A dangerous rock instability risk combined early warning method, device and medium

Abstract

The present application relates to the technical field of geotechnical engineering geological disaster prevention, and discloses a dangerous rock instability risk combined early warning method, device and medium, the present application solves the problem of mechanism modeling and cracking characteristics of dangerous rock in existing technology by constructing a discrete element numerical model specific to dangerous rock. By screening key parameters and using real-time monitoring data to establish a dynamic probability distribution, the defects of traditional method probability calculation static and monitoring data are overcome, so that the instability probability can be dynamically updated with the actual evolution of the dangerous rock; by setting a hierarchical updating mechanism, the dynamic probability distribution and the discrete element numerical model are adjusted hierarchically when the monitoring data reaches the preset threshold, achieving a balance between calculation accuracy and efficiency; finally, by fusing the dynamic instability probability and real-time monitoring data to output early warning information, the status of single early warning decision rule and high false alarm rate is changed.

Description

Technical Field

[0001] This invention relates to the field of geological disaster prevention and control technology in geotechnical engineering, specifically to a joint early warning method, device and medium for the risk of unstable rock formations. Background Technology

[0002] Unstable rock masses are a common type of geological hazard in mountainous and canyon areas. Unlike landslides, which are mainly controlled by the overall shear failure of the sliding surface, the instability evolution of unstable rock masses is essentially a gradual failure process in which primary or secondary tension / shear cracks within the rock mass initiate, expand, and eventually become spatially connected under the influence of gravity, unloading, and water pressure. This instability mode, dominated by cracks, determines that its precursors are hidden, deformation monitoring is difficult, and it often manifests as a sudden collapse or fall. Because unstable rock masses are usually located on high, steep slopes, they have high potential energy and great destructive force. Once unstable, the collapse will rapidly disintegrate, affecting an area far beyond its own boundaries, easily causing the interruption of surrounding transportation arteries, damage to water conservancy facilities, and casualties.

[0003] Existing technologies for assessing and warning of rock instability risks have the following key challenges in practical engineering applications: First, existing methods mostly follow the analysis approach dominated by the landslide slip surface, and have not established a specific model for the instability characteristics of the gradual expansion of the crack surface of unstable rock, resulting in poor mechanism adaptability; Secondly, the computational models generally adopt static probability assessment based on initial physical parameters, lacking a dynamic update mechanism that takes into account parameter degradation, resulting in a serious disconnect between the computational results and the actual evolutionary state. Furthermore, the on-site multi-source monitoring data and the numerical simulation system are independent of each other, and the separation between monitoring and simulation makes it impossible to use measured information to trigger targeted corrections to the model and parameters; Finally, the lack of joint judgment rules that deeply integrate probability quantification indicators with monitored physical characteristics in the early warning process leads to a persistently high rate of false positives and false negatives.

[0004] Overall, existing patents and methods are mostly limited to landslide-type slip surface models, static parameter calculations, and single threshold early warnings. They have not yet formed a complete closed-loop system integrating discrete element modeling, dynamic probability distribution, hierarchical update mechanisms, and monitoring simulation. The engineering applicability and early warning accuracy urgently need to be improved. Summary of the Invention

[0005] This invention provides a method, device, and medium for joint early warning of rock instability risk, which solves the problems in the prior art such as the mismatch between mechanism modeling and rock cracking characteristics, the static nature of instability probability calculation and its separation from monitoring data, the high false alarm and missed alarm rates due to the single early warning judgment rule, and the lack of a complete closed-loop system. It is particularly suitable for long-term dynamic risk assessment and graded early warning of rock instability dominated by cracked surfaces.

[0006] In a first aspect, the present invention provides a joint early warning method for the risk of unstable rock mass, comprising: selecting several key parameters that have the greatest impact on the instability probability based on the discrete element numerical model of the unstable rock mass and the initial probability distribution of the rock mass's physical and mechanical parameters; establishing a dynamic probability distribution of the key parameters based on real-time monitoring data of the key parameters; updating the dynamic probability distribution and the discrete element numerical model in stages when the real-time monitoring data reaches a preset threshold; calculating the dynamic instability probability of the unstable rock mass based on the update results; and outputting early warning information based on the dynamic instability probability and the real-time monitoring data.

[0007] This invention solves the problem of incompatibility between existing technology mechanism modeling and the cracking characteristics of unstable rocks by constructing a dedicated discrete element numerical model for unstable rocks. By selecting key parameters and establishing a dynamic probability distribution using real-time monitoring data, it overcomes the shortcomings of traditional methods, such as static probability calculation and separation from monitoring data, enabling the instability probability to be dynamically updated with the actual evolution of the unstable rock. By setting a hierarchical update mechanism, the dynamic probability distribution and discrete element numerical model are adjusted hierarchically when the monitoring data reaches a preset threshold, achieving a balance between computational accuracy and efficiency. Finally, by integrating the dynamic instability probability and real-time monitoring data to jointly output early warning information, it changes the current situation where early warning judgment rules are singular and the false alarm and missed alarm rates are high.

[0008] This invention establishes a dynamic probability distribution by screening key parameters and utilizing real-time monitoring data. It uses monitoring data to trigger hierarchical updates to achieve coordinated adjustment of the model and parameters, and integrates dynamic instability probability with real-time monitoring data to jointly output early warning information. This solves the problems of static probability, fragmented monitoring, and single early warning in existing technologies, and effectively improves the accuracy, real-time performance, and engineering feasibility of early warning for rock instability risks.

[0009] In one optional implementation, the discrete element numerical model of the unstable rock mass is constructed as follows: a geological model of the unstable rock mass is constructed based on field survey data. The geological model includes the boundary of the unstable rock mass, the distribution of primary crack surfaces, the integrity zoning of the rock mass, and the groundwater seepage field. Based on the geological model, the unstable rock mass is divided into blocks according to the network formed by the primary crack surfaces. The Coulomb-slip contact and tensional fracture criteria are selected as the contact constitutive model. Self-weight load and groundwater pressure are applied, and the free surface is used as the free boundary to obtain the discrete element numerical model. By constructing a geological model based on field survey data, dividing the rock mass into blocks according to the network formed by the primary crack surfaces, and selecting the Coulomb-slip contact and tensional fracture criteria as the contact constitutive model, the established discrete element numerical model can realistically reflect the actual geometric shape of the unstable rock mass cut by multiple sets of cracks and the mechanical mechanism of cracking instability. This solves the problem that the existing technology uses the landslide slip surface analysis approach, which leads to the mismatch between mechanism modeling and the cracking characteristics of unstable rock. It provides an accurate and reliable computational carrier for subsequent key parameter selection, dynamic probability calculation, and hierarchical updating.

[0010] In one optional implementation, the initial probability distribution of rock mass physical and mechanical parameters is constructed as follows: mechanical tests are conducted on the unstable rock mass to obtain statistical samples of rock mass physical and mechanical parameters; distribution tests are performed on the statistical samples to determine the probability distribution type of each rock mass physical and mechanical parameter; based on the probability distribution type, the statistical characteristics of each rock mass physical and mechanical parameter are fitted to form the initial probability distribution. This method is based on measured data and supported by statistical theory, enabling the quantitative expression of parameter uncertainty. It overcomes the shortcomings of traditional methods where parameter values ​​depend on experience or a single definite value, providing a scientific and reliable prior benchmark for subsequent key parameter screening and the establishment of dynamic probability distributions.

[0011] In one optional implementation, several key parameters with the greatest impact on the instability probability are selected, including: determining the value range of each rock mass physical and mechanical parameter based on the initial probability distribution; sampling each parameter within the value range to generate multiple sets of parameter samples; substituting each set of parameter samples into a discrete element numerical model for simulation to calculate the corresponding instability probability; and selecting the parameters with the highest sensitivity as key parameters based on the sensitivity of each parameter to the instability probability. This implementation reduces all parameters to a few highly sensitive key parameters, reducing the simulation sample size and computation time while ensuring the accuracy of the instability probability calculation. It overcomes the problem of low efficiency in existing full-parameter modeling techniques and provides an efficient and engineering-applicable parameter simplification method for long-term dynamic risk assessment of unstable rock masses.

[0012] In one optional implementation, a dynamic probability distribution of the key parameters is established based on real-time monitoring data. This includes: periodically retesting the key parameters to obtain prior statistical features; acquiring real-time monitoring data of the key parameters and correcting the prior statistical features to obtain posterior statistical features of the key parameters; and generating a dynamic probability distribution based on the posterior statistical features and the probability distribution type of the key parameters. This implementation generates a dynamic probability distribution, enabling the parameter distribution to be continuously updated with the actual evolution of the unstable rock mass. This overcomes the shortcomings of existing technologies where probability calculations rely on static initial parameters and are disconnected from the actual evolution state, achieving dynamic synchronization between the instability probability and the actual deterioration process of the unstable rock mass.

[0013] In one optional implementation, when the real-time monitoring data reaches a preset threshold, the dynamic probability distribution and the discrete element numerical model are updated in a tiered manner, including: when the real-time monitoring data reaches a first preset threshold, only the dynamic probability distribution is updated; when the real-time monitoring data reaches a second preset threshold, both the dynamic probability distribution and the discrete element numerical model are updated synchronously; wherein the second preset threshold is higher than the first preset threshold. By setting the first and second preset thresholds, and triggering either a lightweight update that only updates the dynamic probability distribution or a heavy update that simultaneously updates both the dynamic probability distribution and the discrete element numerical model based on the different threshold levels reached by the real-time monitoring data, a tiered adaptation of computational resources and model accuracy driven by monitoring data is achieved. This maintains dynamic probability tracking with low computational cost during the slow evolution phase of the unstable rock, and ensures computational accuracy through model structure adjustments during the accelerated evolution phase. This solves the problems of computational lag or resource waste caused by the separation of monitoring and simulation and the single update mechanism in existing technologies.

[0014] In one optional implementation, the dynamic instability probability of the unstable rock is calculated based on the updated results. This includes: sampling the updated dynamic probability distribution to generate multiple sets of key parameter samples; substituting each set of key parameter samples into a discrete element numerical model for simulation, and counting the number of samples that experience instability; and obtaining the dynamic instability probability based on the ratio of the number of samples that experience instability to the total number of samples. This method of calculating the dynamic instability probability enables probabilistic instability risk assessment, allowing the calculation results to reflect the comprehensive impact of parameter uncertainty on the stability of the unstable rock. It overcomes the shortcomings of traditional deterministic analysis methods that cannot quantify instability risk, providing a scientific and quantitative basis for subsequent early warning level determination.

[0015] In one optional implementation, early warning information is output based on the dynamic instability probability and real-time monitoring data, including: determining a basic early warning level based on the dynamic instability probability and its changing characteristics; verifying or correcting the basic early warning level based on real-time monitoring data to obtain a final early warning level; and outputting early warning information containing the final early warning level. This method of determining the final early warning information achieves dual information fusion decision-making based on probabilistic quantitative trends and actual physical monitoring conditions. It overcomes the shortcomings of existing technologies where pure probability calculations are prone to missed detections due to model errors and pure monitoring indicators are prone to misjudgments due to qualitative ambiguity, thus reducing the false alarm rate and missed alarm rate of early warnings.

[0016] Secondly, this invention provides a joint early warning device for the risk of unstable rock mass, comprising: a key parameter screening module, used to screen several key parameters that have the greatest impact on the instability probability based on the discrete element numerical model of the unstable rock mass and the initial probability distribution of the rock mass's physical and mechanical parameters; a dynamic probability distribution establishment module, used to establish the dynamic probability distribution of the key parameters based on real-time monitoring data of the key parameters; a graded update triggering module, used to perform graded updates on the dynamic probability distribution and the discrete element numerical model when the real-time monitoring data reaches a preset threshold; a dynamic instability probability calculation module, used to calculate the dynamic instability probability of the unstable rock mass based on the update results; and a joint early warning output module, used to output early warning information based on the dynamic instability probability and real-time monitoring data.

[0017] Thirdly, the present invention provides a computer-readable storage medium storing computer instructions, which are used to cause a computer to execute the joint early warning method for rock instability risk described in the first aspect or any corresponding embodiment. Attached Figure Description

[0018] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the joint early warning method for rock instability risk according to an embodiment of the present invention; Figure 2 This is a schematic diagram of a three-dimensional geological model of a dangerous rock in a specific embodiment of the present invention; Figure 3 This is a schematic diagram of the discrete element numerical model in a specific embodiment of the present invention; Figure 4 This is a structural block diagram of the joint early warning device for rock instability risk according to an embodiment of the present invention; Figure 5 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0020] 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 embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0022] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0023] Existing technologies for calculating and warning of rockfall instability probability have the following core limitations, making it difficult to meet actual engineering needs: Static calculation of instability probability is out of touch with actual evolution: Existing technologies mostly use static probability models to calculate the instability probability once based on initial physical parameters, without considering the dynamic changes in parameters caused by the expansion of cracked surfaces of unstable rocks and the deterioration of rock masses; The modeling logic is confusing and the adaptability is poor: most technologies follow the analysis approach dominated by the landslide slip surface, and do not design a special numerical model for the cracking and instability characteristics of unstable rocks. In addition, there is the problem of "full parameter modeling", which results in huge computational load and low efficiency. The monitoring and simulation are disconnected, and there is no dynamic update mechanism: the field monitoring data (displacement, stress, acoustic rupture, etc.) are only used as independent early warning references and are not linked with numerical models and probability calculations. This makes it impossible to trigger targeted updates of the model / parameters, resulting in probability calculations lagging behind the actual evolution of the dangerous rock. The early warning mechanism is too simple and has a high rate of false positives / false negatives: existing early warnings are mostly simple superpositions of "instability probability values" or "monitoring indicator thresholds", without systematic joint judgment rules. Pure probability calculations are prone to false negatives due to model errors, and pure monitoring indicators are prone to false positives due to qualitative ambiguity. There is no complete closed-loop system: the probability calculation, model building, monitoring data, early warning and handling are independent of each other, and an integrated closed loop of "modeling-parameter-calculation-update-early warning" has not been formed. The technical features are scattered and the engineering implementation is poor.

[0024] Based on this, the present invention provides a joint early warning method, device, and medium for rock instability risk, to solve the problems in existing technologies such as incompatibility between mechanism modeling and rock cracking characteristics, static instability probability calculation separated from monitoring data, high false alarm and false negative rates due to single early warning judgment rules, and the lack of a complete closed-loop system. This invention belongs to the field of geotechnical engineering geological disaster prevention technology, and particularly relates to dynamic instability probability calculation and multi-source monitoring joint graded early warning for rockfalls dominated by cracked surfaces. This invention is applicable to long-term dynamic risk assessment and disaster early warning of rockfalls in high-risk areas such as highway slopes, railway slopes, reservoir banks, and urban peripheries.

[0025] According to an embodiment of the present invention, a method for joint early warning of rock instability risk is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0026] This embodiment provides a joint early warning method for the risk of rock instability. Figure 1 This is a flowchart of a joint early warning method for rock instability risk according to an embodiment of the present invention, such as... Figure 1 As shown, the process includes the following steps: Step S101: Based on the discrete element numerical model of the unstable rock mass and the initial probability distribution of the rock mass physical and mechanical parameters, select several key parameters that have the greatest impact on the instability probability.

[0027] The discrete element numerical model of the unstable rock mass is constructed as follows: a geological model of the unstable rock mass is constructed based on field survey data. The geological model includes the boundary of the unstable rock mass, the distribution of primary crack surfaces, the integrity zoning of the rock mass, and the groundwater seepage field. Based on the geological model, the unstable rock mass is divided into blocks according to the network formed by the primary crack surfaces. The Coulomb-slip contact and tension fracture criteria are selected as the contact constitutive model. The self-weight load and groundwater pressure are applied, and the free surface is used as the free boundary to obtain the discrete element numerical model.

[0028] Among them, the Coulomb-slip contact model is used to describe the shear slip behavior of the contact surface between blocks, and its shear strength is determined by the cohesion c and the internal friction angle φ; the tensile fracture criterion is used to describe the tensile failure behavior of the contact surface under tensile stress, and its failure threshold is determined by the contact tensile strength σ. t Confirmed. The instability criterion for unstable rock masses is set as follows: when the spatial penetration rate of the main controlling crack surface reaches or exceeds 80%, or when there is a sudden increase in the tensile displacement of the key block, the unstable rock mass is determined to have become unstable.

[0029] In one specific embodiment, the rock mass physical and mechanical parameters include: contact tensile strength σ t, Crack penetration rate Pc, Contact cohesion c, Internal friction angle φ, Groundwater pressure P w And the rock mass elastic modulus E. The above parameters were obtained as basic values ​​through indoor direct shear tests, tensile tests and creep tests, and were corrected through in-situ tests. The number of test samples for each parameter was no less than 30.

[0030] In one specific embodiment, the discrete element numerical model is constructed primarily using 3DEC software, followed by PFC3D software. Space is reserved for the initiation of secondary crack surfaces when dividing the blocks. The instability criterion is set as follows: the spatial continuity rate of the main controlling crack surface reaches or exceeds 80%, or a sudden increase in the tensile displacement of the key block occurs.

[0031] In a specific embodiment, the method for constructing a geological model of a dangerous rock mass is as follows: (1) Comprehensive on-site investigation The boundaries, distribution of primary crack surfaces, and orientation of structural planes of the unstable rock mass are determined through drilling or pitting; hidden crack surfaces are detected through geophysical methods such as seismic waves or ground-penetrating radar; the precise geometric shape of the unstable rock mass is obtained through three-dimensional laser scanning; and the groundwater seepage path is determined through hydrogeological exploration.

[0032] (2) Defining the dominant instability mode Based on the exploration results, it was determined that the unstable rock mass was dominated by tension or shear cracking, excluding the dominant mode of overall slip of the slip surface, and the core evolution path of cracking from initiation to expansion and final spatial connection was determined.

[0033] (3) Divide risk levels and rock mass zones Based on the size of the unstable rock mass and the importance of the disaster-bearing body, the initial risk level is divided into four levels: low, medium, high, and extremely high. According to the degree of rock mass integrity, the unstable rock mass is divided into intact zone, fractured zone, and cracked zone, which provides a basis for the differentiated parameter assignment of different regions in the subsequent discrete element numerical model.

[0034] (4) Constructing a three-dimensional geological model Based on the above exploration data, a three-dimensional geological model containing the unstable rock mass, parent rock, primary crack surface distribution, and groundwater seepage field is constructed in CAD or GIS software to accurately restore the spatial geometry of the unstable rock mass and the spatial distribution characteristics of the crack surface.

[0035] The method for constructing the initial probability distribution of rock mass physical and mechanical parameters is as follows: conduct mechanical tests on the unstable rock mass to obtain statistical samples of rock mass physical and mechanical parameters; perform distribution tests on the statistical samples to determine the probability distribution type of each rock mass physical and mechanical parameter; based on the probability distribution type, fit the statistical characteristics of each rock mass physical and mechanical parameter to form the initial probability distribution.

[0036] In a specific embodiment, the Kolmogorov-Smirnov test (KS test) is performed on the statistical samples of each parameter obtained from the above tests, and the probability distribution type of each parameter is determined based on the test results. In actual engineering, the physical and mechanical parameters of rock mass usually follow a log-normal distribution or a normal distribution. Among them, the contact tensile strength σt and cohesion c mostly follow a log-normal distribution, while the internal friction angle φ and the rock mass elastic modulus E mostly follow a normal distribution.

[0037] After determining the probability distribution type of each parameter, the sample data is fitted to obtain the statistical characteristic values ​​of each parameter distribution, including the mean μ0, standard deviation σ0, and coefficient of variation CV. The distribution type of each parameter and its corresponding statistical characteristic values ​​are summarized to form the initial probability distribution table of the core parameters.

[0038] Specifically, step S101 above includes: Step S1011: Determine the range of values ​​for each rock mass physical and mechanical parameter based on the initial probability distribution; Step S1012: Sample each parameter within the range of values ​​to generate multiple sets of parameter samples; Step S1013: Substitute each set of parameter samples into the discrete element numerical model for simulation and calculate the corresponding instability probability. Step S1014: Based on the sensitivity of each parameter to the instability probability, select the parameters with the highest sensitivity as key parameters.

[0039] In a specific embodiment, steps S1011 to S1014 above are implemented using the Sobol global sensitivity analysis method. The Sobol method can adapt to the nonlinear and strongly coupled characteristics of discrete element numerical models and can quantitatively calculate the sensitivity coefficients of each parameter to the instability probability. The specific calculation process is as follows: (1) Determining the parameter range Based on the initial probability distribution table of the core parameters, the value range of each parameter is defined according to the 5% to 95th percentile of the probability distribution. The target output is clearly defined as the probability of rockfall instability P. f .

[0040] (2) Parameter sampling and sample set construction Latin hypercube sampling was used to sample m core parameters, generating two independent basic sample matrices A and B, with each matrix containing at least 1000 samples N. For each parameter X to be analyzed... i Construct replacement sample matrix A i The construction method is as follows: replace the i-th column of the base sample matrix A with the i-th column of the base sample matrix B, while keeping the other columns unchanged. This ultimately yields m replacement sample matrices A1 to A2. m .

[0041] (3) Calculate the instability probability using batch simulation. Given the base sample matrices A and B, and m replacement sample matrices A1 to A2... m Substitute the parameters into the discrete element numerical model for batch simulation, calculate the probability of rockfall instability for each set of parameter samples, and obtain the basic output set Pf. A 、Pf B and replace output set Pf A1 To Pf Am .

[0042] (4) Sensitivity index calculation First, calculate the total variance V:

[0043] in, .

[0044] Then calculate the first-order sensitivity index S of each parameter. i :

[0045] Then calculate the overall sensitivity index S for each parameter. ti :

[0046] (5) Key parameter screening The sensitivity coefficient S uses the total sensitivity index S ti S for all parameters ti Normalization is performed so that the sum of the sensitivity coefficients of all parameters is 1. The parameters are then sorted from largest to smallest based on their sensitivity coefficient S, and the top few parameters with a cumulative sensitivity coefficient of 80% or higher are selected as key parameters. Typically, 3 to 5 key parameters are selected; common combinations include contact tension strength σ. t Crack penetration rate P c Contact cohesion c and groundwater pressure P w .

[0047] (6) Computational verification and iterative update The sensitivity analysis process described above can be automated using Python's SALib library combined with the secondary development interface of discrete element numerical software. When the sample size N reaches 1000 or more, if the fluctuation range of the total sensitivity index Sti does not exceed 5%, the calculation result is considered convergent.

[0048] In addition, the effectiveness of the screened key parameters is verified every 1 to 3 months, using on-site monitoring data or supplementary experimental data. If the actual sensitivity of a parameter deviates from the initial analysis results by more than 20%, the list of key parameters is adjusted, and corresponding parameters are added or replaced.

[0049] Step S102: Based on the real-time monitoring data of the key parameters, establish the dynamic probability distribution of the key parameters.

[0050] Specifically, step S102 above includes: Step S1021: Periodically retest the key parameters and fit the prior statistical features. Step S1022: Obtain real-time monitoring data of key parameters, correct prior statistical features, and obtain posterior statistical features of key parameters. Step S1023: Generate a dynamic probability distribution based on the probability distribution type of the posterior statistical features and key parameters.

[0051] It should be noted that the real-time monitoring data mentioned above is obtained by a multi-source monitoring system deployed on site. The specific deployment scheme and data preprocessing method of the monitoring system are described in detail in subsequent step S103.

[0052] In a specific embodiment, the specific implementation process of step S102 above is as follows: (1) A priori calibration Following a pre-set cycle (e.g., monthly or quarterly), on-site or indoor retesting is conducted on the selected key parameters to obtain parameter sample sets at different times t1, t2, t3, etc. Statistical analysis is performed on the sample sets at each time point to fit the prior statistical characteristics of each time point, including the mean μX(tk) and standard deviation σX(tk), where X represents a key parameter, t... k This represents the k-th retest time.

[0053] (2) Fitting of time series evolution function Based on prior statistical characteristics at multiple time points, an exponential decay function or a power function is used to fit the temporal evolution of the statistical characteristics of each key parameter, yielding the evolution functions of the mean μX(t) and standard deviation σX(t) over time. For example, the temporal evolution function for the contact tension strength σt can be expressed as: μσt(t)=μσt0·e^(-α·D(t)) Where μσt0 is the initial mean of the contact tension strength, α is the attenuation coefficient, and D(t) is the microseismic damage variable at time t.

[0054] (3) Bayesian real-time update Establish a correlation model between monitoring indicators and key parameters. For example, establish the correlation between acoustic emission energy and contact tensile strength σt. Using real-time monitoring data as likelihood information, the fitted prior statistical features are corrected in real time using Bayesian correction to obtain the posterior statistical features of each key parameter, including the posterior mean μX*(t) and the posterior standard deviation σX*(t).

[0055] The basic principle of Bayesian correction is to treat prior statistical features as prior distributions, transform real-time monitoring data into observational information of parameters through an association model as the likelihood function, calculate the posterior distribution, and use the statistical features of the posterior distribution as the corrected parameter estimates.

[0056] (4) Dynamic distribution generation Based on the obtained posterior statistical characteristics, and combined with the probability distribution type of each key parameter determined in step S101 (such as log-normal or normal distribution), the dynamic probability distribution function F of each key parameter is generated. X(x，t) This function describes the probability distribution of the key parameter X at any time t.

[0057] When there is a statistical correlation between multiple key parameters, a correlation coefficient matrix is ​​introduced to construct a joint dynamic probability distribution of multiple parameters in order to more accurately reflect the mutual influence between parameters.

[0058] (5) Continuous updating of dynamic probability distribution Steps (2) to (4) above are continuously executed during the long-term monitoring of the unstable rock mass. Whenever new real-time monitoring data is acquired, a Bayesian correction is triggered to update the posterior statistical features and regenerate the dynamic probability distribution function, so that the probability distribution of key parameters always evolves dynamically with the actual damage and deterioration process of the unstable rock mass.

[0059] In one optional implementation, after establishing the dynamic probability distribution of key parameters based on real-time monitoring data, the method further includes: constructing a physical dual-drive simulation system between the dynamic probability distribution and the discrete element numerical model, wherein the dynamic probability distribution drives the model simulation, and the results of the model simulation are used to correct the evolution function of the dynamic probability distribution, thereby achieving bidirectional coupling between parameter evolution and mechanical simulation. The specific construction method is as follows: (1) Coupling interface development By developing an interface using Python and Fish languages, real-time data interaction between dynamic probability parameters and discrete element numerical models can be achieved.

[0060] (2) Dual-drive mechanism The dual-drive mechanism includes: Driver end: The dynamic probability distribution of key parameters outputs random sampled values ​​according to the time step, driving the discrete element numerical model mechanical simulation; Feedback end: The crack spread rate and damage accumulation value obtained from the discrete element numerical model are used to feed back and correct the time-series evolution function of the dynamic probability distribution of key parameters.

[0061] (3) Model validation The model is inverted using initial on-site monitoring data (displacement or microseismic data) to ensure that the crack evolution trend is consistent with the actual situation, with an error of no more than 10%.

[0062] Step S103: When the real-time monitoring data reaches the preset threshold, the dynamic probability distribution and discrete element numerical model are updated in stages.

[0063] Specifically, step S103 above includes: Step S1031: When the real-time monitoring data reaches the preset first preset threshold, only the dynamic probability distribution is updated; Step S1032: When the real-time monitoring data reaches the second preset threshold, the dynamic probability distribution and the discrete element numerical model are updated synchronously; wherein, the second preset threshold is higher than the first preset threshold, the first preset threshold corresponds to the slow evolution state of the unstable rock mass, and the second preset threshold corresponds to the accelerated evolution state of the unstable rock mass.

[0064] In a specific embodiment, the above-mentioned hierarchical update process includes three stages: deployment of the multi-source monitoring system and data preprocessing, setting of hierarchical update trigger thresholds, and implementation of hierarchical updates.

[0065] (1) Deployment of multi-source monitoring system and data preprocessing The multi-source monitoring system deploys various sensors for unstable rock masses, and the monitoring indicators include: tension displacement or crack opening degree, shear stress or tangential displacement, acoustic rupture or micro-seismic activity (including the number of events, energy and source location information), expansion rate of the main control structural surface and groundwater pressure.

[0066] The monitoring points are deployed focusing on the main crack surfaces and stress concentration areas to ensure the capture of key deformation and rupture signals that control rock instability. The sampling frequency is dynamically adjusted according to the risk level: data is collected every 10 minutes under normal conditions, and increased to once per second under high-risk conditions.

[0067] For the collected raw monitoring data, wavelet transform was first used for noise reduction; then, the 3σ criterion was used to identify and remove outliers; for a small number of missing data points, linear interpolation was used to complete them. After preprocessing, core feature parameters, including displacement velocity, displacement acceleration, microseismic event frequency, and microseismic energy, were extracted from the data as the basis for subsequent graded update judgments.

[0068] (2) Setting the trigger threshold for hierarchical updates Based on the different evolution stages of the unstable rock mass, two levels of trigger thresholds are set: 1. Lightweight update threshold, i.e., the first preset threshold. The lightweight update threshold corresponds to the slow evolution stage of the unstable rock mass. The triggering conditions include: the tension displacement velocity is between 0.1 mm and 0.5 mm per day, or the number of acoustic rupture events is between 5 and 20 per day, or the monthly increase in the expansion rate of the main control structure surface does not exceed 1%.

[0069] 2. Weight update threshold, i.e., the second preset threshold The weight update threshold corresponds to the accelerated evolution stage of the unstable rock mass. The triggering conditions include: the tension displacement velocity reaches or exceeds 0.5 mm per day, or the number of acoustic rupture events reaches or exceeds 20 times per day, or the monthly increase in the expansion rate of the main control structural surface reaches or exceeds 1%, or the sudden increase in shear stress reaches or exceeds 0.05 MPa.

[0070] (3) Implementation of graded updates Based on the threshold levels reached by the real-time monitoring data, perform update operations at the corresponding levels: 1. Lightweight update When the real-time monitoring data reaches the lightweight update threshold, only the posterior statistical characteristics of the dynamic probability distribution of key parameters are updated, without changing the geometric shape and contact constitutive parameters of the discrete element numerical model.

[0071] 2. Weight Update When the real-time monitoring data reaches the weight update threshold, the following two updates are performed: First, model update: the geometry of the main control crack surface, the crack surface penetration rate and the block division method are corrected according to the measured data, and the contact constitutive parameters and damage field are updated; Second, parameter update: the dynamic probability distribution of key parameters is corrected synchronously.

[0072] Step S104: Based on the updated results, calculate the dynamic instability probability of the unstable rock.

[0073] Specifically, step S104 above includes: Step S1041: Sample the updated dynamic probability distribution to generate multiple sets of key parameter samples; Step S1042: Substitute each set of key parameter samples into the discrete element numerical model for simulation and count the number of samples that become unstable. Step S1043: Based on the ratio of the number of samples that have become unstable to the total number of samples, obtain the dynamic instability probability.

[0074] In a specific embodiment, the above steps S1041 to S1043 are implemented as follows: (1) Parameter sampling The updated dynamic probability distribution of key parameters is sampled using the Latin hypercube sampling method. The number of sampling iterations N is set to 300 to 500 based on the balance between computational accuracy and efficiency. Each sampling generates a set of key parameter samples, resulting in N sets of key parameter samples in total.

[0075] (2) Batch simulation The N sets of key parameter samples generated by sampling are substituted into the discrete element numerical model for mechanical simulation to determine whether the unstable rock mass will become unstable under each set of parameter conditions, and the number of samples that become unstable, Nf, is counted.

[0076] Since a single discrete element simulation involves a large amount of computation, stacking N simulations will result in significant computational time consumption. To improve simulation efficiency while ensuring computational accuracy, this embodiment adopts a three-level acceleration strategy of "model simplification + parallel computing + result reuse" to control the overall error of batch simulations within 5%. The specific details of the three-level acceleration strategy are as follows: 1. Model simplification and optimization First, adaptive simplification of mesh or particles. For the deep part of the parent rock and the intact area far from the main control cracking surface, coarse mesh or large particles are used for subdivision, with the size being 2 to 3 times that of the mesh or particle size of the main control cracking area; fine mesh or small particles are retained for the main control cracking surface and stress concentration area to ensure the simulation accuracy of key areas.

[0077] Secondly, minor parameters are fixed. For non-critical parameters with a cumulative sensitivity of less than 20% in step S101, the mean of their initial probability distribution is directly used for fixed assignment, and only the critical parameters are used as variables in sampling and simulation.

[0078] Third, the simulation termination conditions were optimized. During a single simulation, when the crack spread rate is below 10% and tends to stabilize, the sample is determined not to be unstable, and the simulation is terminated early; when the crack spread rate reaches or exceeds 80%, the sample is determined to be unstable, and the simulation is also terminated early.

[0079] 2. Computational Parallelization Firstly, parallel sample grouping. N samples are evenly grouped according to the number of CPU or GPU cores, and the Python multiprocessing library or the parallel module built into the discrete element numerical software is used to achieve simultaneous simulation calculations of multiple groups of samples.

[0080] Secondly, GPU acceleration. For discrete element numerical software that supports GPU acceleration, core computational tasks such as contact force solving and displacement iteration are assigned to the GPU for execution; for software that does not support GPU acceleration, GPU computing power is invoked through the CUDA interface to assist in computation.

[0081] 3. Result reuse mechanism Establish a cache library of "parameter range - simulation results". For newly extracted parameter samples, first determine whether they fall within a previously calculated parameter range (with an error not exceeding 3%). If they do, directly reuse the corresponding historical simulation results in the cache library without repeating the complete simulation.

[0082] 4. Implementation tools The above-mentioned sampling parameter substitution, acceleration strategy invocation, and simulation result acquisition are all automated through scripts written in the Fish language built into the discrete element numerical software.

[0083] (3) Calculation of dynamic instability probability After completing the batch simulation, count the number N of samples that became unstable. f Calculate the current time t using the following formula. n The probability of dynamic instability P f(tn) : P f(tn) =N f / N Furthermore, the rate of change v of the dynamic instability probability is calculated. f(tn) and the change in acceleration a f(tn) : v f(tn) =[P f(tn) -P f(tn-1) ] / (t n-tn-1 ) a f(tn) =[v f(tn) -vf (tn-1) ] / (t n-tn-1 ) Among them, P f(tn-1) Let t be the dynamic instability probability at the previous calculation time. n -t n-1 The time interval between the two calculations.

[0084] (4) Time-series update Each time step S103 triggers a graded update, the parameter sampling, batch simulation, and probability calculation are re-executed according to the above process to obtain the dynamic instability probability, its rate of change, and its acceleration at the current moment. The calculation results at each moment are recorded sequentially to form the dynamic instability probability P. f(t) , rate of change v f(t) The time-series evolution curve of the changing acceleration af(t) provides a continuous and quantitative decision-making basis for the joint early warning output in the subsequent step S105.

[0085] Step S105: Output early warning information based on the dynamic instability probability and real-time monitoring data.

[0086] Specifically, step S105 above includes: Step S1051: Determine the basic early warning level based on the dynamic instability probability and its changing characteristics; Step S1052: Verify or correct the basic warning level based on real-time monitoring data to obtain the final warning level; Step S1053: Output the warning information containing the final warning level.

[0087] In a specific embodiment, the joint early warning determination process of steps S1051 to S1053 is as follows: (1) Joint early warning judgment rules The early warning levels for unstable rock masses are divided into four levels, from low to high: no warning, blue warning (trend warning), yellow warning (acceleration warning), and red warning (impending disaster warning).

[0088] First, the dynamic instability probability P calculated in step S104 is... f(t) and its rate of change v f(t) , change in acceleration a f(t) Basic early warning levels are determined. Probabilistic indicators reflect the quantitative risk level of unstable rock masses and their evolution trend.

[0089] Secondly, real-time monitoring data obtained from a multi-source monitoring system is used to verify the basic early warning levels determined based on probability. For blue and yellow alerts, at least one or two cracking-specific monitoring indicators must match the threshold range of the corresponding level to confirm the warning level; if they do not match, the basic early warning level is adjusted accordingly.

[0090] Finally, when real-time monitoring data shows obvious signs of impending disaster, regardless of the current dynamic instability probability calculation results, the warning level will be directly raised to a higher level to ensure timely response to sudden instability events.

[0091] The specific criteria for determining a Level IV warning are shown in Table 1 below. Each threshold can be adjusted appropriately based on the initial risk level of the unstable rock mass. Table 1. Specific Judgment Criteria for Level IV Warning

[0092] (2) Real-time judgment and verification of early warning The early warning determination is carried out by combining automated initial judgment with manual review.

[0093] Automated initial assessment: The early warning system automatically compares the current dynamic instability probability and real-time monitoring data with the above four-level judgment criteria according to the probability calculation update step (usually once per hour or once per day) to give an initial early warning level.

[0094] Manual verification: For cases automatically classified as blue or yellow alerts, professional technicians will conduct on-site inspections and verifications, focusing on the actual condition of the main cracked surface and the loosening of unstable rock masses. The alert level will be confirmed or adjusted based on the on-site verification results. For cases automatically classified as red alerts, no manual verification is required; the emergency alert response will be triggered directly.

[0095] (3) Tiered handling and coordination Based on the final confirmed warning level, corresponding tiered response measures will be initiated: When there is no alarm, maintain routine monitoring and collect data at the established frequency; conduct on-site inspections once a quarter; and conduct periodic retests once a year.

[0096] When the blue alert status is active, the monitoring frequency will be increased to once every 10 minutes; the periodic retesting cycle of key parameters will be shortened to once a month; and a field inspection will be carried out once a week to closely monitor changes in the cracked surface.

[0097] When the yellow alert status is activated, the emergency monitoring mode is initiated, and the monitoring data transmission frequency is increased to the second level; a warning area is delineated within the influence range of the unstable rock mass, and warning signs are set up; an emergency reinforcement plan is prepared, and grouting materials and anchor bolts and other reinforcement materials are prepared; a risk monitoring report is prepared and submitted daily.

[0098] When a red alert is issued, immediately organize the emergency evacuation of personnel and equipment from the danger zone; implement closed management of the danger zone; quickly implement emergency reinforcement measures such as rapid grouting or temporary anchor bolts; and simultaneously notify relevant local departments to activate the joint emergency response mechanism.

[0099] (4) Optimization of early warning downgrading, cancellation and closed-loop management When the warning indicators fall back to the threshold range corresponding to the next lower warning level, and this state remains stable for more than a month, the warning level can be lowered by one level after on-site verification. When all indicators fall back to the no-alarm threshold range and remain stable for more than a month, the warning can be lifted after verification.

[0100] After each early warning event, a post-mortem analysis is conducted on the probability calculation deviations, monitoring indicator response, and accuracy of early warning level determination throughout the entire early warning process. Based on the analysis results, the correlation model between monitoring indicators and key parameters, the evolution function of dynamic probability distribution, and the early warning thresholds at each level are optimized and adjusted to continuously improve the accuracy and reliability of the early warning system.

[0101] The joint early warning method for rock instability risk provided in this embodiment has the following beneficial effects: In terms of calculation accuracy, the deviation between the dynamic instability probability and the actual instability risk can be controlled within 10%, which is far better than the deviation of more than 30% commonly found in existing technologies; In terms of calculation efficiency, dynamic modeling is performed only for 3 to 5 key parameters, and combined with a hierarchical update mechanism and batch simulation acceleration strategy, the calculation efficiency is significantly improved compared with the traditional full-parameter modeling method; In terms of early warning accuracy, the joint judgment rule reduces the false judgment rate and the missed judgment rate by more than 80%; In terms of engineering adaptability, it is designed specifically for rock instability types dominated by cracked surfaces, which is different from the traditional landslide analysis approach and fills the technical gap in the dynamic prevention and control of this type of rock instability; In terms of practicality, each step has clear exploration methods, equipment selection, calculation methods and quantitative thresholds, which can be directly implemented by technical personnel in the relevant technical field and has broad industrial application value.

[0102] The following detailed explanation of the implementation steps, process, and effects of the present invention will be provided using a specific example of the application of early warning technology for unstable rockfalls caused by cracking surfaces on a highway slope. It should be noted that this embodiment is for illustrative purposes only and does not constitute a limitation on the scope of protection of the present invention.

[0103] (I) Project Background The unstable rock mass involved in this embodiment is located on a highway slope with a total volume of approximately 8,000 cubic meters. Below it is a densely populated rural residential area, and the initial risk level is assessed as "high risk".

[0104] Based on a comprehensive on-site investigation, the unstable rock mass is a typical example of a rock mass with predominantly tensile-shear fracture surfaces. Three primary fracture surfaces are present within the rock mass, with a clear spatial distribution and well-defined potential pathways. The area where the rock mass is located has well-developed groundwater, which adversely affects the stability of the rock mass.

[0105] (II) Implementation Steps According to the technical solution of this invention, dynamic instability probability calculation and joint early warning are performed on the unstable rock mass. The specific implementation process is as follows: 1. Geological modeling During the on-site investigation phase, three boreholes and two exploration pits were laid out to clarify the spatial boundary of the unstable rock mass. Three-dimensional laser scanning technology was used to obtain precise geometric data of the unstable rock mass. Through ground-penetrating radar detection, one new concealed crack surface was discovered in addition to the three known primary crack surfaces, resulting in the identification of a total of four crack surfaces.

[0106] Based on the above exploration data, a three-dimensional geological model was constructed, including the unstable rock mass, parent rock, four crack surfaces, and groundwater seepage field. (See attached image.) Figure 2 It accurately reproduces the spatial geometry and crack surface distribution characteristics of the unstable rock mass.

[0107] 2. Construction of Discrete Element Numerical Model and Parameter Probability Distribution A numerical calculation model of the unstable rock mass was constructed using the 3DEC discrete element numerical software. (See [link to documentation]). Figure 3 Based on the spatial network of crack surfaces in the geological model, the unstable rock mass is divided into 1200 block units. The contact constitutive model adopts the "Coulomb-slip contact + tension fracture criterion".

[0108] The initial statistical characteristics of each physical and mechanical parameter were obtained through indoor mechanical tests, as follows: contact tensile strength σ t The mean value of the cohesion c is 3.2 MPa, which follows a log-normal distribution with a coefficient of variation (CV) of 0.2; the mean value of the cohesion c is 0.8 MPa, which follows a normal distribution with a coefficient of variation (CV) of 0.15; the crack penetration rate P c The initial value is 35%, follows a normal distribution, and has a coefficient of variation (CV) of 0.1; groundwater pressure P w The mean value is 0.3 MPa, and it follows a uniform distribution. The above parameters together constitute the initial probability distribution of the core parameters.

[0109] 3. Dynamic probability distribution of key parameters and construction of the dual-drive system Based on the discrete element numerical model, the Sobol global sensitivity analysis method is used to perform sensitivity analysis on the aforementioned core parameters. The calculated sensitivity coefficients S for each parameter are as follows: contact tension strength σ... t 35%, crack penetration rate P c The cohesion c is 18%, and the groundwater pressure P is 28%. w The sensitivity of the four parameters was 12%, and the cumulative sensitivity reached 93%, thus they were identified as key parameters.

[0110] For the selected key parameters, their time-series evolution functions are established. Taking the contact tension strength σ as an example... t For example, the evolution function of its mean over time is fitted as μσ t (t)=3.2·e^(-0.002·D(t)), where D(t) is the microseismic damage variable at time t.

[0111] By developing a coupling interface between Python and Fish languages, a physical dual-drive simulation system is constructed between discrete element numerical models and dynamic probability distributions, realizing bidirectional coupling between parameter evolution and mechanical simulation.

[0112] 4. Monitoring triggers updates and calculates the probability of dynamic instability. A multi-source monitoring system was deployed on-site, including 3 tension displacement gauges, 2 stress gauges, 4 microseismic sensors, and 1 piezometer. Each monitoring point was focused on the main crack surface and stress concentration area, with a sampling frequency of once every 10 minutes.

[0113] Set graded update trigger thresholds: the light update threshold is when the tension displacement speed is between 0.1 mm and 0.5 mm per day; the heavy update threshold is when the tension displacement speed reaches or exceeds 0.5 mm per day.

[0114] During the six months of system operation, real-time monitoring data triggered four lightweight updates and two heavyweight updates. After each update, a three-tiered acceleration strategy of "model simplification + parallel computing + result reuse" was employed to conduct batch simulations. In this embodiment, a 16-core CPU was used for parallel computation of sample groups, and a GPU was used to accelerate the core solution task.

[0115] Based on rolling calculations, the dynamic instability probability P of the unstable rock mass is... f(t) It gradually increased from 0.8% at the initial moment to 7.2% in the 6th month. In the 6th month, the rate of change v... f(t) The daily change in acceleration a is 0.035%. f(t) It is 0.0005% per day squared.

[0116] 5. Joint early warning and response In the 6th month, the probability of dynamic instability P f(t) The probability of reaching 7.2% falls within the probability threshold range (5% to 10%) corresponding to a yellow alert. Simultaneously, real-time monitoring data shows that the tension displacement velocity is 0.6 mm per day and the number of micro-seismic events is 35 per day. Both monitoring indicators meet the matching requirements for a yellow alert, and the system officially issues a yellow alert.

[0117] After the warning is issued, immediately delineate the warning area within the affected area of ​​the unstable rock mass and set up warning signs; prepare an emergency plan for grouting reinforcement and prepare grouting materials and related equipment; and prepare and submit a risk monitoring report daily.

[0118] One month after the warning was issued, the stability of the unstable rock mass was effectively improved after grouting reinforcement. The dynamic instability probability P... f(t) The percentage dropped from 7.2% before reinforcement to 3.8%, falling back to the probability threshold range corresponding to the blue alert. At the same time, monitoring indicators such as tension displacement velocity and the number of microseismic events also recovered to the blue alert level, downgrading the warning level from yellow alert to blue alert.

[0119] (III) Implementation Results In this embodiment, after applying the technical solution of the present invention, the calculation deviation of the dynamic instability probability of the unstable rock mass is controlled within 8.5%, the early warning response time does not exceed 30 minutes, and no misjudgment or missed judgment occurs during the entire monitoring and early warning cycle.

[0120] Compared with traditional methods, this invention effectively improves computational efficiency and significantly enhances early warning accuracy. The successful application of this embodiment effectively safeguards the lives and property of people in surrounding villages and towns, fully verifying the feasibility, effectiveness, and engineering practical value of the technical solution of this invention.

[0121] This embodiment also provides a joint early warning device for the risk of unstable rock formations. This device is used to implement the above embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0122] This embodiment provides a joint early warning device for the risk of unstable rock formations, such as... Figure 4 As shown, it includes: The key parameter screening module 401 is used to screen several key parameters that have the greatest impact on the instability probability based on the discrete element numerical model of the unstable rock mass and the initial probability distribution of the rock mass physical and mechanical parameters. The dynamic probability distribution establishment module 402 is used to establish the dynamic probability distribution of key parameters based on real-time monitoring data of key parameters. The hierarchical update trigger module 403 is used to perform hierarchical updates on the dynamic probability distribution and discrete element numerical model when the real-time monitoring data reaches a preset threshold. The dynamic instability probability calculation module 404 is used to calculate the dynamic instability probability of the unstable rock based on the updated results. The joint early warning output module 405 is used to output early warning information based on the dynamic instability probability and real-time monitoring data.

[0123] The joint early warning device for rock instability risk provided in this embodiment of the invention can execute the joint early warning method for rock instability risk provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method. Further functional descriptions of the above modules and units are the same as in the corresponding embodiments described above, and will not be repeated here.

[0124] Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0125] The following is a detailed reference. Figure 5The diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 501, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 502 or a program loaded from memory 508 into random access memory (RAM) 503. The RAM 503 also stores various programs and data required for the operation of the electronic device. The processor 501, ROM 502, and RAM 503 are interconnected via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0126] Typically, the following devices can be connected to I / O interface 505: input devices 506 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 507 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 508 including, for example, magnetic tapes, hard disks, etc.; and communication devices 509. Communication device 509 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 5 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0127] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 509, or installed from a memory 508, or installed from a ROM 502. When the computer program is executed by the processor 501, it performs the functions defined in the joint early warning method for rock instability risk according to embodiments of the present invention.

[0128] Figure 5 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of the present invention.

[0129] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the joint early warning method for rock instability risk shown in the above embodiments is implemented.

[0130] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0131] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A joint early warning method for the risk of unstable rock formations, characterized in that, The method includes: Based on the discrete element numerical model of the unstable rock mass and the initial probability distribution of the rock mass physical and mechanical parameters, several key parameters that have the greatest impact on the instability probability are selected. Based on the real-time monitoring data of the key parameters, establish the dynamic probability distribution of the key parameters; When the real-time monitoring data reaches a preset threshold, the dynamic probability distribution and the discrete element numerical model are updated in a hierarchical manner. Based on the updated results, the dynamic instability probability of the unstable rock is calculated; Based on the dynamic instability probability and the real-time monitoring data, an early warning message is output.

2. The joint early warning method for rock instability risk according to claim 1, characterized in that, The discrete element numerical model of the unstable rock mass is constructed as follows: A geological model of the unstable rock mass is constructed based on field survey data. The geological model includes the boundary of the unstable rock mass, the distribution of the original crack surface, the rock mass integrity zoning, and the groundwater seepage field. Based on the geological model, the unstable rock mass is divided into blocks according to the network formed by the original crack surfaces. The Coulomb-slip contact and tension fracture criteria are selected as the contact constitutive model. Self-weight load and groundwater pressure are applied, and the free surface is used as the free boundary to obtain the discrete element numerical model.

3. The joint early warning method for rock instability risk according to claim 1, characterized in that, The method for constructing the initial probability distribution of the rock mass physical and mechanical parameters is as follows: Mechanical tests were conducted on the unstable rock mass to obtain statistical samples of its physical and mechanical parameters. The statistical samples were subjected to distribution tests to determine the probability distribution type of each rock mass physical and mechanical parameter; Based on the probability distribution type, the statistical characteristics of the physical and mechanical parameters of each rock mass are fitted to form the initial probability distribution.

4. The joint early warning method for rock instability risk according to claim 1, characterized in that, The screening process focuses on several key parameters that have the greatest impact on the probability of instability, including: The range of values ​​for each rock mass physical and mechanical parameter is determined based on the initial probability distribution. Within the range of values, each parameter is sampled to generate multiple sets of parameter samples; Substitute each set of parameter samples into the discrete element numerical model for simulation and calculate the corresponding instability probability. Based on the sensitivity of each parameter to the instability probability, several parameters with the highest sensitivity are selected as the key parameters.

5. The joint early warning method for rock instability risk according to claim 1, characterized in that, The step of establishing a dynamic probability distribution of the key parameters based on real-time monitoring data of the key parameters includes: The key parameters are periodically retested, and prior statistical features are obtained by fitting. Real-time monitoring data of the key parameters are obtained, and the prior statistical features are corrected to obtain the posterior statistical features of the key parameters. The dynamic probability distribution is generated based on the posterior statistical features and the probability distribution type of the key parameters.

6. The joint early warning method for rock instability risk according to claim 1, characterized in that, When the real-time monitoring data reaches a preset threshold, the dynamic probability distribution and the discrete element numerical model are updated in a hierarchical manner, including: When the real-time monitoring data reaches a preset first threshold, only the dynamic probability distribution is updated; When the real-time monitoring data reaches a preset second preset threshold, the dynamic probability distribution and the discrete element numerical model are updated synchronously. The second preset threshold is higher than the first preset threshold.

7. The joint early warning method for rock instability risk according to claim 1, characterized in that, The calculation of the dynamic instability probability of the unstable rock based on the updated results includes: The updated dynamic probability distribution is sampled to generate multiple sets of key parameter samples; Substitute each set of key parameter samples into the discrete element numerical model for simulation, and count the number of samples that become unstable. The dynamic instability probability is obtained based on the ratio of the number of samples that experienced instability to the total number of samples.

8. The joint early warning method for rock instability risk according to claim 1, characterized in that, The step of outputting early warning information based on the dynamic instability probability and the real-time monitoring data includes: The basic early warning level is determined based on the dynamic instability probability and its changing characteristics; The basic warning level is verified or corrected based on the real-time monitoring data to obtain the final warning level; The output includes warning information containing the final warning level.

9. A joint early warning device for the risk of unstable rock formations, characterized in that, The device includes: The key parameter screening module is used to screen several key parameters that have the greatest impact on the instability probability based on the discrete element numerical model of the unstable rock mass and the initial probability distribution of the rock mass physical and mechanical parameters. The dynamic probability distribution establishment module is used to establish the dynamic probability distribution of the key parameters based on the real-time monitoring data of the key parameters. The hierarchical update triggering module is used to perform hierarchical updates on the dynamic probability distribution and the discrete element numerical model when the real-time monitoring data reaches a preset threshold. The dynamic instability probability calculation module is used to calculate the dynamic instability probability of unstable rocks based on the updated results. The joint early warning output module is used to output early warning information based on the dynamic instability probability and the real-time monitoring data.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to execute the joint early warning method for rock instability risk as described in any one of claims 1 to 8.

Images