Slope reliability analysis method, system and device based on bo-lcmc model

Abstract

The application discloses a slope reliability analysis method, system and equipment based on a BO-LCMC model, and belongs to the technical field of slope engineering data processing. K The traditional deterministic analysis method is prone to neglecting stratum spatial variability and a local coupling Markov chain model when analyzing slope reliability K The slope reliability analysis method in the application first establishes a local coupling Markov chain model, optimizes and solves the local coupling Markov chain model by a Bayesian optimization algorithm, generates a random stratum profile, then, considers stratum spatial variability, effectively maps the random stratum profile to a slope finite element model, calculates the safety factor and slope instability probability of the slope engineering, and analyzes the slope stability. The method effectively improves the modeling precision of complex variable strata and the accuracy of slope reliability evaluation, and is especially suitable for high-speed railway cutting engineering.

Description

Technical Field

[0001] This invention relates to the field of slope engineering data processing technology, and in particular to slope reliability analysis methods, systems and equipment based on the BO-LCMC model. Background Technology

[0002] In the field of road cut slope engineering, slope stability is influenced by a combination of factors, including active fault displacement, spatial distribution of soil and rock mass structures, and transient coupling of rainfall and groundwater seepage. Furthermore, the spatial variability of the strata (such as paleoweathering crust, weak interlayers, and unloading fracture zones) also has a decisive impact on slope stability. However, when analyzing slope stability, traditional limit equilibrium methods are insufficient to fully quantify the potential landslide risks induced by the spatial variability of the strata. Traditional deterministic analysis methods treat stratum parameters as fixed values, ignoring the inherent randomness of the spatial distribution of geological bodies, leading to systematic biases in safety assessment results. These biases manifest in two extremes in practical engineering: either excessively conservative design resulting in resource waste, or underestimation of risk leading to disasters. Therefore, developing reliability analysis methods that integrate the spatial variability of the strata has become a core requirement for improving the safety and economy of slope engineering.

[0003] Compared to traditional deterministic analysis methods, stochastic methods utilize geological exploration data, combined with lithofacies sedimentology theory and probabilistic statistics principles, to construct spatially correlated stochastic geological models. Their algorithms are flexible and diverse, without fixed patterns, and better reflect the randomness and variability of stratigraphic structures. However, existing research mainly focuses on simulating the spatial distribution of binary soil-rock configurations (i.e., a single type of soil embedded within another matrix soil), lacking effective methods for characterizing the multi-component composite configurations commonly found in practical engineering. To address this limitation, the academic community has introduced Markov chain stochastic process models. By constructing a multi-state transition probability matrix (MTBM), probabilistic modeling of the spatially alternating distribution of various soil and rock materials is achieved, providing a theoretical tool for studying complex and variable stratigraphic formations. The locally coupled Markov chain model (LCMC) uses a combination of localized geological profiles and stochastic simulation to simulate stratigraphic structures, improving the model's ability to characterize strata with drastic local changes. However, as the number of lithological categories increases, the dimension of the model's state transition matrix grows rapidly, leading to a sharp decline in computational efficiency and limiting its application in large-scale modeling. Furthermore, in the basic Markov chain model… K Value estimation methods suffer from insufficient accuracy when dealing with nonlinear, multi-peaked strata distributions, and local optima. Furthermore, traditional slope stability analyses, such as the classical limit equilibrium method and simplified mechanical models, while computationally simple and empirically mature and still widely used in slope engineering, struggle to capture the spatial variability prevalent within real geological formations.

[0004] To address the above issues, there is an urgent need to design a simple and efficient slope reliability analysis method to solve the problems existing in the current technology. Summary of the Invention

[0005] To address the aforementioned problems, this invention aims to provide a slope reliability analysis method, system, and equipment based on the BO-LCMC model. The core technical approach is Bayesian (BO) optimization of the LCMC model for geological simulation, finite element mapping, and analysis. On one hand, Bayesian optimization improves the locally coupled Markov chain model, effectively addressing its original shortcomings and thus more accurately reflecting the spatial variability of the geological strata. On the other hand, the geological information obtained from the improved model simulation is mapped to the finite element slope model for slope reliability analysis. This method possesses three core advantages: first, it offers higher accuracy in geological simulation, enabling a more precise interpretation of geological characteristics; second, it simplifies the cumbersome calculation process under the basic LCMC model; and third, it provides reliable support for stability analysis and decision-making in slope and cutting engineering.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0007] On the one hand, the present invention provides a slope reliability analysis method based on the BO-LCMC model, the analysis method comprising the following steps:

[0008] The original borehole data of the slope engineering were obtained, and the borehole profile was divided into mesh cells and multiple CMC (Markov chain) segments based on the locally coupled Markov chain model.

[0009] Extract each CMC segment and construct a vertical transition frequency matrix based on the number of formation state transitions between adjacent strata;

[0010] Randomly generate multiple locally coupled Markov chain models K Values ​​are used to construct a horizontal transition frequency matrix based on the vertical transition frequency matrix.

[0011] Using Bayesian optimization algorithm K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The horizontal transition frequency matrix corresponding to the value is used to construct the final horizontal transition frequency matrix;

[0012] Based on the final horizontal transfer frequency matrix, the borehole profile is simulated using the Monte Carlo method to generate a random stratigraphic profile.

[0013] Spatial matching of random stratigraphic profiles with the finite element model of the slope is performed to calculate the safety factor of the slope engineering. and slope instability probability Slope stability analysis was conducted.

[0014] Optionally, the specific operations for acquiring the original borehole data of the slope engineering and dividing the borehole profile into mesh cells and multiple CMC segments based on the locally coupled Markov chain model include:

[0015] Obtain the original borehole data for the slope engineering and preprocess the original borehole data to obtain the stratigraphic state of the borehole;

[0016] The preprocessed raw borehole data is extracted, and the borehole profile is divided into mesh cells based on the locally coupled Markov chain model. The borehole profile after mesh cell division is then divided into multiple CMC segments.

[0017] Optionally, the method of using Bayesian optimization algorithm to... K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The specific operations for constructing the final horizontal transition frequency matrix, corresponding to the values, include:

[0018] Based on the constructed vertical and horizontal transfer frequency matrices, Monte Carlo simulations are performed to generate stratigraphic profile sequences. The borehole locations and stratigraphic state distributions of the stratigraphic profile sequences are statistically analyzed to construct an initial sample set.

[0019] Based on the initial sample set, the Bayesian optimization algorithm is used to... K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The horizontal transition frequency matrix corresponding to the value is used to construct the final horizontal transition frequency matrix.

[0020] Optionally, the initial sample set is:

[0021] ;

[0022] In the formula, Represents the initial sample set. L ( K ) represents the final stratigraphic matching probability. , P Represents probability. For the first k The first real borehole g Stratigraphic state of each geological unit For all known borehole data, subscript q Indicates the first q indivual K value, for K The number of values.

[0023] Optionally, the method of using Bayesian optimization algorithm to...K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The specific steps for constructing the final horizontal transition frequency matrix, which corresponds to the values, include the following:

[0024] Establish L ( K Gaussian process surrogate model;

[0025] Based on the initial sample set, the parameters of the Gaussian process surrogate model are determined using the maximum marginal likelihood method. ;

[0026] Finding the optimal solution based on iterative optimization of the expected improvement function. K Value, denoted as ;

[0027] Output ;

[0028] according to Calculate the corresponding horizontal transition frequency matrix to construct the final horizontal transition frequency matrix. ,in, express The corresponding adjacent geological units differ in stratigraphic state in the horizontal direction. Transition to stratigraphic state The horizontal transition frequency matrix, ; , The total number of stratigraphic states; Represents the horizontal direction. This is the optimal solution in the horizontal direction.

[0029] Optionally, the optimal solution K The value is represented as:

[0030] ;

[0031] In the formula, D T Indicates up to the T Up to the next iteration, all The set that constitutes the composition.

[0032] Optionally, the final horizontal transition frequency matrix Represented as:

[0033] ;

[0034] In the formula, for K Value The horizontal stratigraphic state transition matrix corresponding to the time.

[0035] Optionally, the random stratigraphic profile is spatially matched with the finite element model of the slope to calculate the safety factor of the slope engineering. and slope instability probability The specific procedures for performing slope stability analysis include:

[0036] Construct a finite element model of the slope and spatially match the coordinates of the center points of the grid cells in the finite element model with a random stratigraphic profile.

[0037] The finite element method with strength reduction is used to analyze the slope stability and calculate the safety factor of the slope engineering. ;

[0038] statistics The sample probability is used as the slope instability probability. ,in, This is the critical safety factor.

[0039] On the other hand, the present invention also provides a slope reliability analysis system based on the BO-LCMC model. This analysis system is used to implement the slope reliability analysis method based on the BO-LCMC model as described above. The analysis system includes...

[0040] The borehole data preprocessing module is used to preprocess the raw borehole data and obtain the formation status of the borehole.

[0041] The borehole profile division module, based on the locally coupled Markov chain model, divides the borehole profile into mesh cells and divides the borehole profile into multiple CMC segments.

[0042] The formation state transition analysis module is used to construct the vertical transition frequency matrix and different K The horizontal transition frequency matrix corresponding to the value;

[0043] The Bayesian optimization module is used to construct an initial sample set based on the data from the formation state transition analysis module, and to optimize the Bayesian optimization module. K The values ​​are optimized to calculate the optimal solution. K The corresponding horizontal transition frequency matrix is ​​used to construct the final horizontal transition frequency matrix;

[0044] The Monte Carlo simulation module is used to perform Monte Carlo simulations on borehole profiles of CMC segments to generate random stratigraphic profiles.

[0045] The slope stability calculation and analysis module is used to spatially match random geological profiles with slope finite element models to calculate the safety factor of slope engineering projects. and slope instability probability Slope stability analysis was conducted.

[0046] In another aspect, the present invention also provides an electronic device, the device including at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions executed by the processor, the instructions being executed by the processor to enable the processor to perform the slope reliability analysis method based on the BO-LCMC model as described above.

[0047] The beneficial effects of this invention are:

[0048] 1. This invention deeply integrates the Bayesian optimization framework with the locally coupled Markov chain model. It establishes a Gaussian process surrogate model with the final stratigraphic matching probability of the stratigraphic simulation as the objective function, and uses the Expected Improvement (EI) criterion to intelligently search for the optimal solution. K This allows for the accurate quantification and efficient simulation of complex and variable geological spatial structures.

[0049] 2. This invention constructs a complete slope stability analysis process, from random geological strata modeling to finite element modeling. By accurately mapping a large number of random geological profiles generated by BO-LCMC simulation to the slope finite element model, the probability distribution of the safety factor of the slope engineering is calculated using the finite element strength reduction method. Finally, the probability of slope instability considering the spatial variability of the geological strata is scientifically provided. This method significantly improves the accuracy and efficiency of slope reliability assessment under complex geological conditions, and is particularly suitable for major projects such as high-speed railway cuttings with extremely high requirements for safety and settlement control. It provides an effective technical approach to solve the dilemma of conservative or overly optimistic results in traditional deterministic analysis. Attached Figure Description

[0050] Figure 1 This is a flowchart of the slope reliability analysis method in this invention.

[0051] Figure 2 This is a geological state diagram of the actual borehole drilled in this invention.

[0052] Figure 3 This is a schematic diagram of the borehole profile grid unit and CMC segment division in this invention.

[0053] Figure 4 Bayesian optimization in this invention K A flowchart of the value. Detailed Implementation

[0054] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0055] Example 1: Refer to Appendix Figure 1This embodiment provides a slope reliability analysis method based on the BO-LCMC model. The analysis method includes the following steps:

[0056] Step 1: Obtain the original borehole data for the slope engineering, and divide the borehole profile into mesh cells and multiple CMC segments based on the locally coupled Markov chain model;

[0057] Optionally, step 1 includes the following sub-steps:

[0058] Sub-step 101: Obtain the original borehole data for the slope engineering and preprocess the original borehole data to obtain the geological conditions of the boreholes;

[0059] See attached document Figure 2 As shown, the specific operations for obtaining and preprocessing the raw borehole data for slope engineering include:

[0060] During the acquisition of the original borehole data, it is assumed that the slope engineering is equipped with... N There are 1 drill hole, labeled ZK1, ZK2, ..., ZK. N The formation condition information of each borehole is collected vertically, and different formation conditions are marked sequentially from top to bottom. S 1, S 2,…, S n , n This represents the total number of stratigraphic states, with different values ​​indicating different stratigraphic states. For example, during drilling, a total of four different soil layers were found: sand, clay, silt, and silty clay, represented by yellow, cyan, purple, and red, respectively, corresponding to the stratigraphic states. S 1, S 2, S 3, S 4. The order of formation conditions differs in different boreholes. The maximum elevation is obtained based on the actual borehole elevation range within the study area. and minimum elevation Then the elevation range Divided into b There are two equal intervals, and the length of each interval is... L Then set to The unified stratigraphic sequence of the slope engineering project can be determined based on the borehole distribution.

[0061] Sub-step 102: Extract the preprocessed raw borehole data, divide the borehole profile into mesh cells based on the locally coupled Markov chain model, and divide the borehole profile into multiple CMC segments.

[0062] Specifically, the preprocessed raw borehole data is extracted, and the borehole profile is divided into grid cells based on the minimum formation thickness information of the borehole. A local coupled Markov chain model is constructed to simulate the geological profile. The borehole profile is then divided into multiple CMC segments, and the Markov property of the preprocessed raw borehole data is checked.

[0063] Optionally, refer to the appendix Figure 3 Sub-step 102 includes the following sub-steps:

[0064] Sub-step 1021: Standardize the elevation range of each borehole, and determine the borehole profile length and vertical depth accordingly. Divide the borehole profile into sections based on the minimum formation thickness information. The grid cells, that is Each geological unit.

[0065] Then, the entire borehole profile was divided into multiple CMC segments. The principle for dividing the CMC segments was based on the profile to be inferred (including surface data and...). N (One borehole), constructed by adjacent borehole triplets (two boundary boreholes + center borehole). N-2 A CMC segment, as shown in the attached image. Figure 3 As shown.

[0066] Sub-step 1022: Establish the transition frequency matrix of the borehole profile based on the preprocessed borehole data, and calculate the expected frequency of each geological unit. , x and y Let represent the number of rows and columns of the transition frequency matrix, respectively. x =1,2,…, n ; y =1,2,…, n Calculate the chi-square statistic. ,in, This indicates the stratigraphic state of two adjacent geological units. Transition to stratigraphic state Number of times, i =1,2,…, n ; j =1,2,…, n .

[0067] Sub-step 1023: Perform a Markov test on the calculation results in sub-step 1022;

[0068] like If the value is less than the critical value, then the Markov test is satisfied. The optimal order of the Markov chain is then determined using the AIC criterion. If a Markov chain of a certain order passes the test... If the AIC value is minimized, then the Markov chain of this order can well describe the state transition process of the soil and rock mass.

[0069] The AIC criterion selects the optimal model order by balancing the model's goodness of fit (e.g., log-likelihood) and model complexity (e.g., number of parameters). The formula for calculating the AIC value is:

[0070] ;

[0071] In the formula: This represents the maximum likelihood function value of the model. This represents the number of independent parameters in the model.

[0072] Step 2: Extract the CMC fragment from Step 1 and construct the Vertical Transfer Frequency Matrix (VPTM) based on the number of stratigraphic state transfers between adjacent geological units;

[0073] Specifically, adjacent geological units are divided vertically from stratigraphic state. Transition to stratigraphic state The frequency is expressed as , v Representing the vertical direction, the vertical direction of the stratigraphic state transition Represented as:

[0074] ;

[0075] Therefore, the vertical transition frequency matrix VPTM can be expressed as:

[0076] ;

[0077] In the formula, .

[0078] Step 3: Randomly generate multiple locally coupled Markov chain models. K The value is used to construct the horizontal transfer frequency matrix HTPM based on the vertical transfer frequency matrix VTPM.

[0079] Specifically, during initialization, the locally coupled Markov chain model is determined based on geological conditions and engineering experience. K value space , for K The minimum value, for K The maximum value; in Randomly select from within the range For each K value, K The corresponding horizontal transition frequency matrix HTPM can be calculated from the vertical transition frequency matrix VTPM to which the value belongs.

[0080] Optionally, the construction process of the horizontal transition frequency matrix HTPM specifically includes,

[0081] Adjacent geological units are classified in the horizontal direction based on their stratigraphic state. Transition to stratigraphic state The frequency is expressed as , Representing the horizontal direction, the horizontal stratigraphic state transition matrix is... Represented as:

[0082] ;

[0083] The horizontal transition frequency matrix HTPM is then represented as:

[0084] ;

[0085] In the formula, .

[0086] Step 4: Use the Bayesian optimization algorithm to... K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The horizontal transition frequency matrix corresponding to the value is used to construct the final horizontal transition frequency matrix;

[0087] Optionally, step 4 includes the following sub-steps:

[0088] Sub-step 401: Based on the constructed vertical transfer frequency matrix VTPM and horizontal transfer frequency matrix HTPM, perform Monte Carlo simulation to generate a stratigraphic profile sequence, statistically analyze the borehole locations and stratigraphic state distribution of the stratigraphic profile sequence, and construct an initial sample set.

[0089] Specifically, based on the aforementioned constructed vertical transfer frequency matrix VTPM and horizontal transfer frequency matrix HTPM, the final formation matching probability is calculated. L ( K ), represented as:

[0090] ;

[0091] In the formula, P Represents probability. For the first k The first real borehole g Stratigraphic state of each geological unit For all known borehole data. Information obtained from the slope engineering site includes spatial location information of boreholes (such as the coordinates of each borehole) and stratigraphic sequence information of each borehole (the stratigraphic state of each borehole at different depths, such as borehole ZK1: 0-2m sand (S1), 2-5m clay (S2), 5-8m silt (S3); borehole ZK2: 0-3m silty clay (S4), 3-6m sand (S1)).

[0092] according to K Value and L ( K ) Construct the initial dataset In the formula, the subscript q Indicates the first q indivual K value.

[0093] Sub-step 402: Based on the initial sample set, use the Bayesian optimization algorithm to... K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The corresponding horizontal transition frequency matrix HTPM is used to construct the final horizontal transition frequency matrix;

[0094] Specifically, based on the Bayesian optimization algorithm, the formation with the highest probability of matching the actual borehole location is selected. K value as the optimal solution K Value, denoted as ,calculate The corresponding horizontal transition frequency matrix HTPM is used to construct the final horizontal transition frequency matrix.

[0095] Optionally, the specific operation process of sub-step 402 is as follows: Figure 4 As shown, it includes the following sub-steps:

[0096] Sub-step 4021, establish L ( K Gaussian process surrogate model;

[0097] Assumption L ( K The objective function follows a Gaussian process. L ( K The probability distribution model of )

[0098] ;

[0099] ;

[0100] In the formula, To conform to a Gaussian distribution, It is a mean function. For two random value and covariance function Indicates the signal variance. Indicates the length scale. It is a hyperparameter that controls the objective function. L ( K The range of fluctuation of the value. The larger the value, the better the objective function. L ( K The greater the fluctuation around its mean, the better. This is another key hyperparameter that controls the smoothness of the objective function; it defines the range of "similar" distances. The larger the value, the stronger the objective function. L ( K The slower the value changes, the smoother the result. The smaller the value, the stronger the objective function. L ( K The more drastic the value changes, the more convoluted the story becomes.

[0101] Sub-step 4022: Based on the initial sample set, determine the parameters of the Gaussian process surrogate model using the maximum marginal likelihood method. :

[0102] ;

[0103] in, Marginal likelihood Represented as:

[0104] ;

[0105] In the formula, This represents the vector of observed target values, containing all evaluated values. K The objective function corresponding to the value A column vector of values , S Indicates that it has been evaluated K The number of values; express The covariance matrix, This represents the transpose of a matrix.

[0106] Sub-step 4023: Iterative optimization to find the optimal solution. K value;

[0107] Adopting the desired improvement ( EI Function-guided search:

[0108] ;

[0109] ;

[0110] ;

[0111] ;

[0112] In the formula, Best at present The observation is all the values ​​in the current iteration. K Among the values, the one with the highest matching probability. K value, This represents the standard deviation of the Gaussian process prediction, which is the Gaussian process surrogate model's prediction of itself. Uncertainty measure The larger the value, the better the Gaussian process surrogate model is at that point. K The less data there is around a value, the higher the uncertainty. This is a tradeoff parameter, a very small positive number (usually set to 0.01), used to balance exploration and exploitation; This represents the probability density function of the standard normal distribution. The cumulative distribution function represents the standard normal distribution.

[0113] For newly selected during the iteration process K Value, denoted as Perform a complete formation simulation and calculate Then update the dataset and retrain the Gaussian process proxy model until the Gaussian process proxy model converges, at which point the updates stop.

[0114] The update of the dataset is represented as:

[0115] ;

[0116] Convergence conditions include (updates will stop if any one of the convergence conditions is met):

[0117] ;

[0118] In the formula, This indicates the training of the Gaussian process surrogate model; For the updated dataset, The dataset before the update; for EI Threshold; To improve the threshold; The maximum number of iterations, This represents the current iteration number;

[0119] Sub-step 4024, output the optimal solution. K value;

[0120] Optimal solution K value Represented as:

[0121] ;

[0122] In the formula, D T Indicates up to the T Up to the next iteration, all The set that constitutes the composition.

[0123] Sub-step 4025, based on the optimal solution K The corresponding HTPM is calculated to construct the final horizontal transition frequency matrix. , That is to say The corresponding adjacent geological units differ in stratigraphic state in the horizontal direction. Transition to stratigraphic state The horizontal transition frequency matrix, Represented as:

[0124] ;

[0125] In the formula, for K Value The corresponding horizontal stratigraphic state transition matrix at that time. This is the optimal solution in the horizontal direction.

[0126] Step 5: Based on the final horizontal transfer frequency matrix, simulate the borehole profile using the Monte Carlo method to generate a random formation profile.

[0127] Specifically, based on the vertical transfer frequency matrix and the final horizontal transfer frequency matrix, the borehole profile of the CMC segment is simulated using the Monte Carlo method. By tracking and calibrating the location and attributes of geological units in the borehole profile, a color coding method is used to color-code different types of geological units to generate random stratigraphic profile maps.

[0128] In the Monte Carlo simulation process, to ensure the continuity of the simulation results, the boundary strata distribution should be smoothed to reduce the abrupt changes at the segment transitions. Since each CMC segment undergoes the same number of simulations, the probability distribution of shared units is calculated using the mean.

[0129] Specifically, no. x The first CMC segment and the first x The probability of the final stratigraphic state of the geological unit in the overlapping region of +1 CMC segments can be expressed as:

[0130] ;

[0131] In the formula, and These represent the geological unit Z in the first place. x The first CMC segment and the first x +1 CMC fragment is in formation state The probability of.

[0132] Step 6: Spatial matching of the random stratigraphic profile with the slope finite element model to calculate the safety factor of the slope engineering. and slope instability probability Slope stability analysis was conducted.

[0133] Optionally, step 6 includes the following sub-steps,

[0134] Sub-step 601, considering the spatial variability of strata, is crucial for slope engineering stability analysis. The key lies in effectively mapping the data from the random stratigraphic profile generated by Monte Carlo simulation to the slope finite element model. First, the coordinates of the center points of the grid cells in the slope finite element model are spatially matched with the random stratigraphic profile. Then, based on the location of the grid cells, the type of geological unit and its physical and mechanical parameters (cohesion, internal friction angle, elastic modulus, and Poisson's ratio, etc.) corresponding to each grid cell are assigned to the corresponding grid cells in the slope finite element model, thus achieving spatial mapping of strata properties.

[0135] In an embodiment of the present invention, sub-step 601 specifically includes the following steps:

[0136] Sub-step 6011: In the "Part" module of Abaqus / CAE, based on the preliminary design cross-sectional data of the slope engineering, construct a two-dimensional geometric model of the slope engineering through sketches.

[0137] Sub-step 6012: In the "Property" module of Abaqus / CAE, define the corresponding material parameters for different geological conditions. Due to the variability of geological strata in road cut slopes, for soil and rock masses with different geological conditions, assign corresponding material properties based on engineering geological survey reports and geotechnical test data, and select the Mohr-Coulomb constitutive model as the finite element model for the slope. The formula for the Mohr-Coulomb constitutive model is:

[0138] ;

[0139] In the formula, Shear stress; It is the soil cohesion; It is normal stress; It is the internal friction angle; F This is the strength reduction factor.

[0140] Sub-step 6013: Enter the "Mesh" module of Abaqus to mesh the slope finite element model. To ensure effective mapping of stratigraphic properties between the random stratigraphic profile and the slope finite element model, it is necessary to ensure that the element meshing and size are consistent in both.

[0141] Sub-step 6014 assigns a set of random stratigraphic distribution data (including soil and rock types and corresponding parameters: cohesion, internal friction angle, elastic modulus, Poisson's ratio, etc.) from the random stratigraphic profile to the slope finite element model element by element, thereby achieving effective spatial mapping of stratigraphic properties.

[0142] Sub-step 6015: Based on the actual stress conditions of the slope engineering, apply fixed constraints and horizontal constraints to the bottom and sides of the slope finite element model, respectively, and apply self-weight load to restore the true stress state of the slope engineering under self-weight.

[0143] Sub-step 602 involves using the finite element strength reduction method to perform slope stability analysis and calculate the safety factor for the slope engineering. .

[0144] Specifically, the strength reduction method is used in the "Step" module of Abaqus / CAE, and an initial strength reduction coefficient is set. F By analyzing the gradual reduction of soil cohesion With internal friction angle This study simulates the deformation evolution of a slope engineering project under the continuous reduction of shear strength parameters (soil cohesion and internal friction angle). When the numerical calculation fails to converge, the slope is deemed to have reached an unstable state, and the corresponding strength reduction factor is the safety factor of the slope engineering project. .

[0145] The shear strength parameter after reduction using the strength reduction method can be expressed as:

[0146] ;

[0147] In the formula, The reduced soil cohesion, This is the reduced internal friction angle.

[0148] Statistical analysis was performed on the results of multiple Monte Carlo simulations to obtain the safety factor for slope engineering. The statistical characteristics are expressed by the formula:

[0149] ;

[0150] ;

[0151] In the formula, This represents the average safety factor for slope engineering. The standard deviation of the safety factor for slope engineering; For the number of Monte Carlo simulations, Indicates the first Monte Carlo simulation.

[0152] Sub-step 603, statistics The sample probability ( The critical safety factor (usually taken as 1) serves as the probability of slope instability. .

[0153] When considering geological variations, the failure probability is derived from the probability distribution of the safety factor for slope engineering. Assume the safety factor for slope engineering... Let be a continuous random variable, satisfying the probability density function, when Less than the critical safety factor When the value is typically 1, the slope engineering is considered to have failed. The calculation formula is as follows:

[0154] .

[0155] Example 2: This application provides a slope reliability analysis system based on the BO-LCMC model. The analysis system is used to implement the slope reliability analysis method described in Example 1.

[0156] Optionally, the analysis system includes,

[0157] The borehole data preprocessing module is used to preprocess the raw borehole data and obtain the formation status of the borehole.

[0158] The borehole profile division module, based on the locally coupled Markov chain model, divides the borehole profile into mesh cells and divides the borehole profile into multiple CMC segments.

[0159] The formation state transition analysis module is used to construct the vertical transition frequency matrix (VPTM) and different... K The horizontal transition frequency matrix HTPM corresponding to the value;

[0160] The Bayesian optimization module is used to construct an initial sample set based on the data from the formation state transition analysis module, and to optimize the Bayesian optimization module. K The values ​​are optimized to calculate the optimal solution. K The corresponding HTPM values ​​are used to construct the final horizontal transition frequency matrix;

[0161] The Monte Carlo simulation module is used to perform Monte Carlo simulations on borehole profiles of CMC segments to generate random stratigraphic profiles.

[0162] The slope stability calculation and analysis module is used to spatially match random geological profiles with slope finite element models to calculate the safety factor of slope engineering projects. and slope instability probability Slope stability analysis was conducted.

[0163] This application also provides an electronic device, including at least one processor and a memory communicatively connected to the processor; wherein the memory stores instructions executed by the processor, the instructions being executed by the processor to enable the processor to perform the slope reliability analysis method described in Embodiment 1. The electronic device can be a terminal or other devices besides a terminal. For example, the electronic device can be a GPU BOX, mobile phone, tablet computer, laptop computer, PDA, in-vehicle electronic device, mobile internet device (MID), augmented reality (AR) / virtual reality (VR) device, robot, wearable device, ultra-mobile personal computer (UMPC), netbook, or personal digital assistant (PDA), etc., and can also be a server, network attached storage (NAS), personal computer (PC), television (TV), ATM, or self-service machine, etc. This application does not specifically limit the scope of the electronic device.

[0164] The device in this application embodiment can be a device with an operating system. The operating system can be Android, Linux, Windows, or other possible operating systems; this application embodiment does not specifically limit it.

[0165] This application also provides a computer-readable storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements the steps in the slope reliability analysis method disclosed in Embodiment 1 of this application.

[0166] This application also provides a computer program product that, when run on an electronic device, enables a processor to execute the steps in the slope reliability analysis method disclosed in Embodiment 1 of this application.

[0167] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0168] This application describes embodiments with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, electronic devices, and computer program products according to embodiments of this application. It should 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 terminal device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing terminal device, 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.

[0169] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing terminal device to operate 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.

[0170] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal equipment, causing a series of operational steps to be performed on the computer or other programmable terminal equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable terminal 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.

[0171] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A slope reliability analysis method based on the BO-LCMC model, characterized in that, The analytical method includes the following steps. The original borehole data of the slope engineering were obtained, and the borehole profile was divided into mesh cells and multiple CMC segments based on the locally coupled Markov chain model. Extract each CMC segment and construct a vertical transition frequency matrix based on the number of formation state transitions between adjacent strata; Randomly generate multiple locally coupled Markov chain models K Values ​​are used to construct a horizontal transition frequency matrix based on the vertical transition frequency matrix. Using Bayesian optimization algorithm K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The horizontal transition frequency matrix corresponding to the value is used to construct the final horizontal transition frequency matrix; Based on the final horizontal transfer frequency matrix, the borehole profile is simulated using the Monte Carlo method to generate a random stratigraphic profile. Spatial matching of random stratigraphic profiles with the finite element model of the slope is performed to calculate the safety factor of the slope engineering. and slope instability probability Slope stability analysis was conducted. Among them, the use of Bayesian optimization algorithm for K The values ​​are optimized to determine the optimal solution. K Value, calculate the optimal solution K The specific operations for constructing the final horizontal transition frequency matrix, corresponding to the values, include: Based on the constructed vertical and horizontal transfer frequency matrices, Monte Carlo simulations are performed to generate stratigraphic profile sequences. The borehole locations and formation states of these sequences are statistically analyzed to construct an initial sample set. The initial sample set is as follows: ； In the formula, Represents the initial sample set. L ( K ) represents the final stratigraphic matching probability. , P Represents probability. For the first k The first real borehole g Stratigraphic state of each geological unit For all known borehole data, subscript q Indicates the first q indivual K value, for K The number of values; Establish L ( K Gaussian process surrogate model; Based on the initial sample set, the parameters of the Gaussian process surrogate model are determined using the maximum marginal likelihood method. ; Finding the optimal solution based on iterative optimization of the expected improvement function. K Value, denoted as ; Output ; In the formula, D T Indicates up to the T Up to the next iteration, all The set that constitutes; according to Calculate the corresponding horizontal transition frequency matrix to construct the final horizontal transition frequency matrix. ,in, express The corresponding adjacent geological units differ in stratigraphic state in the horizontal direction. Transition to stratigraphic state The horizontal transition frequency matrix, ; , The total number of stratigraphic states; Represents the horizontal direction. This is the optimal solution in the horizontal direction.

2. The slope reliability analysis method based on the BO-LCMC model according to claim 1, characterized in that, The specific operations for acquiring raw borehole data for slope engineering and dividing the borehole profile into mesh cells and multiple CMC segments based on a locally coupled Markov chain model include: Obtain the original borehole data for the slope engineering and preprocess the original borehole data to obtain the stratigraphic state of the borehole; The preprocessed raw borehole data is extracted, and the borehole profile is divided into mesh cells based on the locally coupled Markov chain model. The borehole profile after mesh cell division is then divided into multiple CMC segments.

3. The slope reliability analysis method based on the BO-LCMC model according to claim 1, characterized in that, Final horizontal transition frequency matrix Represented as: ； In the formula, for K Value The horizontal stratigraphic state transition matrix corresponding to the time.

4. The slope reliability analysis method based on the BO-LCMC model according to claim 3, characterized in that, Spatial matching of random stratigraphic profiles with the finite element model of the slope is performed to calculate the safety factor of the slope engineering. and slope instability probability The specific procedures for performing slope stability analysis include: Construct a finite element model of the slope and spatially match the coordinates of the center points of the grid cells in the finite element model with a random stratigraphic profile. The finite element method with strength reduction is used to analyze the slope stability and calculate the safety factor of the slope engineering. ; statistics The sample probability is used as the slope instability probability. ,in, This is the critical safety factor.

5. A slope reliability analysis system based on the BO-LCMC model, wherein the analysis system is used to implement the slope reliability analysis method based on the BO-LCMC model as described in any one of claims 1-4, characterized in that, The analysis system includes, The borehole data preprocessing module is used to preprocess the raw borehole data and obtain the formation status of the borehole. The borehole profile division module, based on the locally coupled Markov chain model, divides the borehole profile into mesh cells and divides the borehole profile into multiple CMC segments. The formation state transition analysis module is used to construct the vertical transition frequency matrix and different K The horizontal transition frequency matrix corresponding to the value; The Bayesian optimization module is used to construct an initial sample set based on the data from the formation state transition analysis module, and to optimize the Bayesian optimization module. K The values ​​are optimized to calculate the optimal solution. K The corresponding horizontal transition frequency matrix is ​​used to construct the final horizontal transition frequency matrix; The Monte Carlo simulation module is used to perform Monte Carlo simulations on borehole profiles of CMC segments to generate random stratigraphic profiles. The slope stability calculation and analysis module is used to spatially match random geological profiles with slope finite element models to calculate the safety factor of slope engineering projects. and slope instability probability Slope stability analysis was conducted.

6. An electronic device, characterized in that, The device includes at least one processor; and a memory communicatively connected to the processor; wherein the memory stores instructions that are executed by the processor to enable the processor to perform the slope reliability analysis method based on the BO-LCMC model as described in any one of claims 1-4.

Images