAR image wavelength, The radar incident angle is π, where π is the mathematical constant pi.

[0116] Based on the time series of deformation, InSAR regional deformation rate data are obtained.

[0117] In step 2, GNSS technology receives BeiDou + GPS satellite signals and uses carrier phase differential technology (RTK) to obtain the three-dimensional coordinates of the monitoring point. The change in coordinates at different times is the displacement value. Based on the time series of displacement values, the displacement rate data of the GNSS point is obtained.

[0118] In step 2, InSAR regional deformation rate data are used as prior values, and GNSS point displacement rate data are used as observed values. The time series of spatiotemporally continuous fused surface deformation data is output, and the fusion formula is as follows:

[0119] ;

[0120] ;

[0121] in, This is the fused deformation state vector. This is InSAR regional deformation rate data. For GNSS point displacement rate data, For Kalman gain, For the observation matrix, To predict the error covariance, To observe the noise covariance, the superscript T indicates transpose.

[0122] The Kalman filter fusion formula provides a specific calculation method for fusing InSAR and GNSS data, achieving optimal estimation by dynamically adjusting the weights through Kalman gain.

[0123] Based on the acquired preliminary data of MEMS deep deformation, reference data of fiber optic grating deep deformation, and temperature data, a temperature drift compensation model is established. The temperature drift coefficient of the MEMS is calibrated using the high stability of the fiber optic grating data, and real-time compensation is performed on the preliminary data of MEMS deep deformation. The temperature drift compensation model is as follows:

[0124]

[0125] in, The data is the compensated deep deformation data. Preliminary data on deep deformation of MEMS. The temperature drift coefficient is obtained by calibration using the high stability of fiber Bragg grating data. This represents the change in temperature.

[0126] The core logic of "calibrating the temperature drift coefficient k of MEMS using the high stability of fiber Bragg grating data" is as follows: using the stable period fiber Bragg grating data as a "reference", combined with temperature changes, the temperature drift coefficient of MEMS is quickly calculated. The specific steps include the following 4 steps:

[0127] (1) Selecting data

[0128] During the stable period of the slope, take the original MEMS data, the original fiber Bragg grating data, and the temperature data at the corresponding depth, and delete obvious outliers (such as MEMS mutation data and fiber Bragg grating demodulation failure data).

[0129] (2) Self-correction of fiber optic grating temperature drift:

[0130] Calculate the temperature drift of the fiber Bragg grating:

[0131]

[0132] in, The coefficient of thermal expansion of the fiber Bragg grating (typical value ≈ 5.5 × 10⁻⁶). -6 / ℃),

[0133] This refers to the length of the fiber Bragg grating sensing section (typical value = 100mm).

[0134] For actual measured temperature, This is the fiber grating temperature drift (this value is approximately 0 during the stable period).

[0135] Corrected fiber Bragg grating data:

[0136]

[0137] in, For the corrected fiber Bragg grating data,

[0138] These are the original values ​​for the fiber Bragg grating;

[0139] (3) Calculate the temperature drift of MEMS :

[0140]

[0141] Record the temperature change value :

[0142]

[0143] Among them, 25℃ is the reference temperature. This represents the temperature change value. This is raw MEMS data. This refers to the temperature drift of MEMS.

[0144] (4) Fitting temperature drift coefficient k:

[0145] Constructing a calibration dataset: using temperature change values MEMS temperature drift is the independent variable. Using the dependent variable, a sample set is formed, with a total sample size of ≥30, to ensure the reliability of subsequent fitting.

[0146] Perform linear regression using Excel or a simple fitting tool to obtain the formula: k is the temperature drift coefficient to be calibrated, and b is the offset. The goodness of fit is required to be greater than or equal to 0.9. If the error does not meet the requirements, piecewise linear fitting optimization is adopted, and the temperature drift coefficient k value is calibrated separately according to different temperature ranges to improve the calibration accuracy.

[0147] Methods for calculating landslide stability coefficients include:

[0148] The Morganston-Price method from the limit equilibrium method is used to determine this, as shown in the following formula:

[0149]

[0150] in, This is the landslide stability coefficient.

[0151] For sandstone cohesion, The segment length of the sliding surface,

[0152] For segmented soil and rock gravity, The inclination angle of the sliding surface segment,

[0153] The internal friction angle of the rock and soil. This refers to the pore water pressure.

[0154] The temperature drift compensation model uses a simple and efficient linear formula to correct for the effects of temperature. The Morgensdon-Price method, a classic limit equilibrium method, calculates the stability coefficient in real time using monitoring data, providing a solid theoretical foundation for stability assessment. These formulas together constitute the core algorithm of the intelligent monitoring and early warning subsystem, ensuring the scientific rigor and accuracy of data processing and early warning judgments.

[0155] In step 3, MEMS technology employs a MEMS inclinometer integrating a triaxial accelerometer and a gyroscope. The accelerometer measures the gravitational field component to obtain the tilt angle, while the gyroscope compensates for dynamic interference. Through integration calculations, the angular displacement and axial deformation of the deep soil and rock mass are obtained. The core formula is as follows:

[0156]

[0157]

[0158] Where d is the axial deformation. These are measurements from a triaxial accelerometer. The distance between adjacent MEMS measurement points. This represents the change in tilt angle.

[0159] In step 5, the time series prediction model is an improved LSTM intelligent early warning model, which is a long short-term memory neural network based on an attention mechanism. Its input is multi-source synchronous time series data, and its output is multi-source synchronous time series data for future time periods.

[0160] An improved LSTM intelligent early warning model is constructed: an attention-based LSTM network is built. The input layer receives multi-source synchronous time-series data from the past 7 days (a total of 144 time steps, with feature dimensions: surface deformation rate, deep deformation rate, pore pressure, rainfall, and temperature). The model includes two LSTM layers (128 hidden units), an attention layer, and a fully connected layer, outputting predicted time-series data for the next 7 days (in this embodiment, to improve accuracy, data from the previous 1-3 days is retained as the basis for judgment).

[0161] The training set comes from the database constructed in step 4. The fused surface deformation data, compensated deep deformation data, pore water pressure values ​​and temperature data are unified with a time resolution, and the data are cleaned and normalized. The data are then stitched together according to time to obtain multi-source synchronous time series data. The loss function is mean squared error, the optimizer is Adam, the learning rate is 0.001, and the prediction error of the model after training is ≤8%.

[0162] Combining multidimensional time-series features for prediction results in a more perceptive model and more accurate and efficient predictions.

[0163] Preferably, the improved LSTM intelligent early warning model adopts an encoder-decoder structure, takes multi-source synchronous time-series data as input, enhances the expression of key features through an attention mechanism, and finally outputs predicted time-series data for future periods, mainly including the following levels:

[0164] Input layer: Receives multi-source synchronized timing data with a preset time window length;

[0165] Multi-source data feature fusion layer: Parallel LSTM branches process different features;

[0166] Attention mechanism layer: Adaptively learns the importance weights of each time step and each feature;

[0167] LSTM coding layer: extracts deep temporal features;

[0168] LSTM decoding layer: generates future time series predictions;

[0169] Output layer: Outputs multi-source synchronized time-series data for future time periods.

[0170] Based on the model's predicted future data, the surface deformation rate, compensated deep deformation rate, pore water pressure, and temperature data are extracted and decomposed; combined with pre-stored geotechnical parameters (sandstone cohesion)... internal friction angle Porosity Permeability coefficient (etc.), using the Morganston-Price method to calculate the future landslide stability coefficient in real time.

[0171] Note that the multi-source synchronous time series data input to the improved LSTM intelligent early warning model usually contains noise, missing values, or outliers, and data cleaning and normalization are required to form a high-quality time series.

[0172] Data cleaning: Use interpolation methods (such as linear interpolation and spline interpolation) to fill in missing data;

[0173] Normalization: To avoid the influence of different units on model training, each monitored variable is normalized. Common methods include Min-Max scaling to the [0,1] interval or Z-score standardization.

[0174] After preprocessing, the time series of multi-source synchronized time series data is obtained.

[0175] By inputting the time series data of multi-source synchronous time series into the time series prediction model constructed above, future multi-source synchronous time series data can be obtained.

[0176] After predicting the future multi-source synchronous time series data in step 6, an inverse transformation needs to be performed based on the normalization parameters saved during normalization to restore the predicted values ​​to the original physical dimensions, thereby obtaining the fused surface deformation data, compensated deep deformation data, pore water pressure values, and temperature data.

[0177] In step 7, the preset warning thresholds include:

[0178] Attention level: The surface deformation rate after integration is 0.3-0.5 mm / day, or the deep deformation rate after compensation is 0.1-0.3 mm / day, or the landslide stability coefficient is 1.15-1.25;

[0179] Preparation level: The surface deformation rate after fusion is 0.5-1 mm / day, or the deep deformation rate after compensation is 0.3-0.5 mm / day, or the landslide stability coefficient is 1.05-1.15;

[0180] Action level: The surface deformation rate after fusion is greater than 1 mm / day, or the deep deformation rate after compensation is greater than 0.5 mm / day, or the landslide stability coefficient is less than 1.05.

[0181] By setting three levels of early warning thresholds, refined hierarchical management of landslide risk has been achieved. The "Attention" level alerts for attention, the "Preparation" level alerts for preparation, and the "Action" level alerts for immediate action. Different levels correspond to different emergency response measures, avoiding false alarms or missed alarms caused by a single threshold. The threshold range comprehensively considers three dimensions of indicators: surface displacement, deep deformation, and stability coefficient, forming a multi-parameter joint judgment mechanism. This significantly improves the accuracy and reliability of early warnings, providing a scientific basis for disaster prevention and mitigation.

[0182] In addition, to prevent inaccurate or insufficient real-time predictions, and to improve feedback speed and prevent prediction model failure, early warnings can be issued based on real-time InSAR regional deformation rate data, GNSS point displacement rate data, compensated deep deformation rate, and stability coefficient. These data are then compared with preset early warning thresholds to obtain the corresponding real-time early warning level. Based on landslide soil and rock parameters and engineering realities, historical data is used to invert and adjust the weights of each early warning indicator, setting three levels of early warning thresholds as shown in Table 1.

[0183] Table 1. Classified Real-time Early Warning Level Table

[0184]

[0185] Simulate different levels of early warning scenarios (such as simulating action-level early warnings by adjusting sensor output values) to test whether early warning information is accurately pushed (mobile APP, SMS, on-site audible and visual alarms), with an early warning response time of ≤15 minutes.

[0186] This invention constructs a rapid response channel for sudden events by introducing compensated deep deformation rate and stability coefficient as real-time early warning criteria. The stability coefficient, as a comprehensive indicator for evaluating slope stability, directly reflects the current safety reserve; the deep deformation rate can detect the initiation of deep slip zones immediately. Once these real-time indicators exceed safety thresholds, the system can immediately trigger an alarm, compensating for the potential lag in time-series prediction when dealing with sudden and transient scenarios, and achieving full-scenario coverage from "trend prediction" to "transient response".

[0187] The accuracy of early warnings is further improved through a multi-parameter joint early warning mechanism based on real-time parameters. Different monitoring technologies (InSAR, GNSS, MEMS) reflect the slope deformation state from different dimensions, and joint judgment can effectively avoid false alarms that may occur with a single technology (such as InSAR being affected by atmospheric interference, GNSS being blocked, etc.). The stability coefficient, as a comprehensive indicator, corroborates the deformation rate, forming a double guarantee. The setting of three-level thresholds, combined with engineering practice, provides clear and operable decision-making basis for on-site management personnel.

[0188] The dual-modal redundancy design of this invention is not simply redundant backup, but rather a deep fusion and mutual verification of two early warning logics. Specifically, the system can dynamically compare the future trend data predicted by the LSTM model with the real-time measured data. For example, when the predicted data indicates an increased risk, but the real-time measured data remains within a stable range, the system can classify it as a potential risk and continuously track it; conversely, if the real-time measured data (such as deep deformation rate or stability coefficient) shows an abnormal sudden change, and the prediction model has not yet captured the change due to learning cycle limitations, the real-time early warning channel can respond immediately. This cross-verification of "predicted trend" and "real-time situation" effectively avoids the false alarms and missed alarms caused by single-dimensional judgment, making the early warning results more scientific and reliable.

[0189] Meanwhile, key data collected by the real-time early warning channel (especially abnormal events that triggered real-time alarms but were not predicted by the model) can be automatically labeled and used as high-quality training samples, feeding back into the iterative optimization process of the LSTM model. This enables the prediction model to continuously learn new and rare instability patterns, gradually expanding its knowledge boundaries, thereby achieving continuous evolution and self-improvement of its predictive capabilities.

[0190] Example 2:

[0191] like Figure 2 As shown in the figure, this embodiment provides a precision prevention and ecological collaborative restoration system, including a solid waste-based low-carbon grouting reinforcement subsystem, an ecological slope protection subsystem, and an intelligent monitoring and early warning subsystem.

[0192] (I) Solid waste-based low-carbon grouting reinforcement subsystem:

[0193] The solid waste-based low-carbon grouting reinforcement subsystem includes grouting conduits and solid waste-based low-carbon grouting material. The grouting conduits are made of seamless steel pipes with a diameter of 80mm, arranged in a staggered pattern within the slope body at 3m intervals and a 15° inclination angle, with a drilling depth of 11m, ensuring penetration to 2m below the potential sliding surface. Grouting holes are pre-drilled in the conduit walls, and the bottom of the conduit is sealed. The solid waste-based low-carbon grouting material is injected into the slope body through the grouting conduits, diffusing and solidifying under grouting pressure to form a solid waste-based grouting reinforcement zone. This zone covers both the upper and lower parts of the potential sliding surface, significantly improving the shear strength and overall stability of the slip zone soil.

[0194] (II) Ecological Slope Protection Subsystem:

[0195] The ecological slope protection subsystem includes a grid-type ecological concrete frame, foundation anchors, vegetation system, and drainage system.

[0196] A grid-type ecological concrete frame is laid on the slope surface and is precast or cast-in-place using low-carbon ecological concrete based on solid waste, which is the same material used for grouting. The grid is rectangular with a grid size of 2m × 2m. The cross-sectional dimensions of the horizontal and vertical beams are 300mm × 400mm. Reinforcing ribs are provided at the nodes to improve the overall integrity of the frame, and monitoring holes with a diameter of 80-100mm are reserved for the installation of monitoring equipment or for vegetation growth channels.

[0197] The foundation anchors are made of 32mm diameter threaded steel and are embedded in the stable soil layer of the slope. The pull-out force of a single anchor is ≥80kN. The foundation anchors are arranged one-to-one with the nodes of the lattice-type ecological concrete frame, firmly anchoring the frame to the slope and forming a reliable force transmission path.

[0198] The vegetation system consists of native grass and shrub composite plants planted within a lattice-structured ecological concrete frame. The shrub roots, along with the foundation anchors, penetrate into the soil, forming a soil-stabilizing coupling mechanism with the grouting reinforcement area of ​​the solid waste base; the herbaceous plant roots interweave between the surface soil of the slope and the lattice-structured ecological concrete frame, forming a shallow reinforcement network.

[0199] The interception and drainage system includes a slope-top interception ditch, an internal drainage channel within the grid structure, and longitudinal drainage ditches. The interception ditch has a cross-sectional dimension of 300mm × 300mm and is constructed using porous ecological concrete. The internal drainage channel within the grid structure connects the interception ditch and the longitudinal drainage ditches. A longitudinal drainage ditch is installed every 10m, with a built-in 50mm diameter permeable hose to effectively guide slope runoff and reduce the adverse effects of rainwater infiltration on slope stability.

[0200] (III) Intelligent Monitoring and Early Warning Subsystem:

[0201] The intelligent monitoring and early warning subsystem includes a surface displacement monitoring module, a deep deformation monitoring module, a pore water pressure monitoring module, a micro-meteorological monitoring module, a data transmission system, and a cloud platform server.

[0202] The surface displacement monitoring module includes InSAR and GNSS monitoring stations. InSAR acquires satellite SAR images covering the landslide area and a surrounding 1-2 km buffer zone to achieve wide-area surface deformation monitoring; GNSS monitoring stations are deployed at key locations and in the anomalous areas of the InSAR monitoring units, using BeiDou + GPS dual-mode reception and carrier phase differential technology to acquire high-precision point displacement data.

[0203] The deep deformation monitoring module includes a MEMS inclinometer and a fiber Bragg grating inclinometer, which are interleaved and arranged in monitoring holes within the landslide body. The MEMS inclinometer integrates a triaxial accelerometer and a gyroscope to acquire raw acceleration and angular velocity data; the fiber Bragg grating inclinometer acquires raw wavelength data. Working together, the MEMS inclinometer provides highly sensitive continuous deformation monitoring, while the fiber Bragg grating inclinometer provides highly stable reference data, jointly and accurately locating the position and deformation characteristics of the potential sliding surface.

[0204] The pore water pressure monitoring module includes a vibrating wire pore water pressure gauge embedded in the slope to collect raw frequency data of pore water pressure in real time. The data is then converted to obtain pore water pressure values, providing key parameters for calculating the landslide stability coefficient.

[0205] The micro-meteorological monitoring module is deployed at the top of the slope to acquire meteorological data such as temperature, humidity, and rainfall in the landslide area. The temperature data is used for temperature drift compensation in deep deformation monitoring.

[0206] The data transmission system adopts LoRa+4G / 5G dual-mode communication. Each monitoring device on site is connected to the wireless data acquisition and transmission terminal via an RS485 interface. The terminal aggregates data through a LoRa self-organizing network, and then uploads the multi-source monitoring data to the cloud platform server in real time via the 4G / 5G network, ensuring the stability of data transmission in complex mountainous environments.

[0207] The cloud platform server, as the core processing unit of the intelligent monitoring and early warning subsystem, receives and processes the data uploaded by each monitoring module, executes the landslide prediction method as described in Example 1, and performs multi-source data fusion, intelligent early warning model inference, and hierarchical early warning output.

[0208] Specifically, in this embodiment, the solid waste-based low-carbon grouting reinforcement subsystem, the ecological slope protection subsystem, and the intelligent monitoring and early warning subsystem achieve integrated landslide management through structural coupling and functional synergy.

[0209] The solid waste-based low-carbon grouting reinforcement subsystem includes grouting conduits and solid waste-based low-carbon grouting materials. The grouting conduits are laid within the landslide body and penetrate 1-2m into the potential sliding surface. The solid waste-based low-carbon grouting materials, by weight, include 40-60 parts of steel slag powder, 20-30 parts of fly ash, 10-20 parts of slag powder, 5-10 parts of alkali activator, 0.5-1.5 parts of thickener, and 20-30 parts of water. The alkali activator is a composite solution of water glass with a modulus of 1.2-1.5 and NaOH (sodium hydroxide) solution with a concentration of 8-12 mol / L mixed in a mass ratio of 3:1. The thickener is hydroxypropyl methylcellulose.

[0210] The ecological slope protection subsystem includes a lattice-type ecological concrete frame, foundation anchors, a vegetation system, and a drainage system. The lattice-type ecological concrete frame is made of solid waste-based low-carbon ecological concrete of the same origin as the grouting material. The lattice is rectangular with a grid size of 2m × 2m, and the cross-sectional dimensions of the horizontal and vertical beams are 300mm × 400mm. The 28-day compressive strength is ≥20MPa, the flexural strength is ≥3MPa, and the porosity is 15-20%. Reinforcing ribs are added to the nodes, and monitoring holes with a diameter of 80-100mm are reserved. The anchors are 32mm diameter threaded steel bars, implanted into the stable soil layer of the slope, with a single anchor pull-out force ≥80kN, and are arranged one-to-one with the grid nodes; the vegetation system is a composite of native grasses and shrubs; the drainage system includes intercepting ditches, drainage channels within the grid, and longitudinal drainage ditches; the drainage channels within the grid are connected to the intercepting ditches and longitudinal drainage ditches respectively; the intercepting ditches have a cross-sectional size of 300mm×300mm, are poured with porous ecological concrete, and a longitudinal drainage ditch is set every 10m, with a 50mm diameter permeable hose inside.

[0211] The surface displacement monitoring module is a monitoring system that integrates InSAR and GNSS technologies. InSAR covers the landslide area and a 1-2 km buffer zone around it, while GNSS is deployed at key locations and in the InSAR anomaly zone. The two are fused using a Kalman filter algorithm. It is used to collect InSAR and GNSS displacement information in the landslide area and fuse them to generate a time series of surface deformation rates.

[0212] The deep deformation monitoring module is a fusion monitoring system of MEMS and fiber optic inclinometer. The MEMS inclinometer and the fiber optic inclinometer are arranged alternately, and the two work together to locate the sliding surface.

[0213] The pore water pressure monitoring module uses a vibrating wire pore water pressure gauge;

[0214] The micro-meteorological monitoring module is used to obtain the temperature in the landslide area;

[0215] The data transmission system adopts LoRa+4G / 5G dual-mode communication;

[0216] The cloud platform server has a built-in improved LSTM intelligent early warning model with a multi-source data feature fusion layer.

[0217] A three-level early warning threshold is set based on multiple indicators.

[0218] Solid waste-based low-carbon grouting materials have a 28-day compressive strength ≥ 30 MPa and a permeability coefficient ≤ 1 × 10⁻⁶. -7 The shrinkage rate is ≤0.5%, and the carbon emission is ≤0.1t (tons) of carbon dioxide / t; the frost resistance strength loss rate of the lattice-structured ecological concrete is ≤15%, and the impermeability grade is ≥P6; the permeability coefficient of the ecological filling matrix is ​​≥5×10 -3 cm / s. The grouting conduit uses seamless steel pipe with a diameter of 80mm. The conduit is laid out in a quincunx pattern with a spacing of 3m and an inclination angle of 10-20°. The layout parameters are determined based on the limit equilibrium method, specifically the Morganston-Price method within the limit equilibrium method.

[0219] like Figure 3 As shown in the figure, this embodiment also provides a method for installing and using a precision prevention and control and ecological synergistic restoration system, including the following steps:

[0220] Step 1: Preliminary investigation and scheme design: Using UAVs equipped with LiDAR sensors, geological boreholes, and ground-penetrating radar combined with InSAR for preliminary survey, three-dimensional spatial information of the target area is collected to obtain landslide parameters and risk zones. Based on the landslide parameters and risk zones, grouting, slope protection and monitoring schemes are generated and optimized.

[0221] Step 2: Preparation of solid waste-based low-carbon grouting material: Pre-treat industrial solid waste materials such as steel slag powder, fly ash, and slag powder, prepare alkali activator, mix according to the ratio to prepare slurry, and prepare ecological concrete of the same origin as the grouting material, and prepare ecological filling matrix;

[0222] Step 3: Construction of the solid waste-based low-carbon grouting reinforcement subsystem: Drill holes using a geological drilling rig, insert and seal the grouting conduit, use segmented retreating grouting, pressure 0.5-1.2MPa, cure for 7 days after grouting and test the quality;

[0223] Step 4: Construction of the ecological slope protection subsystem: slope pretreatment, construction of the grid-type ecological concrete frame and foundation anchors, ecological matrix filling and vegetation planting, and construction of the interception and drainage system;

[0224] Step 5: Deployment and debugging of intelligent monitoring and early warning subsystem: Deploy monitoring equipment such as InSAR, GNSS, and MEMS, fuse InSAR and GNSS data through Kalman filtering algorithm, establish temperature drift compensation model to form MEMS-fiber grating complementary mechanism, and set three-level early warning thresholds based on LoRa+4G / 5G data transmission system.

[0225] Step 6: System Operation, Maintenance and Effect Evaluation: Routine maintenance of sensors and slope protection structures; regular calibration of sensors and evaluation of system mechanical stability, ecological effects, low-carbon effects and economic effects; optimization of system parameters.

[0226] In step 4, the plant seeds are native grass and shrub species adapted to the local climate. A combination of grass and shrubs is adopted. The herbaceous plants are bermudagrass and ryegrass, and the shrubs are Amorpha fruticosa and Lespedeza bicolor. The lattice frame and the deep grouting material are of the same origin and provide channels for the growth of vegetation roots. The shrub roots and foundation anchors drill into the soil and form a soil-fixing coupling mechanism with the grouting reinforcement. The herbaceous plant roots penetrate the slope soil layer and form a reinforcement network with the lattice concrete. The volume ratio of the mixed matrix is ​​nutrient soil: organic fertilizer: water-retaining agent: seeds = 85:10:4.99:0.01. The nutrient soil is composed of 70% topsoil, 20% river sand, and 10% decomposed straw.

[0227] The data transmission system debugging requires ensuring that the LoRa gateway signal strength is ≥-80dBm, the data upload rate is ≥100kbps, and there is no disconnection for 24 consecutive hours; the intelligent early warning model debugging requires collecting on-site data for 3-7 days to ensure that the deviation between the model prediction results and the actual data is ≤10%.

[0228] Step 6 includes mechanical stability assessment (LiDAR analysis of surface displacement by UAV, annual displacement ≤50mm) and deep deformation monitoring (no obvious slippage), ecological effect assessment (vegetation coverage ≥80%), species diversity (≥5 species) and soil organic matter content (≥2%), low-carbon effect assessment (carbon emission reduction = traditional cement consumption × 0.9 - carbon emissions of solid waste-based materials), and economic effect assessment (cost savings rate ≥15% compared with traditional schemes).

[0229] The specific steps of the system installation method in this embodiment are as follows:

[0230] (1) First, conduct preliminary surveys and design the plan:

[0231] Historical SAR images of the landslide area were acquired and processed using SBAS-InSAR technology to obtain historical deformation fields, identify potential landslide areas and deformation rates, and a digital elevation model (DEM) was acquired using UAV LiDAR scanning to identify landslide boundaries. Geological boreholes were deployed, and core samples were collected to obtain sandstone cohesion. internal friction angle Porosity Permeability coefficient The Morganston-Price method in the limit equilibrium method was used to determine the stability coefficient of the landslide after grouting reinforcement. .

[0232] The stability coefficient after grouting reinforcement must be ensured. For a depth of ≥1.25, the grouting pipes are designed in a quincunx pattern with a spacing of 3m, a depth of 11m (penetrating the sliding surface 1-2m), and an inclination angle of 15°.

[0233] The lattice-type ecological concrete frame is made of solid waste-based low-carbon ecological concrete of the same origin as the grouting material. The lattice is rectangular with a grid size of 2m×2m. The cross-sectional dimensions of the horizontal and vertical beams are 300mm×400mm. The 28-day compressive strength is ≥20MPa, the flexural strength is ≥3MPa, and the porosity is 15-20%. Reinforcing ribs are added to the nodes, and monitoring holes with a diameter of 80-100mm are reserved. The foundation anchors are 32mm diameter threaded steel bars, 3m in length, implanted into the stable soil layer of the slope. The pull-out force of a single anchor is ≥80kN, and they are arranged one-to-one with the lattice nodes. The vegetation system consists of native grass and shrub species adapted to the local climate, using a combination of grass and shrubs. The herbaceous plants are bermudagrass and ryegrass, and the shrubs are purple acacia and lespedeza. The interception and drainage system is designed with a 300mm×300mm interception ditch at the top of the slope, a 100mm×150mm drainage channel inside the lattice, and a longitudinal drainage ditch every 12m, connecting to the collection well at the foot of the slope.

[0234] Based on the preliminary InSAR survey results, and according to historical landslide experience values, high, medium, and low risk zones were divided. In high-risk zones, the density of GNSS monitoring points was increased to one per 200 square meters; in medium-risk zones, one per 300 square meters; and in low-risk zones, one per 500 square meters. In high-risk zones, the spacing between monitoring holes was increased to 15-20 meters, the spacing between MEMS inclinometer measuring points was 0.8-1 meters, and the spacing between fiber optic inclinometer measuring points was 1.5-2 meters, ensuring that no landslide surface was missed during monitoring.

[0235] (2) Preparation of solid waste-based low-carbon grouting materials:

[0236] Raw materials: Steel slag powder (specific surface area 480m²) 2 50 parts ( / kg) of Grade I fly ash, 25 parts of S95 slag powder, 20 parts of alkaline activator (water glass modulus 1.3 + NaOH concentration 10mol / L, mass ratio 3:1) 8 parts, 1 part of HPMC (hydroxypropyl methylcellulose), and 25 parts of water.

[0237] Slurry preparation: dry mixing for 3 min → adding activator aqueous solution and stirring for 4 min → adding HPMC and stirring for 2 min. The viscosity and fluidity of the slurry are tested and found to meet the target requirements.

[0238] (3) Construction of the solid waste-based low-carbon grouting reinforcement subsystem:

[0239] Drilling: The XY-1 drilling rig was used to drill holes at the designed locations with a diameter of 90mm and a depth of 11m. The depth of the sliding surface was verified to be 8-9m, which was consistent with the exploration results. Grouting pipe installation: Seamless steel pipes with a diameter of 80mm were used and inserted into the drilled holes and then sealed with quick-setting cement mortar.

[0240] Grouting: Segmented retreating grouting, each segment is 1.5m long, grouting pressure is 0.8-1.0MPa, and the grouting volume Q per hole is calculated according to the following formula:

[0241] ;

[0242] in, Let π be the radius of slurry diffusion, and π be pi. The length of the grouting section. Porosity of soil and rock mass The fill factor is 1.1-1.3;

[0243] Quality inspection: The permeability coefficient obtained from the water pressure test is 8×10⁻⁶. -8 cm / s, the compressive strength of the core sample after 28 days is 32MPa, which is qualified.

[0244] (4) Construction of the ecological slope protection subsystem:

[0245] Construction of the lattice frame: The foundation anchors are made of 28mm diameter threaded steel, 2.5m in length, and are inserted at an angle perpendicular to the slope, with a spacing of 2m (corresponding to the lattice nodes). The average pull-out force is 86kN.

[0246] Solid waste-based low-carbon ecological concrete mix proportion: 35% steel slag powder, 25% fly ash, 20% slag powder, 15% engineering waste aggregate, 5% alkali activator, water-cement ratio 0.35, 28-day compressive strength 23MPa, porosity 18%; verticality error of formwork installation 2‰, moist curing after pouring for 16 days, no cracks or honeycomb defects on the frame surface.

[0247] Ecological filling and vegetation planting: Filling substrate ratio (volume ratio): 72% improved nutrient soil + 12% solid waste biochar + 4% water-retaining agent + 2% slow-release fertilizer + 0.015% mixed seeds; layered filling and compaction, filling height is 25mm lower than the top surface of the grid, hydroseeding seeding rate is 25g / m², transplanting Amorpha fruticosa seedlings at nodes, survival rate is 92%;

[0248] Construction of the interception and drainage system: The interception ditch at the top of the slope is constructed using solid waste-based porous concrete with a porosity of 22% and a permeability coefficient of 8×10⁻⁶. -3 cm / s; The drainage channel inside the grid has a built-in permeable hose with a diameter of 50mm, is filled with 10-20mm gravel, and the surface is covered with geotextile.

[0249] (5) Deployment and debugging of intelligent monitoring and early warning subsystem:

[0250] Surface displacement monitoring: Acquire InSAR data and interface with satellite data providers to subscribe to Sentinel-1A / B satellite data; GNSS monitoring station installation: Install GNSS monitoring stations at the designed locations with antenna height of 1.5m, unobstructed surroundings, and base station deployment distance of 5-10km from the monitoring station. Use static measurement method for initial calibration and observation duration ≥24h.

[0251] GNSS technology: Global Navigation Satellite System (GNSS) is a space-based radio navigation and positioning system that can provide users with all-weather 3D coordinates, velocity, and time information at any location on the Earth's surface or in near-Earth space. It is a general term for satellite navigation systems that can achieve global coverage.

[0252] InSAR technology: Interferometric radar refers to Synthetic Aperture Radar (InSAR), a newly developed space-based Earth observation technology that combines traditional SAR remote sensing technology with radio astronomy interferometry. It utilizes radar to transmit microwaves towards a target area and then receives the echoes reflected from the target, obtaining SAR complex image pairs of the same target area. If there is coherence between the complex image pairs, conjugate multiplication of the SAR complex image pairs yields an interferogram. Based on the phase value of the interferogram, the path difference of the microwaves in the two images can be obtained, thus calculating the topography, landforms, and minute surface changes of the target area. It can be used for digital elevation modeling, crustal deformation detection, etc.

[0253] Deep deformation monitoring: Install MEMS inclinometers, embed MEMS inclinometer strings into the monitoring borehole, and connect the sensors via flexible cables. The bottom is fixed to a stable rock layer (penetrating the sliding surface to a depth of 2m). The space between the borehole wall and the sensor is filled with a mixture of fine sand and bentonite (ratio 3:1) to reduce temperature changes and vibration interference. Fiber optic inclinometers and MEMS inclinometer strings are arranged alternately, and armored optical fibers are used to avoid damage during construction. The grating demodulator is placed next to the on-site data acquisition terminal, and its demodulation accuracy is ≤1pm.

[0254] InSAR data transmission: Satellite imagery data is automatically downloaded via the cloud platform API interface. The download cycle is set, and the test data download rate is ≥10Mbps. Preliminary data transmission of deep deformation in MEMS: The MEMS inclinometer is connected to the LoRa gateway via the RS485 interface. The sampling frequency (100Hz) and transmission frequency (1 time / 10min, data is compressed before transmission) are configured, and the communication stability is tested.

[0255] (6) System operation, maintenance, and performance evaluation:

[0256] Routine operation and maintenance:

[0257] InSAR data processing: Automatically process newly acquired satellite imagery, update regional deformation fields, compare and analyze consistency with GNSS data, and promptly calibrate any discrepancies; MEMS inclinometer maintenance: Every 3 months, read the temperature compensation coefficient of the MEMS inclinometer through the field data acquisition terminal and update the temperature drift compensation model; Check sensor connection cables every 6 months to prevent aging and breakage; Multi-source data consistency verification: Conduct GNSS static measurements (observation duration ≥ 8 hours) monthly to calibrate InSAR deformation data; Compare the sliding surface positioning results of MEMS and fiber optic inclinometers quarterly to ensure data reliability.

[0258] System effectiveness evaluation:

[0259] The mechanical stability assessment involves scanning the slope surface every six months using a LiDAR drone and comparing and analyzing changes in surface displacement; the displacement of the deep sliding surface is calculated using inclinometer data; the ecological effect assessment involves measuring vegetation coverage (≥80%), species diversity (≥4 native plant species), and soil organic matter content (≥2%, measured using the potassium dichromate oxidation-external heating method) quarterly; the low-carbon effect assessment involves statistically analyzing the amount of solid waste used in the project (steel slag, fly ash, slag, and construction waste) and calculating carbon emission reduction (carbon emission reduction = traditional cement usage × 0.9 - carbon emissions from solid waste-based materials, where the carbon emission coefficient for traditional cement is 0.9 tons of carbon dioxide / ton, and the carbon emission coefficient for solid waste-based materials is 0.1 tons of carbon dioxide / ton); the economic effect assessment involves comparing the project with traditional treatment schemes (anti-slide piles + ordinary grouting + separate ecological restoration), calculating the project cost of this system (material cost, construction cost, and maintenance cost), and evaluating the cost saving rate (≥15%).

[0260] This system can achieve:

[0261] (1) Precise monitoring and efficient early warning: By integrating multi-source sensing technologies such as InSAR, GNSS, MEMS, and fiber optic inclinometer, the deformation, hydrological and meteorological data of the slope body’s “macro-surface-micro-positioning-deep soil and rock” are collected in real time. Combined with machine learning models, the landslide risk is assessed and warned in a refined manner, so as to buy enough time for disaster response.

[0262] (2) Structural stability and ecological synergy: In terms of mechanical performance, the grouting reinforcement + grid-type slope protection + multi-type plant root system coupling mechanism forms a three-dimensional slope reinforcement system with a high landslide stability coefficient. The grid-type frame transfers the load to the stable stratum, improving the overall anti-overturning and anti-sliding ability. In terms of ecological effect, solid waste-based biochar improves the soil's water and fertilizer retention capacity and adsorbs heavy metals. Plant roots penetrate the filling matrix and slope soil layer to form a "biological reinforcement network". It improves the ecological landscape effect of the slope and improves the regional microclimate.

[0263] (3) Low-carbon environmental protection and resource recycling: Use solid waste-based materials such as steel slag, fly ash, and engineering waste to improve the utilization rate of waste resources.

[0264] (4) Strong system compatibility: Homogeneous materials ensure coordinated structural deformation, and the pre-reserved monitoring holes of the lattice-type ecological concrete realize deep synergy of "monitoring-structure-ecology", which can be widely adapted to various landslide scenarios such as highways, railways, and mines.

[0265] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

[0266] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0267] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0268] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0269] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0270] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A landslide intelligent monitoring and early warning method, characterized in that, Includes the following steps: Acquire InSAR data, GNSS data, pore water pressure values, raw acceleration and angular velocity data from MEMS inclinometers, and temperature data; InSAR stands for Synthetic Aperture Radar Interferometry, and GNSS stands for Global Navigation Satellite System. Differential interferometry is performed on InSAR data to obtain deformation rate data of the InSAR region; Carrier phase differential processing is performed on GNSS data to obtain GNSS point displacement rate data; Based on the InSAR regional deformation rate data and GNSS point displacement rate data, the Kalman filter algorithm is used to perform spatiotemporal fusion and output the fused surface deformation data. Preliminary data on deep deformation of MEMS are obtained from the original acceleration and angular velocity data of MEMS inclinometer. Combined with temperature data and a pre-established temperature drift compensation model, the preliminary data on deep deformation of MEMS is compensated in real time to obtain the compensated deep deformation data; MEMS stands for micro-motor system. The fused surface deformation data, compensated deep deformation data, pore water pressure values ​​and temperature data are integrated into multi-source synchronous time-series data. Input multi-source synchronous time series data with a preset time window length into a trained time series prediction model to predict multi-source synchronous time series data for future time periods. Based on the multi-source synchronous time-series data of the future time period, the fused surface deformation rate, compensated deep deformation rate, pore water pressure value and temperature data are obtained, and the landslide stability coefficient is calculated in combination with the pre-stored test parameters. The early warning system is classified based on the fused surface deformation rate, the compensated deep deformation rate, and the landslide stability coefficient, along with the preset warning threshold.

2. The landslide intelligent monitoring and early warning method according to claim 1, characterized in that, Differential interferometry was performed on InSAR data to obtain InSAR region deformation rate data, including: By acquiring two SAR images of the same area at different times, the surface deformation is retrieved using the phase difference between the images, thus forming a time series of deformation: ； in, The total phase difference, The phase difference is caused by surface deformation. The phase difference is caused by topographic relief. The phase difference is caused by atmospheric delay. Noise phase; Deformation Calculate using the following formula: ； in, For deformable variables, For SAR image wavelength, π is the radar incident angle, and π is pi. Based on the time series of deformation, InSAR regional deformation rate data are obtained.

3. The landslide intelligent monitoring and early warning method according to claim 2, characterized in that, Based on the InSAR regional deformation rate data and GNSS point displacement rate data, a Kalman filter algorithm is used for spatiotemporal fusion to output fused surface deformation data, including: Using InSAR regional deformation rate data as prior values ​​and GNSS point displacement rate data as observed values, the time series of spatiotemporally continuous fused surface deformation data is output. The fusion formula is as follows: ； ； in, This is the fused deformation state vector. This is InSAR regional deformation rate data. For GNSS point displacement rate data, For Kalman gain, For the observation matrix, To predict the error covariance, To observe the noise covariance, the superscript T indicates transpose.

4. The landslide intelligent monitoring and early warning method according to claim 3, characterized in that, The temperature drift compensation model is as follows: ； in, The data is the compensated deep deformation data. Preliminary data on deep deformation of MEMS. This is the temperature drift coefficient. This represents the change in temperature.

5. The landslide intelligent monitoring and early warning method according to claim 4, characterized in that, Methods for calculating landslide stability coefficients include: The Morganston-Price method from the limit equilibrium method is used to determine this, as shown in the following formula: ； in, This is the landslide stability coefficient. For sandstone cohesion, The segment length of the sliding surface, For segmented soil and rock weight, The inclination angle of the sliding surface segment, The internal friction angle of the rock and soil. This refers to the pore water pressure.

6. The landslide intelligent monitoring and early warning method according to claim 2, characterized in that, The preset warning thresholds include: Attention level: The surface deformation rate after integration is 0.3-0.5 mm / day, or the deep deformation rate after compensation is 0.1-0.3 mm / day, or the landslide stability coefficient is 1.15-1.25; Preparation level: The surface deformation rate after fusion is 0.5-1 mm / day, or the deep deformation rate after compensation is 0.3-0.5 mm / day, or the landslide stability coefficient is 1.05-1.15; Action level: The surface deformation rate after fusion is greater than 1 mm / day, or the deep deformation rate after compensation is greater than 0.5 mm / day, or the landslide stability coefficient is less than 1.

05.

7. The landslide intelligent monitoring and early warning method according to claim 6, characterized in that, The method further includes: Real-time InSAR regional deformation rate data and GNSS point displacement rate data, compensated deep deformation rate and landslide stability coefficient are acquired and compared with preset real-time warning thresholds to obtain the corresponding real-time warning level. Attention level: InSAR area deformation rate data is 0.3-0.5 mm / day, or GNSS point displacement rate data is 0.5-1 mm / day, or compensated deep deformation rate is 0.1-0.3 mm / day, or stability coefficient meets 1.15-1.25; Preparation level: InSAR area deformation rate data is 0.5-1 mm / day, or GNSS point displacement rate data is 1-3 mm / day, or compensated deep deformation rate is 0.3-0.5 mm / day, or stability coefficient meets 1.05-1.15; Action level: InSAR area deformation rate data >1 mm / day, or GNSS point displacement rate data >3 mm / day, or compensated deep deformation rate >0.5 mm / day, or stability coefficient <1.

05.

8. A precision prevention and control system for ecological synergistic restoration, characterized in that, include: A solid waste-based low-carbon grouting reinforcement subsystem includes a grouting conduit and a solid waste-based low-carbon grouting material. The solid waste-based low-carbon grouting material is injected into the grouting conduit to form a deep reinforcement area. The grouting conduit is deployed within the landslide body and penetrates the potential sliding surface. An ecological slope protection subsystem includes a grid-type ecological concrete frame, foundation anchors, and a vegetation system. The grid-type ecological concrete frame is laid on the slope surface of the landslide body, the foundation anchors anchor the grid-type ecological concrete frame to the slope body of the landslide body, and the vegetation system is set inside the grid-type ecological concrete frame. as well as The intelligent monitoring and early warning subsystem includes: The surface displacement monitoring module is used to acquire InSAR and GNSS data of the landslide area; The deep deformation monitoring module is used to acquire the raw acceleration and angular velocity data of the MEMS inclinometer in the landslide area; A pore water pressure monitoring module is used to acquire pore water pressure values ​​in the landslide area. The micro-meteorological monitoring module is used to acquire temperature data in the landslide area; The data transmission system is used to connect the cloud platform server with the surface displacement monitoring module, the deep deformation monitoring module, the pore water pressure monitoring module, and the micro-meteorological monitoring module. The cloud platform server is used to perform landslide deformation prediction and early warning based on the data uploaded by the surface displacement monitoring module, deep deformation monitoring module, pore water pressure monitoring module and micro-meteorological monitoring module as described in claim 1.

9. The precision prevention and control and ecological synergistic restoration system according to claim 8, characterized in that, The solid waste-based low-carbon grouting material comprises, by weight, 40-60 parts steel slag powder, 20-30 parts fly ash, 10-20 parts slag powder, 5-10 parts alkali activator, 0.5-1.5 parts thickener, and 20-30 parts water. The alkali activator is a composite solution of water glass with a modulus of 1.2-1.5 and sodium hydroxide solution with a concentration of 8-12 mol / L mixed at a mass ratio of 3:1, and the thickener is hydroxypropyl methylcellulose.

10. The precision prevention and control and ecological synergistic restoration system according to claim 8, characterized in that, The lattice-type ecological concrete frame is made of solid waste-based low-carbon ecological concrete of the same origin as the solid waste-based low-carbon grouting material. Its lattice form is rectangular, and the nodes are provided with reinforcing ribs and reserved monitoring holes. The basic anchors are implanted into the stable soil layer of the slope, and the pull-out force of a single anchor is ≥80kN. They are arranged one-to-one with the grid nodes of the grid-type ecological concrete frame. The vegetation system is a grass-shrub composite plant, whose root system forms a solidified body with the foundation anchor and the solid waste-based low-carbon grouting material.

Images