Intelligent early warning method and device for landslide
By constructing a personalized two-dimensional discrete element model using Spencer's method and discrete element simulation technology, and combining it with long short-term memory networks and rigid body limit equilibrium method, the problem of traditional landslide early warning methods being unable to accurately capture the evolution law of slopes is solved, thus realizing personalized landslide risk prediction and dynamic early warning.
Patent Information
- Application Number
- CN202510853472.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-24
- Publication Date
- 2025-11-04
AI Technical Summary
Traditional landslide early warning methods are unable to accurately capture the unique evolution patterns of different slopes, making it impossible to achieve accurate prediction and dynamic early warning. Furthermore, the uniform threshold is difficult to adapt to the mechanical properties of different slopes.
The Spencer method was used to screen potential slip surfaces, and a personalized two-dimensional discrete element model was constructed by combining discrete element simulation technology. Long short-term memory network was used to learn slope geological information, and the rigid body limit equilibrium method was combined for early warning judgment.
It achieves precise capture of the differences in geological structure and geotechnical parameters of different slopes, dynamically generates exclusive training data and prediction models, realizes personalized early warning, and improves the timeliness and accuracy of early warning.
Smart Images
Figure CN120894871A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of landslide early warning technology, and more specifically, to a landslide intelligent early warning method and device. Background Technology
[0002] In landslide early warning technology, traditional methods primarily rely on historical case data to train general neural network models. However, different slopes exhibit significant differences in geological structure, sliding surface geometry, and geotechnical parameters, making it difficult for traditional methods to accurately capture the unique evolutionary patterns of target slopes. Furthermore, existing technologies often employ fixed parameters for stability calculations, failing to effectively simulate the nonlinear decay process of these parameters over time, thus reducing the timeliness and accuracy of early warnings. Simultaneously, the difficulty in accurately adapting uniform thresholds to the mechanical properties of different slopes hinders the accurate prediction and dynamic early warning of landslide risks.
[0003] Therefore, there is an urgent need for a landslide intelligent early warning method and device to solve the problem of being unable to accurately predict landslide risks and achieve dynamic early warning. Summary of the Invention
[0004] The purpose of this invention is to provide a landslide intelligent early warning method and device to improve the above-mentioned problems. To achieve the above objective, the technical solution adopted by this invention is as follows:
[0005] Firstly, this application provides a landslide intelligent early warning method, including:
[0006] Obtain slope geological information, which includes slope topographic data, soil and rock parameters, and field monitoring data;
[0007] A vector graphic is drawn from the geological information of the slope. The slope stability is calculated and the slip surface is screened using the vector graphic and the Spencer method. A set of potential slip surfaces is output.
[0008] Based on the discrete element simulation technology, a two-dimensional discrete element model is constructed from the geological information of the slope. By cutting the two-dimensional discrete element model and introducing random parameter perturbation, multiple slope stability analysis models are generated.
[0009] Multiple slope stability analysis models are input into a long short-term memory network for training, and a slope prediction model is output. By inputting the slope geological information and the set of potential slip surfaces into the slope prediction model for prediction, slip surface characteristic parameters are obtained.
[0010] The landslide stability coefficient is calculated based on the rigid body limit equilibrium method using the slip surface characteristic parameters, and a pre-warning judgment is made on the landslide stability coefficient by using a preset dynamic threshold.
[0011] Secondly, this application also provides a landslide intelligent early warning device, comprising:
[0012] The acquisition module is used to acquire slope geological information, which includes slope topographic data, soil and rock parameters, and field monitoring data.
[0013] The drawing module is used to draw vector graphics on the slope geological information, perform slope stability calculation and slip surface screening using vector graphics and the Spencer method, and output a set of potential slip surfaces.
[0014] The module is used to construct a two-dimensional discrete element model based on the slope geological information using discrete element simulation technology. By cutting the two-dimensional discrete element model and introducing random parameter perturbations, multiple slope stability analysis models are generated.
[0015] The training module is used to input multiple slope stability analysis models into a long short-term memory network for training and output a slope prediction model. By inputting the slope geological information and the set of potential slip surfaces into the slope prediction model for prediction, slip surface characteristic parameters are obtained.
[0016] The calculation module is used to calculate the landslide stability coefficient based on the rigid body limit equilibrium method of the slip surface characteristic parameters, and to make early warning judgments on the landslide stability coefficient through a preset dynamic threshold.
[0017] The beneficial effects of this invention are as follows:
[0018] This invention uses the Spencer method to screen potential sliding surfaces and combines it with discrete element method (DEM) simulation technology to construct a refined two-dimensional DEM model. Based on the differences in conditions within the slope geological information, it generates a personalized two-dimensional DEM model with "one model per slope" characteristics. Secondly, it employs a long short-term memory (LSTM) network for deep learning to study the complex nonlinear relationship between the slope geological information and the aforementioned two-dimensional DEM model, constructing a customized slope prediction model for specific slopes to align with the personalized modeling concept of "one model per slope." The "one model per slope" early warning system constructed by this invention dynamically generates exclusive training data and prediction models based on the differences in geological structure, sliding surface morphology, and geotechnical parameters of different slopes, accurately capturing the unique evolutionary patterns of slopes and achieving personalized early warning. In summary, this invention addresses the problem of not being able to accurately predict landslide risks and achieve dynamic early warning.
[0019] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing embodiments of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description
[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.
[0021] Figure 1 This is a schematic diagram of the landslide intelligent early warning method described in this embodiment of the invention;
[0022] Figure 2 This is a schematic diagram of slope stability analysis as described in an embodiment of the present invention;
[0023] Figure 3 This is a schematic diagram of screening unstable slip surfaces as described in an embodiment of the present invention;
[0024] Figure 4 This is a schematic diagram of the landslide intelligent early warning device described in an embodiment of the present invention.
[0025] The markings in the diagram are: 800, landslide intelligent early warning device; 801, processor; 802, memory; 803, multimedia component; 804, I / O interface; 805, communication component. Detailed Implementation
[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.
[0027] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this invention, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.
[0028] Example 1:
[0029] This embodiment provides a landslide intelligent early warning method.
[0030] See Figure 1 The figure shows that the method includes steps S1 to S5, including:
[0031] S1: Obtain slope geological information, which includes slope topographic data, soil and rock parameters, and field monitoring data;
[0032] In this step, the slope topographic data is acquired using lidar scanning technology and UAV photography technology. This data includes information on the slope's shape, slope gradient, aspect, elevation, and soil thickness.
[0033] The parameters of the soil and rock mass are obtained through on-site investigation and sampling. The parameters of the soil and rock mass include the lithology, structure, fracture, joint, density, elastic modulus, Poisson's ratio, cohesion, internal friction angle and tensile strength of the underground soil and rock mass.
[0034] The on-site monitoring data is obtained through on-site monitoring equipment and includes cumulative displacement, velocity, acceleration, and displacement angle.
[0035] S2: Draw vector graphics for the geological information of the slope, calculate the slope stability and screen the slip surfaces using vector graphics and Spencer method, and output a set of potential slip surfaces;
[0036] To clarify the specific method for obtaining the potential slip surface set, step S2 includes S21 to S23, specifically:
[0037] S21: Draw the geometric feature data, faults, and joints in the slope geological information to obtain vector graphics;
[0038] In this step, the vector graphics are used to visually represent the geometric features and geological structure of the slope.
[0039] S22: Import the vector graphics into the preset two-dimensional limit equilibrium slope stability analysis model for construction and assign values to the soil and rock strength parameters to obtain the soil and rock strength calculation model;
[0040] In this step, based on the soil and rock parameters (cohesion, internal friction angle) in the slope geological information, parameter values are assigned to the soil and rock elements in the two-dimensional limit equilibrium slope stability analysis model. Through the two-dimensional limit equilibrium slope stability analysis model and the assignment of soil and rock strength parameters, an accurate calculation basis is provided for slope stability analysis.
[0041] S23: Perform a slip surface search on the soil and rock strength calculation model according to the Spencer method to obtain a set of potential slip surfaces.
[0042] like Figure 2 and Figure 3As shown, the number of slip surfaces is set to 5000, and the searched slip surfaces are as follows: Figure 2 As shown, all sliding surfaces with a stability coefficient less than 1.3 were selected. Figure 3 As shown, the smooth surfaces with a stability coefficient of less than 1.3 are considered the most unstable parts.
[0043] S3: Based on the discrete element simulation technology, a two-dimensional discrete element model is constructed from the geological information of the slope. By cutting the two-dimensional discrete element model and introducing random parameter perturbation, multiple slope stability analysis models are generated.
[0044] In this step, potential sliding surfaces are screened based on the Spencer method, and a refined two-dimensional discrete element model is constructed in combination with the discrete element simulation technology. Based on the differences in conditions in the slope geological information, a two-dimensional discrete element model with personalized characteristics of "one model for each slope" is generated.
[0045] To clarify the specific method for obtaining the slope stability analysis model, step S3 includes S31 to S34, specifically:
[0046] S31: Based on discrete element simulation technology, the slope topographic line in the slope geological information is initially statically calculated and constructed to obtain a two-dimensional discrete element model;
[0047] In this step, the slope topographic lines from the slope geological information are imported based on discrete element method (DEM) simulation technology to delineate the boundaries of the soil and rock layers and generate an initial mesh. Next, the Mohr-Coulomb model is used to assign differentiated parameters to each soil and rock layer region. The initial static calculation program is run to verify the convergence of the model and the rationality of the displacement distribution, ensuring that the model does not exhibit initial instability, thus constructing the two-dimensional discrete element model.
[0048] S32: Input the set of potential slip surfaces into the two-dimensional discrete element model, and obtain the slip surface parameter model by cutting the model of each slip surface in the set of potential slip surfaces and introducing random parameter perturbation;
[0049] To clarify the specific method for obtaining the slip surface parameter model, step S32 includes S321 to S323, specifically:
[0050] S321: Input the set of potential slip surfaces into the two-dimensional discrete element model, and perform model cutting on each potential slip surface through the discrete element automated operation to obtain multiple discrete element models of different slip surfaces;
[0051] In this step, a model cutting operation is performed on each potential sliding surface to divide the potential sliding surface into three parts: the sliding body, the sliding surface contact zone, and the sliding bed, while keeping the parameters of the sliding body and sliding bed regions unchanged.
[0052] S322: Perform strength parameter distribution statistics on the contact bands of the sliding surfaces in the discrete element models of the different sliding surfaces, and generate random parameter combinations by randomly perturbing the strength parameters of multiple contact bands of the sliding surfaces using the Monte Carlo method;
[0053] In this step, a statistical distribution is set for the contact bands of the sliding surfaces in the discrete element models of the different sliding surfaces, and then the Monte Carlo method is used to randomly perturb the intensity parameters of multiple contact bands of the sliding surfaces to generate random parameter combinations.
[0054] The sliding contact zone includes cohesion and internal friction angle. The statistical distribution includes a normal distribution in which the cohesion follows a preset mean and a preset standard deviation. The random parameter combination is 100 random samples for each group of sliding surfaces.
[0055] S323: Based on the combination of the sliding surface geometric parameters and the random parameters, a sliding surface parameter model is obtained by fusing them.
[0056] S33: Based on the strength reduction method, simulate the dynamic deterioration of the sliding surface parameters in the sliding surface parameter model, extract and correlate the slope displacement data and monitoring point displacement data to obtain the correlated model;
[0057] To clarify the specific method for obtaining the association model, step S33 includes steps S331 to S333, specifically:
[0058] S331: Based on the strength reduction method, the sliding surface strength parameter in the sliding surface parameter model is reduced, the dynamic deterioration process of the sliding surface parameter is simulated and the slope displacement data is extracted to obtain the slope displacement data at different deterioration stages.
[0059] In this step, the dynamic degradation process is simulated as follows: the reduction factor is gradually reduced from the initial value to a value close to the instability threshold, and the mechanical response of the slip surface parameter model is recalculated under each reduction factor.
[0060] At each stage of deterioration, slope displacement data is extracted, which is used to reflect the deformation characteristics of the slope under different strength conditions.
[0061] S332: Perform quasi-static calculations on the reduced slip surface parameter model and extract displacement time series data of monitoring points. Based on the slope displacement data of different deterioration stages and the displacement time series data of monitoring points, determine the instability model.
[0062] In this step, quasi-static calculations are performed on the reduced slip surface parameter model to simulate the mechanical response of the slope during slow loading, and displacement time-series data from monitoring points are extracted. Based on slope displacement data at different deterioration stages and displacement time-series data from monitoring points, the instability characteristics of the model are analyzed, and the instability model is determined. Quasi-static calculations continue until the model exhibits instability (abrupt displacement or calculation non-convergence).
[0063] S333: Correlate the instability model, the slope displacement data at different deterioration stages, the displacement time series data of the monitoring points, and the slip surface strength parameters to obtain a correlation model.
[0064] In this step, the correlation model is used to analyze the mechanical behavior and deformation characteristics during the slope instability process.
[0065] S34: Integrate the slope displacement data, the sliding surface geometric parameters and sliding surface strength parameters in the potential sliding surface set, and the associated model to obtain multiple slope stability analysis models.
[0066] In this step, after integration, working condition labels are added to obtain multiple slope stability analysis models, which are used for LSTM model training.
[0067] Among them, the working condition label is the slip surface number and / or reduction stage.
[0068] S4: Input multiple slope stability analysis models into a long short-term memory network for training, output a slope prediction model, and obtain the slip surface characteristic parameters by inputting the slope geological information and the potential slip surface set into the slope prediction model for prediction.
[0069] In this step, the depth of the long short-term memory network is used to learn the complex nonlinear relationship between the slope geological information and the two-dimensional discrete element model, and to construct a slope prediction model customized for a specific slope, so as to fit the personalized modeling concept of "one slope, one model".
[0070] To clarify the specific method for obtaining the surface characteristic parameters, step S4 includes S41 to S44, specifically:
[0071] S41: Extract the spatiotemporal evolution features of the slope from the two-dimensional discrete element model, and construct a multidimensional input matrix by monitoring the slope surface. Then, concatenate the spatiotemporal evolution features of the slope with the multidimensional input matrix to output an enhanced feature matrix.
[0072] To clarify the specific method for obtaining the enhanced feature matrix, step S41 includes steps S411 to S413, specifically:
[0073] S411: The two-dimensional discrete element model is denoised, and the spatiotemporal evolution features of the slope are extracted from the denoised two-dimensional discrete element model to construct the original time series features;
[0074] In this step, the two-dimensional discrete element model includes cumulative displacement, velocity, acceleration, and displacement angle. The denoising process is a moving average filter, used to remove abnormal samples caused by non-convergence of numerical calculations. The spatiotemporal evolution characteristics of the slope include the displacement, stress, and strain of the slope at different time points.
[0075] S412: Monitor the slope surface and construct a multidimensional input matrix by aligning the time-series data from different monitoring points according to timestamps;
[0076] S413: Extract statistical features of the displacement curve based on the multidimensional input matrix, and concatenate the statistical features of the displacement curve with the original time series data to output an enhanced feature matrix.
[0077] In this step, statistical features of the displacement curve are extracted based on the multidimensional input matrix. The statistical features of the displacement curve are then Z-score standardized to eliminate the dimensional differences between different monitoring points. The standardized statistical features of the displacement curve are then concatenated with the original time-series data to output an enhanced feature matrix. The statistical features of the displacement curve include mean, variance, and extreme values.
[0078] S42: Construct a bidirectional long short-term memory network based on the enhanced feature matrix, dynamically allocate weights for different temporal features through an attention mechanism, and output a slope prediction model;
[0079] In this step, the network structure of the bidirectional long short-term memory network is as follows: the input layer receives 4-dimensional temporal features, the hidden layer has 128 neurons, and an attention mechanism is introduced to focus on key time nodes. The output layer adopts a multi-task learning framework: geometric parameters and intensity parameters are predicted by independent fully connected branches, the loss function is a weighted mean squared error, and a Dropout layer and L2 regularization are added to prevent overfitting.
[0080] Training strategy: The Adam optimizer is used in conjunction with a cosine annealing scheduler to dynamically adjust the learning rate; the training set and validation set are divided in an 8:2 ratio, an early stopping method is set, and the loss of the validation set is monitored.
[0081] Model validation: After training, the feature embedding space is analyzed using visualization tools to verify the model's ability to distinguish surface parameters.
[0082] Preferably, the bidirectional long short-term memory network is a Bi-LSTM network.
[0083] S43: Optimize the slope prediction model based on knowledge distillation technology to obtain a lightweight prediction model;
[0084] In this step, knowledge distillation technology can compress complex models into lightweight models without significantly reducing model performance, thereby improving the model's running efficiency.
[0085] S44: Input the slope geological information into the lightweight prediction model to perform error prediction of potential slip surface and strength parameters, and construct slip surface characteristic parameters.
[0086] To clarify the specific method for obtaining the surface feature parameters, step S44 includes steps S441 to S444, specifically:
[0087] S441: Input the slope geological information into the lightweight prediction model to predict the potential sliding surface and strength parameters respectively, and obtain the sliding surface prediction results and strength parameter prediction results;
[0088] In this step, the predicted sliding surface results include the position, shape, and tilt angle geometry of the sliding surface, and the predicted strength parameters include cohesion and internal friction angle.
[0089] S442: Calculate the error between the predicted slip surface result and the preset actual slip surface value to obtain the slip surface prediction error value;
[0090] S443: Calculate the error between the predicted strength parameter and the actual value of the preset strength parameter to obtain the strength parameter error value;
[0091] S444: Based on the slip surface prediction error value, the strength parameter error value, and the preset geomechanical constraints, the slip surface characteristic parameters are constructed to obtain the slip surface characteristic parameters.
[0092] In this step, the geomechanical constraints include the stability conditions of the slip surface and the range of physical and mechanical properties of the rock and soil. By comprehensively considering the prediction error and the geomechanical constraints, the characteristics of the slip surface are fully described, which solves the problem of relying on only a single prediction result in traditional methods.
[0093] S5: Calculate the landslide stability coefficient based on the rigid body limit equilibrium method for the slip surface characteristic parameters, and make an early warning judgment on the landslide stability coefficient by using a preset dynamic threshold.
[0094] To clarify the specific method for obtaining early warning judgments, step S5 includes S51 to S53, specifically:
[0095] S51: The landslide stability coefficient is obtained by calculating the characteristic parameters of the slip surface based on the rigid body limit equilibrium method;
[0096] In this step, the landslide stability coefficient is calculated by the rigid body limit equilibrium method, which can accurately evaluate the stability state of the slope and solve the problem of lack of accurate calculation of the landslide stability coefficient in traditional methods.
[0097] S52: Set a dynamic threshold based on the engineering safety level and specification requirements, and judge the warning level of the landslide stability coefficient through the dynamic threshold to obtain a warning level result;
[0098] In this step, a dynamic threshold is set based on the engineering safety level and specification requirements to judge whether the slope stability is within the safe range;
[0099] Preferably, the engineering safety level is as follows: for a first-level slope, Fs < 1.05 triggers a red warning; 1.05 ≤ Fs < 1.10 triggers a yellow warning;
[0100] S53: Trigger a warning according to the warning level result.
[0101] In this step, when Fs ≤ 1.05, a red warning (text message, sound and light alarm) is triggered, and an emergency plan is started; when 1.05 < Fs ≤ 1.10, a yellow warning is triggered, and the monitoring frequency is increased to 1 minute / time; the warning result is pushed to the engineering management platform in real time, and the risk area is associated with the GIS map positioning.
[0102] Embodiment 2:
[0103] This embodiment provides a landslide intelligent warning device, which includes:
[0104] An acquisition module, used to acquire slope geological information, and the slope geological information includes slope terrain data, rock and soil body parameters and on-site monitoring data;
[0105] A drawing module, used to draw a vector graph for the slope geological information, calculate the slope stability and screen the slip surface through the vector graph and the Spencer method, and output a set of potential slip surfaces;
[0106] A construction module, used to construct a two-dimensional discrete element model for the slope geological information based on the discrete element simulation technology, and generate multiple slope stability analysis models by cutting the two-dimensional discrete element model and introducing random parameter perturbations;
[0107] To clarify the specific acquisition method of the construction module, specifically:
[0108] A calculation unit, used to perform initial static calculation and construction on the slope terrain line in the slope geological information based on the discrete element simulation technology to obtain a two-dimensional discrete element model;
[0109] The cutting unit is used to input the set of potential slip surfaces into the two-dimensional discrete element model. By cutting each slip surface in the set of potential slip surfaces and introducing random parameter perturbations, a slip surface parameter model is obtained.
[0110] To clarify the specific methods for obtaining the cutting units, the following are included:
[0111] The cutting subunit is used to input the set of potential slip surfaces into the two-dimensional discrete element model, and to cut each potential slip surface into a model through the automated discrete element operation to obtain discrete element models of multiple different slip surfaces.
[0112] The statistical subunit is used to perform intensity parameter distribution statistics on the contact zones of the sliding surfaces in the discrete element models of the different sliding surfaces, and to generate random parameter combinations by randomly perturbing the intensity parameters of multiple contact zones of the sliding surfaces using the Monte Carlo method.
[0113] The fusion subunit is used to fuse the sliding surface geometric parameters and the combination of random parameters to obtain a sliding surface parameter model.
[0114] The simulation unit is used to simulate the dynamic deterioration of the sliding surface parameters in the sliding surface parameter model based on the strength reduction method, extract and correlate the slope displacement data and monitoring point displacement data to obtain the correlated model.
[0115] To clarify the specific methods for obtaining the simulation unit, the following are included:
[0116] The simulation sub-unit is used to reduce the sliding strength parameter in the sliding surface parameter model based on the strength reduction method, simulate the dynamic deterioration process of the sliding surface parameter and extract the slope displacement data to obtain the slope displacement data at different deterioration stages.
[0117] The calculation subunit is used to perform quasi-static calculations on the reduced slip surface parameter model and extract displacement time series data of the monitoring points. Based on the slope displacement data of different deterioration stages and the displacement time series data of the monitoring points, the instability model is determined.
[0118] The association subunit is used to associate the instability model, the slope displacement data at different deterioration stages, the displacement time series data of the monitoring points, and the slip surface strength parameters to obtain the association model.
[0119] The integration unit is used to integrate the slope displacement data, the sliding surface geometric parameters and sliding surface strength parameters in the potential sliding surface set, and the associated model to obtain multiple slope stability analysis models.
[0120] The training module is used to input multiple slope stability analysis models into a long short-term memory network for training and output a slope prediction model. By inputting the slope geological information and the set of potential slip surfaces into the slope prediction model for prediction, slip surface characteristic parameters are obtained.
[0121] The calculation module is used to calculate the landslide stability coefficient based on the rigid body limit equilibrium method of the slip surface characteristic parameters, and to make early warning judgments on the landslide stability coefficient through a preset dynamic threshold.
[0122] It should be noted that the specific manner in which each module performs its operation in the apparatus described in the above embodiments has been described in detail in the embodiments of the method, and will not be elaborated here.
[0123] Example 3:
[0124] Corresponding to the above method embodiments, this embodiment also provides a landslide intelligent early warning device. The landslide intelligent early warning device described below and the landslide intelligent early warning method described above can be referred to each other.
[0125] Figure 4 This is a block diagram illustrating a landslide intelligent early warning device 800 according to an exemplary embodiment. Figure 4 As shown, the landslide intelligent early warning device 800 may include: a processor 801 and a memory 802. The landslide intelligent early warning device 800 may also include one or more of the following: a multimedia component 803, an I / O interface 804, and a communication component 805.
[0126] The processor 801 controls the overall operation of the landslide intelligent early warning device 800 to complete all or part of the steps in the aforementioned landslide intelligent early warning method. The memory 802 stores various types of data to support the operation of the landslide intelligent early warning device 800. This data may include, for example, instructions for any application or method used to operate on the landslide intelligent early warning device 800, as well as application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 802 can be implemented using any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted via the communication component 805. The audio component also includes at least one speaker for outputting audio signals. I / O interface 804 provides an interface between processor 801 and other interface modules, such as keyboards, mice, and buttons. These buttons can be virtual or physical. Communication component 805 is used for wired or wireless communication between the landslide intelligent early warning device 800 and other devices. Wireless communication includes, for example, Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, or 4G, or a combination thereof. Therefore, the corresponding communication component 805 may include a Wi-Fi module, a Bluetooth module, and an NFC module.
[0127] In an exemplary embodiment, the landslide intelligent early warning device 800 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the landslide intelligent early warning method described above.
[0128] Example 4:
[0129] Corresponding to the above method embodiments, this embodiment also provides a medium. The medium described below can be referred to in correspondence with the landslide intelligent early warning method described above.
[0130] A medium storing a computer program, which, when executed by a processor, implements the steps of the landslide intelligent early warning method described in the above method embodiments.
[0131] The medium can specifically be any medium capable of storing program code, such as a USB flash drive, external hard drive, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.
[0132] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.
[0133] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.
Claims
1. A landslide intelligent early warning method, characterized in that, include: Obtain slope geological information, which includes slope topographic data, soil and rock parameters, and field monitoring data; A vector graphic is drawn from the geological information of the slope. The slope stability is calculated and the slip surface is screened using the vector graphic and the Spencer method. A set of potential slip surfaces is output. Based on the discrete element simulation technology, a two-dimensional discrete element model is constructed from the geological information of the slope. By cutting the two-dimensional discrete element model and introducing random parameter perturbation, multiple slope stability analysis models are generated. Multiple slope stability analysis models are input into a long short-term memory network for training, and a slope prediction model is output. By inputting the slope geological information and the set of potential slip surfaces into the slope prediction model for prediction, slip surface characteristic parameters are obtained. The landslide stability coefficient is calculated based on the rigid body limit equilibrium method using the slip surface characteristic parameters, and a pre-warning judgment is made on the landslide stability coefficient by using a preset dynamic threshold.
2. The landslide intelligent early warning method according to claim 1, characterized in that, A two-dimensional discrete element model is constructed based on the slope geological information using discrete element simulation technology. Multiple slope stability analysis models are generated by segmenting the two-dimensional discrete element model and introducing random parameter perturbations, including: Based on discrete element simulation technology, the slope topographic lines in the slope geological information are initially statically calculated and constructed to obtain a two-dimensional discrete element model. The set of potential slip surfaces is input into the two-dimensional discrete element model. By cutting the model for each slip surface in the set of potential slip surfaces and introducing random parameter perturbations, a slip surface parameter model is obtained. Based on the strength reduction method, the sliding surface strength parameters in the sliding surface parameter model are simulated to dynamically deteriorate the sliding surface parameters. Slope displacement data and monitoring point displacement data are extracted and correlated to obtain a correlation model. By integrating the slope displacement data, the sliding surface geometric parameters and sliding surface strength parameters in the potential sliding surface set, and the correlation model, multiple slope stability analysis models are obtained.
3. The landslide intelligent early warning method according to claim 2, characterized in that, The set of potential slip surfaces is input into the two-dimensional discrete element model. By dividing each slip surface in the set of potential slip surfaces into models and introducing random parameter perturbations, a slip surface parameter model is obtained, including: The set of potential slip surfaces is input into the two-dimensional discrete element model. The model is then cut into each potential slip surface through the automated discrete element operation to obtain multiple discrete element models of different slip surfaces. The strength parameter distribution of the contact zone in the discrete element model of the different sliding surfaces is statistically analyzed, and the strength parameters of multiple contact zones are randomly perturbed by the Monte Carlo method to generate random parameter combinations. The sliding surface parameter model is obtained by fusing the combination of the sliding surface geometric parameters and the random parameters.
4. The landslide intelligent early warning method according to claim 2, characterized in that, The sliding surface strength parameters in the sliding surface parameter model are simulated for dynamic deterioration based on the strength reduction method. Slope displacement data and monitoring point displacement data are extracted and correlated to obtain a correlated model, including: The sliding surface strength parameter in the sliding surface parameter model is reduced based on the strength reduction method, the dynamic deterioration process of the sliding surface parameter is simulated and the slope displacement data is extracted to obtain the slope displacement data at different deterioration stages. Quasi-static calculations were performed on the reduced slip surface parameter model, and displacement time series data of monitoring points were extracted. Based on the slope displacement data of different deterioration stages and the displacement time series data of monitoring points, the instability model was determined. The instability model, the slope displacement data at different deterioration stages, the displacement time series data of the monitoring points, and the slip surface strength parameters are correlated to obtain a correlation model.
5. The landslide intelligent early warning method according to claim 1, characterized in that, Multiple slope stability analysis models are input into a long short-term memory network for training, and a slope prediction model is output. By inputting the slope geological information and the set of potential slip surfaces into the slope prediction model for prediction, slip surface characteristic parameters are obtained, including: The spatiotemporal evolution features of the slope are extracted from the two-dimensional discrete element model, and a multidimensional input matrix is constructed by monitoring the slope surface. The spatiotemporal evolution features of the slope are concatenated with the multidimensional input matrix to output an enhanced feature matrix. A bidirectional long short-term memory network is constructed based on the enhanced feature matrix. The weights of different temporal features are dynamically allocated through the attention mechanism to output a slope prediction model. The slope prediction model is optimized using knowledge distillation technology to obtain a lightweight prediction model. The slope geological information is input into the lightweight prediction model to predict the errors of potential slip surface and strength parameters, and to construct slip surface characteristic parameters.
6. The landslide intelligent early warning method according to claim 5, characterized in that, The slope geological information is input into the lightweight prediction model to predict the errors of potential slip surface and strength parameters, and to construct slip surface characteristic parameters, including: The slope geological information is input into the lightweight prediction model to predict the potential slip surface and strength parameters, respectively, and the slip surface prediction results and strength parameter prediction results are obtained. The error is calculated by comparing the predicted surface value with the actual value of the preset surface to obtain the predicted surface error value. The error value of the strength parameter is obtained by calculating the error between the predicted strength parameter and the actual value of the preset strength parameter. Based on the slip surface prediction error value, the strength parameter error value, and the preset geomechanical constraints, the slip surface characteristic parameters are constructed to obtain the slip surface characteristic parameters.
7. A landslide intelligent early warning device, characterized in that, include: The acquisition module is used to acquire slope geological information, which includes slope topographic data, soil and rock parameters, and field monitoring data. The drawing module is used to draw vector graphics on the slope geological information, perform slope stability calculation and slip surface screening using vector graphics and the Spencer method, and output a set of potential slip surfaces. The module is used to construct a two-dimensional discrete element model based on the slope geological information using discrete element simulation technology. By cutting the two-dimensional discrete element model and introducing random parameter perturbations, multiple slope stability analysis models are generated. The training module is used to input multiple slope stability analysis models into a long short-term memory network for training and output a slope prediction model. By inputting the slope geological information and the set of potential slip surfaces into the slope prediction model for prediction, slip surface characteristic parameters are obtained. The calculation module is used to calculate the landslide stability coefficient based on the rigid body limit equilibrium method of the slip surface characteristic parameters, and to make early warning judgments on the landslide stability coefficient through a preset dynamic threshold.
8. The landslide intelligent early warning device according to claim 7, characterized in that, The building module includes: The computing unit is used to perform initial static calculations on the slope topographic lines in the slope geological information based on discrete element simulation technology and construct a two-dimensional discrete element model. The cutting unit is used to input the set of potential slip surfaces into the two-dimensional discrete element model. By cutting each slip surface in the set of potential slip surfaces and introducing random parameter perturbations, a slip surface parameter model is obtained. The simulation unit is used to simulate the dynamic deterioration of the sliding surface parameters in the sliding surface parameter model based on the strength reduction method, extract and correlate the slope displacement data and monitoring point displacement data to obtain the correlated model. The integration unit is used to integrate the slope displacement data, the sliding surface geometric parameters and sliding surface strength parameters in the potential sliding surface set, and the associated model to obtain multiple slope stability analysis models.
9. The landslide intelligent early warning device according to claim 8, characterized in that, The cutting unit includes: The cutting subunit is used to input the set of potential slip surfaces into the two-dimensional discrete element model, and to cut each potential slip surface into a model through the automated discrete element operation to obtain discrete element models of multiple different slip surfaces. The statistical subunit is used to perform intensity parameter distribution statistics on the contact zones of the sliding surfaces in the discrete element models of the different sliding surfaces, and to generate random parameter combinations by randomly perturbing the intensity parameters of multiple contact zones of the sliding surfaces using the Monte Carlo method. The fusion subunit is used to fuse the sliding surface geometric parameters and the combination of random parameters to obtain a sliding surface parameter model.
10. The landslide intelligent early warning device according to claim 8, characterized in that, The simulation unit includes: The simulation sub-unit is used to reduce the sliding strength parameter in the sliding surface parameter model based on the strength reduction method, simulate the dynamic deterioration process of the sliding surface parameter and extract the slope displacement data to obtain the slope displacement data at different deterioration stages. The calculation subunit is used to perform quasi-static calculations on the reduced slip surface parameter model and extract displacement time series data of the monitoring points. Based on the slope displacement data of different deterioration stages and the displacement time series data of the monitoring points, the instability model is determined. The association subunit is used to associate the instability model, the slope displacement data at different deterioration stages, the displacement time series data of the monitoring points, and the slip surface strength parameters to obtain the association model.