Intelligent perception monitoring and early warning method and system for open-pit mine blasting disturbance landslide

Abstract

The application discloses an open-pit mine blasting disturbance landslide intelligent perception monitoring and early warning method and system, and relates to the technical field of engineering monitoring. The method combines numerical simulation analysis and an intelligent perception network model to realize accurate quantification and intelligent early warning of the stability of a slope under the influence of blasting disturbance. Through discrete-continuous coupling numerical simulation technology, the stability of the slope is simulated before blasting construction, the safety factor of the slope is extracted, and a high-precision slope stability evaluation system is constructed. A fuzzy mathematics method is introduced to construct a stability membership function, multiple parameters obtained through numerical simulation calculation are mapped into fuzzy variables, and quantitative description of the stability grade of the slope under different blasting conditions is formed. Subsequently, an intelligent perception network model is used in combination with numerical simulation results and field monitoring data to train the stability membership function through improvement of physical intelligence, a slope stability prediction model considering multiple factors is established, and seamless connection from numerical analysis to intelligent early warning is realized.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of engineering monitoring, and particularly relates to an open-pit mine blasting disturbance landslide intelligent perception monitoring and early warning method and system. BACKGROUND

[0002] In recent years, with the expansion of open-pit mine mining scale, the increase of slope engineering construction intensity and the rapid expansion of infrastructure, slope stability has become a key issue in the field of geotechnical engineering safety management and geological disaster prevention and control. Among them, the strong disturbance produced in the blasting construction process of open-pit mine is particularly prominent. Blasting vibration is easy to cause internal structure loosening of rock mass, accelerate the expansion of joint fissure, and even cause slope instability or landslide, which brings serious challenges to construction safety and environmental protection. With the gradual development of mining into deeper, larger scale and higher intensity mining stage, the influence of blasting disturbance on slope stability shows high complexity and uncertainty, and the traditional method faces more severe test in landslide prediction and dynamic early warning.

[0003] The current mainstream landslide monitoring and early warning method is mostly based on displacement monitoring, stress analysis and numerical simulation technology, but its limitations are increasingly prominent. Specifically, the traditional monitoring scheme usually adopts point-like and single sensing measurement technology, which is difficult to effectively capture the transient response characteristics of the entire slope area under the action of blasting vibration, resulting in incomplete identification of potential instability area; static or semi-static numerical simulation method also cannot accurately restore the real dynamics behavior of rock mass structure under the action of blasting transient, which seriously affects the accuracy of stability evaluation. In addition, the traditional early warning method based on fixed threshold lacks environmental adaptability and parameter dynamic adjustment ability, which is difficult to adapt to complex factors such as blasting parameter change, lithology difference, geological condition fluctuation and climate factor, thereby reducing the reliability of early warning result and increasing the risk of false alarm or missed alarm. SUMMARY

[0004] In view of the deficiencies of the prior art, the present application provides an open-pit mine blasting disturbance landslide intelligent perception monitoring and early warning method and system to solve the problems that the prior art is difficult to obtain the slope state in time under the condition of blasting disturbance, difficult to accurately identify the potential instability position, and the monitoring and early warning do not form an integrated linkage mechanism.

[0005] The technical scheme of the present application is as follows:

[0006] On the one hand, the present application provides an open-pit mine blasting disturbance landslide intelligent perception monitoring and early warning method, which specifically comprises the following steps:

[0007] Obtaining mine slope geological survey data and historical case data of the area to be excavated;

[0008] According to the mine slope geological survey data of the area to be excavated, a three-dimensional slope stability numerical model of the area to be excavated is constructed;

[0009] The three-dimensional slope stability numerical model of the area to be excavated was simulated under different blasting slope stability influencing factors to clarify the influence law of blasting slope stability influencing factors on slope safety factor. Fuzzy mathematics theory was used to construct a membership function to quantitatively map the mapping relationship between blasting slope stability influencing factor conditions and slope safety factor. The blasting slope stability influencing factors include blasting parameters and rock physical and mechanical parameters.

[0010] A slope stability prediction model that comprehensively considers the influence of multiple factors is constructed based on the membership function, and trained using historical case data to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors; the slope stability prediction model that comprehensively considers the influence of multiple factors is used to predict the slope safety factor based on the input blasting parameters and rock physical and mechanical parameters.

[0011] By using a slope stability prediction model that comprehensively considers the influence of multiple factors after training, the real-time monitoring data during the actual blasting construction process is analyzed, the predicted slope safety factor is automatically output, the landslide risk level is obtained, and early warning information is generated in real time to optimize and adjust the blasting construction plan.

[0012] Furthermore, the geological survey data of the mine slope includes the rock mass structure characteristics, joint distribution, and rock physical and mechanical parameters of the area to be excavated; the historical case data includes a set of rock physical and mechanical parameters, blasting parameters, and their corresponding slope safety factors; the rock physical and mechanical parameters include rock weathering intensity, degree of fracture development, and shear strength; the blasting parameters include charge amount, hole mesh parameters, resistance line, and detonation sequence.

[0013] Furthermore, the obtained three-dimensional slope stability numerical model of the area to be excavated is simulated under different blasting slope stability influencing factors to clarify the influence law of blasting slope stability influencing factors on slope safety factor, and a membership function is established, specifically including:

[0014] A1: Construct the boundary conditions of the three-dimensional slope stability numerical model, and use discrete continuous coupled simulation analysis to calculate the slope safety factor of each region in the three-dimensional slope stability numerical model under different blasting parameter conditions, that is, the slope safety factor under different blasting slope stability influencing factors, and determine the influence law of blasting slope stability influencing factors on the slope safety factor.

[0015] A2: Based on the influence of blasting slope stability factors on slope safety factor, a membership function is constructed using fuzzy mathematics theory to quantitatively map the relationship between blasting slope stability factors and slope safety factor.

[0016] Furthermore, the construction of a slope stability prediction model based on the membership function that comprehensively considers the influence of multiple factors, and the training of this model using historical case data, to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors, specifically includes:

[0017] B1: Optimize the membership function by constructing a training dataset using historical case data;

[0018] Specifically: a membership function calibration method based on error minimization and parameter adjustment is adopted, while monotonicity and boundary constraints are applied during the optimization process;

[0019] B2: Based on the optimized membership function, construct a slope stability prediction model that comprehensively considers the influence of multiple factors;

[0020] The slope stability prediction model that comprehensively considers the influence of multiple factors includes an input layer, a fuzzy mapping layer, a membership-driven risk attention mechanism module, and an output layer.

[0021] The input layer is used to receive input blasting parameters and rock physical and mechanical parameters;

[0022] The fuzzy mapping layer is used to map blasting parameters and rock physical and mechanical parameters to corresponding membership values ​​under fuzzy rules according to the membership function, and to form a fuzzy evaluation vector by all membership values.

[0023] The membership-driven risk attention mechanism module takes as input a fuzzy evaluation vector and a feature vector composed of normalized blasting parameters and rock physical and mechanical parameters. The membership-driven risk attention mechanism then weights each feature to obtain a risk-weighted feature vector. Input to the output layer;

[0024] The output layer receives the risk-weighted feature vector output by the membership-driven risk attention mechanism module. By extracting high-order features through several layers of nonlinear sensing units, the predicted value of the slope safety factor under this blasting disturbance is finally obtained.

[0025] The calculation process of the membership-driven risk attention mechanism is as follows:

[0026] First, to incorporate each membership value in the fuzzy evaluation vector into the relevance calculation, based on membership... Define risk amplification factor :

[0027] (8);

[0028] in, For indexing, These are weight parameters;

[0029] Based on this, the risk amplification factor Embedding the self-attention score yields a risk relevance score after fusing membership:

[0030] (9);

[0031] in, For self-attention score; The risk relevance score after merging membership degrees;

[0032] Then, the risk relevance score is calculated by fusing the membership degrees of all features. Normalization is performed to obtain the risk attention weights for each feature. ;

[0033] Utilizing risk attention weights Numerical vectors for each feature We perform weighted aggregation to obtain the risk-weighted feature vector. ;

[0034] (10);

[0035] in, The number of features;

[0036] B3: Use the training dataset to train the slope stability prediction model that comprehensively considers the influence of multiple factors, and obtain the trained slope stability prediction model that comprehensively considers the influence of multiple factors.

[0037] Specifically, the residuals of the dynamic equilibrium equation, the generalized constitutive equation, and the wave propagation equation are introduced into the Loss function as follows:

[0038] (11);

[0039] in, For loss function, , , These are the weighting coefficients for the corresponding loss terms. For the first The measured output of each sample For the network's predicted output, The physical constraint sampling point number; , , These are the residuals obtained by substituting the model output into the dynamic equilibrium equation, the generalized constitutive equation, and the wave propagation equation, respectively. For network parameters, and These represent the number of data samples and the number of physical constraint sampling points, respectively.

[0040] Furthermore, the slope stability prediction model, which comprehensively considers the influence of multiple factors and has been trained, is used to analyze real-time monitoring data during actual blasting construction. The model automatically outputs the predicted slope safety factor, thereby obtaining the landslide risk level and generating real-time early warning information. This process optimizes and adjusts the blasting construction plan. Specifically, this includes:

[0041] C1: During actual blasting operations, real-time monitoring data of the slope is collected, including blasting parameters and rock physical and mechanical parameters;

[0042] C2: Using a slope stability prediction model that has been trained and comprehensively considers the influence of multiple factors, the predicted slope safety factor is output in real time, and landslide risk level classification and early warning information are automatically generated based on the slope safety factor.

[0043] C3: Dynamically optimize and adjust the blasting construction plan based on the landslide risk level classification and early warning information.

[0044] On the other hand, the present invention also provides an intelligent sensing, monitoring and early warning system for landslides caused by blasting in open-pit mines, used to implement an intelligent sensing, monitoring and early warning method for landslides caused by blasting in open-pit mines, including:

[0045] The data acquisition module is used to acquire geological survey data and historical case data of the mine slope in the area to be excavated;

[0046] The model building module is used to construct a three-dimensional numerical model of slope stability in the area to be excavated based on the geological survey data of the mine slope in the area to be excavated.

[0047] The membership function construction module is used to simulate the three-dimensional slope stability numerical model of the area to be excavated under different blasting slope stability influencing factors, clarify the influence law of blasting slope stability influencing factors on slope safety factor, and establish membership function.

[0048] The model building and training module is used to build a slope stability prediction model that comprehensively considers the influence of multiple factors based on the membership function, and to train it using historical case data to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors.

[0049] The risk level prediction module is used to analyze real-time monitoring data during actual blasting construction using a slope stability prediction model that has been trained and comprehensively considers the influence of multiple factors. It automatically outputs the predicted slope safety factor, thereby obtaining the landslide risk level and generating early warning information in real time, and optimizing and adjusting the blasting construction plan.

[0050] Thirdly, this application proposes an electronic device, comprising: one or more processors, and a memory, wherein the memory is used to store instructions, and when the instructions are executed by the one or more processors, the one or more processors execute the aforementioned intelligent sensing, monitoring and early warning method for landslides caused by blasting in open-pit mines.

[0051] Fourthly, this application proposes a computer-readable storage medium storing executable instructions that, when executed, cause a processor to perform the aforementioned intelligent sensing, monitoring, and early warning method for landslides caused by blasting disturbance in open-pit mines.

[0052] Fifthly, this application proposes a computer program product, including a computer program or instructions, which, when executed by a processor, implements the aforementioned intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines.

[0053] Compared with the prior art, the advantages of this invention are as follows:

[0054] This innovative method combines numerical simulation analysis with an intelligent sensing network model, overcoming the limitations of traditional methods that rely solely on on-site monitoring data. It achieves precise quantification and intelligent early warning of slope stability under the influence of blasting disturbances. Through discrete-continuous coupled numerical simulation technology, slope stability is simulated before blasting, extracting the slope safety factor and constructing a high-precision slope stability evaluation system. Based on this, fuzzy mathematics is introduced to construct a stability membership function, mapping the multiple parameters obtained from numerical simulation calculations into fuzzy variables, forming a quantitative description of slope stability levels under different blasting conditions. Subsequently, an intelligent sensing network model is used, combining numerical simulation results with on-site monitoring data. By improving the physical intelligent sensing network to train the stability membership function, a slope stability prediction model that comprehensively considers the influence of multiple factors is established, achieving seamless integration from numerical analysis to intelligent early warning. This method can dynamically adjust blasting construction parameters and optimize excavation schemes in real time based on numerical simulation prediction results, thereby effectively reducing landslide risks and improving the safety and economic benefits of mining. It is applicable to open-pit mining engineering. By integrating numerical simulation with intelligent sensing networks, this invention achieves efficient and intelligent prediction of slope stability in complex blasting environments, providing a new generation of intelligent solutions for safety management in mines and rock engineering. Attached Figure Description

[0055] Figure 1 This is a flowchart of the intelligent sensing, monitoring and early warning method for landslides caused by blasting in open-pit mines, as described in this embodiment of the invention.

[0056] Figure 2 This is a diagram of the numerical calculation model in an embodiment of the present invention;

[0057] Figure 3This is a diagram illustrating the factors influencing the stability of a blasted slope in an embodiment of the present invention.

[0058] Figure 4 This is a framework diagram of the improved PINN-FNN intelligent sensing network (a slope stability prediction model that comprehensively considers the influence of multiple factors) in an embodiment of the present invention.

[0059] Figure 5 This is a monitoring and early warning curve diagram derived from the slope safety factor in an embodiment of the present invention. Detailed Implementation

[0060] The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0061] Example 1:

[0062] Due to the limitations of existing technologies, there is an urgent need to develop a novel landslide monitoring and early warning system that integrates intelligent sensing technology, multi-parameter simulation of blasting disturbances, and real-time intelligent risk level identification. This invention addresses this key technological bottleneck by innovatively proposing an intelligent sensing and identification method based on the fusion of fuzzy mathematics theory and physical neural networks. By constructing an intelligent sensing network model driven by slope stability membership functions and physical field data, it achieves high-precision quantitative assessment and dynamic intelligent early warning of slope stability. This method significantly enhances the real-time perception, accurate prediction, and adaptive response capabilities for slope instability risks under complex blasting disturbance environments, effectively improving the safety assurance level and disaster prevention capabilities during open-pit mining construction.

[0063] like Figure 1 As shown in the figure, this invention provides an intelligent sensing, monitoring, and early warning method for landslides caused by blasting disturbance in open-pit mines, specifically including the following steps:

[0064] S1: Obtain geological survey data and historical case data of the mine slope in the area to be excavated; the geological survey data of the mine slope includes the rock mass structure characteristics, joint (fracture) distribution and rock physical and mechanical parameters of the area to be excavated; the historical case data includes a set of rock physical and mechanical parameters, blasting parameters and their corresponding slope safety factors;

[0065] The rock physical and mechanical parameters used in this embodiment were selected based on collected historical case data, with parameters that have a significant impact on the slope safety factor being chosen. The rock physical and mechanical parameters mentioned in this embodiment include rock weathering strength, degree of fracture development, and shear strength. The blasting parameters mentioned in this embodiment include charge amount, perforation parameters, resistance line, and detonation sequence.

[0066] S2: Based on the geological survey data of the mine slope in the area to be excavated, construct a three-dimensional numerical model of slope stability in the area to be excavated;

[0067] Specifically, based on the geological survey data of the mine slope, an appropriate numerical simulation method is selected to establish a three-dimensional numerical model of the slope stability of the area to be excavated, as shown in the attached figure. Figure 2 As shown; in this embodiment, the geological survey data of the mine slope provides the basis for the material parameters in the three-dimensional slope stability numerical model, and the numerical simulation method selected for comprehensive calculation efficiency is FLAC3D-PFC coupling technology;

[0068] S3: The three-dimensional slope stability numerical model of the area to be excavated, obtained in S2, is simulated under different blasting slope stability influencing factors to clarify the influence of blasting slope stability influencing factors on the slope safety factor, and a membership function is established; the blasting slope stability influencing factors include blasting parameters and rock physical and mechanical parameters, such as... Figure 3 As shown;

[0069] S3.1: Based on the actual mine site conditions, construct the boundary conditions of a three-dimensional slope stability numerical model that truly reflects the geological and topographical characteristics. Fully consider the constraints of important factors such as the actual topographic undulations and geological structural surfaces in the mining area. Use discrete-continuous coupling simulation analysis to calculate the slope safety factor of each region in the three-dimensional slope stability numerical model under different blasting parameter conditions, that is, the slope safety factor under different blasting slope stability influencing factors, and determine the influence law of blasting slope stability influencing factors on the slope safety factor.

[0070] In this embodiment, the reduced strength is used to calculate the slope safety factor Fs, which is used to comprehensively evaluate the overall stability of the slope, and finally to obtain the influence law of the blasting slope stability influencing the slope safety factor, as follows:

[0071] (1) When the charge is large, the concentration of detonation energy increases, the amplitude and radius of the blasting stress wave increase, the range of plastic failure zone and through crack in the slope expands, and the safety factor decreases significantly; when the charge is small, the blasting disturbance is relatively weakened, the range of slope damage is limited, the safety factor remains at a high level and increases as the charge decreases.

[0072] (2) When the resistance line is small, the distance between the blast hole and the free surface is too close, and the blasting energy is concentrated near the slope, which easily forms a significant local weakening zone, resulting in a decrease in the slope safety factor; as the resistance line increases, the propagation path of energy in the slope body is lengthened, the slope disturbance is weakened, and the safety factor gradually increases and tends to stabilize.

[0073] (3) When the hole mesh parameters are small, the superposition effect of blasting stress waves from adjacent blast holes is significant, the stress concentration increases, the overall stiffness of the slope is weakened, and the safety factor shows a downward trend. As the hole mesh parameters increase, the mutual interference of stress waves weakens, the blasting disturbance is more dispersed in space, and the slope safety factor increases accordingly.

[0074] (4) When micro-delay detonation is used, the blasting energy is dispersed over time, the peak stress wave at a single moment decreases, the dynamic response of the slope weakens, and the corresponding slope safety factor is higher; when instantaneous detonation is used, the energy is released in a concentrated manner in a very short time, the peak stress wave increases, the impact on the slope is more severe, and the slope safety factor is significantly reduced.

[0075] (5) When the rock weathering intensity is high, the structure is loose and the strength and stiffness are significantly reduced. Under the combined action of self-weight and blasting disturbance, it is more likely to deform and be damaged. The slope safety factor decreases significantly as weathering intensifies. When the rock weathering intensity is low or basically fresh, the integrity and mechanical properties are better, the ability to resist damage is stronger, the slope safety factor remains at a relatively high level and increases as weathering weakens.

[0076] (6) When the joints and fissures are highly developed, dense, and interconnected, the rock mass is divided into blocks or thin plates, and potential slip surfaces are more likely to form and connect. The slope safety factor decreases significantly with the increase of the degree of joint and fissure development. When the joints and fissures are few, relatively sparsely distributed, and have limited extension, the rock mass has better integrity, and potential failure surfaces are difficult to form continuous channels. The slope safety factor increases with the decrease of the degree of joint and fissure development.

[0077] (7) When the shear strength (including the shear capacity reflected by the cohesion and internal friction angle) of the rock mass and its joint surfaces is high, the potential sliding surface is not easy to undergo shear failure under external load and blasting disturbance, and the slope safety factor increases with the increase of shear strength; when the shear strength is low, the rock mass is not strong enough to resist shear sliding, and it is more likely to become unstable along the weak surface under the same working conditions, and the slope safety factor decreases significantly with the decrease of shear strength.

[0078] S3.2: Based on the influence law of the blasting slope stability influencing factors on the slope safety factor, the membership function is constructed using fuzzy mathematics theory to quantitatively map the mapping relationship between the blasting slope stability influencing factors and the slope safety factor, as shown in the following example;

[0079] (1) Membership function of charge quantity and slope safety factor:

[0080] (1);

[0081] in, It is the ratio of the actual charge to the design allowable charge. This represents the membership value of the slope safety factor corresponding to the charge quantity.

[0082] (2) Membership function of perforated grid parameters and slope safety factor:

[0083] (2);

[0084] in, This represents the deviation between the actual and optimized perforated mesh parameters. This represents the membership value of the slope safety factor corresponding to the perforated mesh parameters;

[0085] (3) Membership function of resistance line and slope safety factor:

[0086] (3);

[0087] in, As the actual line of resistance, The lower limit threshold of the resistance line (a level that is significantly too small and has a low slope safety factor). This is the upper limit threshold of the resistance line (increasing it after it is reached will not significantly improve the safety factor). The membership value of the slope safety factor corresponding to the resistance line;

[0088] (4) Membership function of detonation sequence and slope safety factor:

[0089] (4);

[0090] in, These are the detonation timing parameters. The membership value of the slope safety factor corresponding to the detonation sequence;

[0091] (5) Membership function of rock weathering intensity and slope safety factor:

[0092] (5);

[0093] in, This represents the membership value of the slope safety factor corresponding to the rock weathering intensity. The rock weathering intensity index (can be normalized to 0–1 or graded quantification); This is the lower threshold at which weathering begins to significantly affect the safety factor; The upper limit threshold is for areas that are clearly severely weathered and are generally considered unsafe.

[0094] (6) Membership function of crack development degree and slope safety factor:

[0095] (6);

[0096] in, The membership value of the slope safety factor corresponding to the degree of crack development; As an indicator of the degree of joint and fracture development; This is the lower threshold at which the development of joints and fissures begins to have a significant impact on stability. The upper limit threshold for highly developed joints and fissures that easily form through-slip surfaces;

[0097] (7) Membership function of shear strength and slope safety factor:

[0098] (7);

[0099] in, This represents the membership value of the slope safety factor corresponding to the shear strength. It is a rock mass shear strength index (which can be a normalized quantity of the combined index of equivalent shear strength, cohesion and internal friction angle). This is the lower threshold value for areas with low shear strength and prone to instability. The upper limit threshold is set for materials with high shear strength that basically meet safety requirements.

[0100] S4: Construct a slope stability prediction model that comprehensively considers the influence of multiple factors based on the membership function, and train it using historical case data to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors; the slope stability prediction model that comprehensively considers the influence of multiple factors is used to predict the slope safety factor based on the input blasting parameters and rock physical and mechanical parameters.

[0101] S4.1: Utilize historical case data to construct a training dataset, optimize the membership function, and improve prediction accuracy and applicability;

[0102] In this embodiment, a membership function calibration method based on error minimization and parameter adjustment is adopted to enable the membership function to adapt to different blasting conditions. At the same time, basic monotonicity and boundary constraints are applied during the optimization process (e.g., the membership degree of charge decreases monotonically with the charge ratio, and the membership degree of shear strength increases monotonically with the strength index) to prevent the membership function from exhibiting reverse changes that contradict the engineering mechanism, and to ensure that the membership calibration results are numerically convergent and physically reasonable.

[0103] S4.2: Based on the optimized membership function, construct a slope stability prediction model that comprehensively considers the influence of multiple factors;

[0104] In this embodiment, an intelligent prediction framework is constructed with the mapping path of "influencing factors of blasted slope stability → slope response → stability level". The slope stability prediction model that comprehensively considers the influence of multiple factors is as follows: Figure 4 As shown, it includes an input layer, a fuzzy mapping layer, a membership-driven risk attention mechanism module, and an output layer;

[0105] The input layer is used to receive input blasting parameters and rock physical and mechanical parameters;

[0106] The fuzzy mapping layer is used to map blasting parameters and rock physical and mechanical parameters to corresponding membership values ​​under fuzzy rules according to the membership function, and to form a fuzzy evaluation vector with all membership values ​​to characterize the membership degree of each influencing factor under the current blasting condition.

[0107] The membership-driven risk attention mechanism module takes as input a fuzzy evaluation vector and a feature vector composed of normalized blasting parameters and rock physical and mechanical parameters. The membership-driven risk attention mechanism then weights each feature to obtain a risk-weighted feature vector. Input to the output layer;

[0108] The output layer receives the risk-weighted feature vector output by the membership-driven risk attention mechanism module. By extracting high-order features through several layers of nonlinear sensing units (MLP), the predicted value of the slope safety factor under this blasting disturbance is finally obtained.

[0109] The calculation process of the membership-driven risk attention mechanism is as follows:

[0110] First, to incorporate each membership value in the fuzzy evaluation vector into the relevance calculation, based on membership... Define risk amplification factor :

[0111] (8);

[0112] in, For indexing, For weight parameters, when When smaller, 1- If it is larger, then Larger; when When it is large, Close to 1;

[0113] Based on this, the risk amplification factor Embedding the self-attention score yields a risk relevance score after fusing membership:

[0114] (9);

[0115] in, For self-attention score; The risk relevance score after merging membership degrees;

[0116] Then, the risk relevance score is calculated by fusing the membership degrees of all features. Normalization is performed to obtain the risk attention weights for each feature. ;

[0117] Utilizing risk attention weights Numerical vectors for each feature We perform weighted aggregation to obtain a risk-weighted feature vector that highlights the risk factors. ;

[0118] (10);

[0119] in, The number of features;

[0120] Through the above calculation process, the membership-driven risk attention mechanism explicitly introduces the safety membership prior of each influencing factor when calculating the attention weight: when the blasting factor is significant, parameters such as charge quantity, blasting sequence, and resistance line receive higher attention weights due to their lower safety membership; when the geological structure has significant control, the weights of parameters such as rock weathering degree, fracture development degree, and shear strength are correspondingly increased, thereby automatically highlighting the risk factors that have the greatest impact on the slope safety factor during the feature aggregation process, and improving the ability to identify critical instability conditions;

[0121] S4.3: Use the training dataset to train the slope stability prediction model that comprehensively considers the influence of multiple factors, and obtain the trained slope stability prediction model that comprehensively considers the influence of multiple factors.

[0122] During training, the physical control equations related to the slope dynamic response are simultaneously embedded as constraints into the loss function. These physical control equations include at least the following:

[0123] Ph1: Dynamic equilibrium equation (conservation of momentum).

[0124] Ph2: Generalized constitutive equation (Hooke's law);

[0125] Ph3: Wave propagation equation (the propagation behavior of blast waves in rock mass);

[0126] The residuals of the aforementioned dynamic equilibrium equations, generalized constitutive equations, and wave propagation equations are incorporated into the loss function for joint optimization with data errors. This aims to improve the physical consistency and generalization ability of the model. The specific loss function is as follows:

[0127] (11);

[0128] in, For loss function, , , These are the weighting coefficients for the corresponding loss terms. For the first The measured output of each sample For the network's predicted output, The physical constraint sampling point number; , , These are the residuals obtained by substituting the model output into the physical control relationships such as the dynamic equilibrium equation, the generalized constitutive equation, and the wave propagation equation. For network parameters, and These are the number of data samples and the number of physical constraint sampling points, respectively.

[0129] Model Output and Closed-Loop Evaluation: When the prediction error meets the target convergence condition, the model outputs the predicted value of the slope safety factor, thereby obtaining the landslide risk level. If the error exceeds the limit, it enters the next round of iterative training. Finally, a slope stability prediction model integrating fuzzy inference, physical constraints, and depth perception is formed, which has high adaptability, high interpretability, and real-time feedback capability.

[0130] S5: Using a slope stability prediction model that comprehensively considers the influence of multiple factors after training, analyze the real-time monitoring data during the actual blasting construction process, automatically output the predicted slope safety factor, and then obtain the landslide risk level and generate early warning information in real time to guide the dynamic optimization and timely adjustment of construction parameters; the landslide risk level includes safe, critical, and unstable.

[0131] Specifically, it includes the following steps:

[0132] S5.1: During actual blasting operations, real-time monitoring data of the slope is collected, including blasting parameters and rock physical and mechanical parameters;

[0133] S5.2: Utilizing the trained slope stability prediction model that comprehensively considers the influence of multiple factors, the model outputs the predicted slope safety factor in real time. Based on the slope safety factor, it automatically classifies landslide risk levels and generates early warning information. The early warning results are displayed intuitively in the form of a curve graph, such as... Figure 5 As shown, it is clearly divided into three risk levels;

[0134] White (Safe): Slope safety factor Fs > 1.3, slope condition is stable, blasting construction can be carried out normally;

[0135] Yellow (critical): The slope safety factor is 1.0 < Fs ≤ 1.3. The slope is in a critical stability state. Monitoring needs to be strengthened and blasting parameters need to be adjusted appropriately.

[0136] Red (Instability): The slope safety factor Fs ≤ 1.0, indicating a high risk of slope instability. Construction should be suspended immediately and emergency slope support and reinforcement measures should be taken.

[0137] S5.3: Dynamically optimize and adjust the blasting construction plan based on the landslide risk level classification and early warning information;

[0138] When the warning level is yellow (critical) or red (instability), the amount of explosive charge should be reduced promptly, and the layout of blast holes and the detonation sequence should be adjusted appropriately to reduce the impact of vibration caused by construction. In addition, buffering measures such as vibration damping pads, micro-delay, or staged detonation should be implemented to further reduce the negative impact of disturbance on slope stability. When slope stability is significantly reduced, corresponding support and reinforcement measures, such as anchor bolt reinforcement and shotcrete support, should be implemented immediately before the next stage of construction. Simultaneously, continuous slope response monitoring should be maintained, and monitoring data should be fed back to the model in real time to achieve continuous dynamic optimization and risk control of blasting construction parameters, ensuring construction safety and long-term slope stability.

[0139] Example 2:

[0140] An intelligent sensing, monitoring, and early warning system for landslides caused by blasting in open-pit mines is provided to implement an intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines, comprising:

[0141] The data acquisition module is used to acquire geological survey data and historical case data of the mine slope in the area to be excavated;

[0142] The model building module is used to construct a three-dimensional numerical model of slope stability in the area to be excavated based on the geological survey data of the mine slope in the area to be excavated.

[0143] The membership function construction module is used to simulate the three-dimensional slope stability numerical model of the area to be excavated under different blasting slope stability influencing factors, clarify the influence law of blasting slope stability influencing factors on slope safety factor, and establish membership function.

[0144] The model building and training module is used to build a slope stability prediction model that comprehensively considers the influence of multiple factors based on the membership function, and to train it using historical case data to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors.

[0145] The risk level prediction module is used to analyze real-time monitoring data during actual blasting construction using a slope stability prediction model that has been trained and comprehensively considers the influence of multiple factors. It automatically outputs the predicted slope safety factor, thereby obtaining the landslide risk level and generating early warning information in real time, and optimizing and adjusting the blasting construction plan.

[0146] Example 3:

[0147] This embodiment proposes an electronic device, including: one or more processors, and a memory, wherein the memory is used to store instructions, and when the instructions are executed by the one or more processors, the one or more processors execute the aforementioned intelligent sensing, monitoring and early warning method for landslides caused by blasting in open-pit mines.

[0148] The electronic device can be a mobile phone, computer, or tablet computer, etc., and includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, implements a method for intelligent sensing, monitoring, and early warning of landslides caused by blasting in open-pit mines, as described in the embodiments. It is understood that the electronic device may also include input / output (I / O) interfaces and communication components.

[0149] The processor is used to execute all or part of the steps in the intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines, as described in the above embodiments. The memory is used to store various types of data, which may include, for example, instructions for any application or method in an electronic device, as well as application-related data.

[0150] The processor can be implemented as an Application Specific Integrated Circuit (ASIC), Digital Signal Processor (DSP), Programmable Logic Device (PLD), Field Programmable Gate Array (FPGA), controller, microcontroller, microprocessor, or other electronic components, and is used to execute the intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines described in the above embodiments.

[0151] Example 4:

[0152] This embodiment proposes a computer-readable storage medium that stores executable instructions. When these instructions are executed, if they are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium.

[0153] The computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the intelligent sensing, monitoring and early warning method for open-pit mine blasting disturbance landslides described in the various embodiments of this application.

[0154] The aforementioned storage media include: flash memory, hard disk, multimedia card, card-type memory (e.g., SD (Secure Digital Memory Card) or DX (Memory Data Register, MDR) memory), random access memory (RAM), static random-access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic storage, disk, optical disk, server, APP (Application) application store, and other media capable of storing program verification codes. These media store computer programs, which, when executed by a processor, can implement the various steps of the aforementioned intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines.

[0155] Example 5:

[0156] This embodiment proposes a computer program product, including a computer program or instructions, which, when executed by a processor, implements the aforementioned intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines.

[0157] Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a computer program product.

[0158] The various embodiments in this application are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0159] The scope of protection of this application is not limited to the embodiments described above. Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from the scope and spirit of this disclosure. If such modifications and variations fall within the scope of this disclosure and its equivalents, then the intent of this disclosure also includes these modifications and variations.

Claims

1. A method for intelligent sensing, monitoring, and early warning of landslides caused by blasting in open-pit mines, characterized in that, Specifically, the following steps are included: Obtain geological survey data and historical case data of the mine slope in the area to be excavated; Based on the geological survey data of the mine slope in the area to be excavated, a three-dimensional numerical model of slope stability in the area to be excavated is constructed. The three-dimensional slope stability numerical model of the area to be excavated was simulated under different blasting slope stability influencing factors to clarify the influence law of blasting slope stability influencing factors on slope safety factor. Fuzzy mathematics theory was used to construct a membership function to quantitatively map the mapping relationship between blasting slope stability influencing factor conditions and slope safety factor. The blasting slope stability influencing factors include blasting parameters and rock physical and mechanical parameters. A slope stability prediction model that comprehensively considers the influence of multiple factors is constructed based on the membership function, and trained using historical case data to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors; the slope stability prediction model that comprehensively considers the influence of multiple factors is used to predict the slope safety factor based on the input blasting parameters and rock physical and mechanical parameters. By using a slope stability prediction model that comprehensively considers the influence of multiple factors after training, the real-time monitoring data during the actual blasting construction process is analyzed, the predicted slope safety factor is automatically output, the landslide risk level is obtained, and early warning information is generated in real time to optimize and adjust the blasting construction plan.

2. The intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines according to claim 1, characterized in that, The geological survey data of the mine slope includes the rock mass structure characteristics, joint distribution and rock physical and mechanical parameters of the area to be excavated; the historical case data includes a set of rock physical and mechanical parameters, blasting parameters and their corresponding slope safety factors; the rock physical and mechanical parameters include rock weathering intensity, degree of fracture development and shear strength; the blasting parameters include charge amount, hole mesh parameters, resistance line and detonation sequence.

3. The intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines according to claim 1, characterized in that, The obtained three-dimensional slope stability numerical model of the area to be excavated was simulated under different blasting slope stability influencing factors to clarify the influence of blasting slope stability influencing factors on the slope safety factor, and a membership function was established, specifically including: A1: Construct the boundary conditions of the three-dimensional slope stability numerical model, and use discrete continuous coupled simulation analysis to calculate the slope safety factor of each region in the three-dimensional slope stability numerical model under different blasting parameter conditions, that is, the slope safety factor under different blasting slope stability influencing factors, and determine the influence law of blasting slope stability influencing factors on the slope safety factor. A2: Based on the influence of blasting slope stability factors on slope safety factor, a membership function is constructed using fuzzy mathematics theory to quantitatively map the relationship between blasting slope stability factors and slope safety factor.

4. The intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines according to claim 1, characterized in that, The slope stability prediction model based on the membership function, which comprehensively considers the influence of multiple factors, is constructed and trained using historical case data to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors. Specifically, this includes: B1: Optimize the membership function by constructing a training dataset using historical case data; Specifically: a membership function calibration method based on error minimization and parameter adjustment is adopted, while monotonicity and boundary constraints are applied during the optimization process; B2: Based on the optimized membership function, construct a slope stability prediction model that comprehensively considers the influence of multiple factors; The slope stability prediction model that comprehensively considers the influence of multiple factors includes an input layer, a fuzzy mapping layer, a membership-driven risk attention mechanism module, and an output layer. The input layer is used to receive input blasting parameters and rock physical and mechanical parameters; The fuzzy mapping layer is used to map blasting parameters and rock physical and mechanical parameters to corresponding membership values ​​under fuzzy rules according to the membership function, and to form a fuzzy evaluation vector by all membership values. The membership-driven risk attention mechanism module takes as input a fuzzy evaluation vector and a feature vector composed of normalized blasting parameters and rock physical and mechanical parameters. The membership-driven risk attention mechanism then weights each feature to obtain a risk-weighted feature vector. Input to the output layer; The output layer receives the risk-weighted feature vector output by the membership-driven risk attention mechanism module. By extracting high-order features through several layers of nonlinear sensing units, the predicted value of the slope safety factor under this blasting disturbance is finally obtained. The calculation process of the membership-driven risk attention mechanism is as follows: First, to incorporate each membership value in the fuzzy evaluation vector into the relevance calculation, based on membership... Define risk amplification factor : （8）； in, For indexing, These are weight parameters; Based on this, the risk amplification factor Embedding the self-attention score yields a risk relevance score after fusing membership: （9）； in, For self-attention score; The risk relevance score after merging membership degrees; Then, the risk relevance score is calculated by fusing the membership degrees of all features. Normalization is performed to obtain the risk attention weights for each feature. ; Utilizing risk attention weights Numerical vectors for each feature We perform weighted aggregation to obtain the risk-weighted feature vector. ; （10）； in, The number of features; B3: Use the training dataset to train the slope stability prediction model that comprehensively considers the influence of multiple factors, and obtain the trained slope stability prediction model that comprehensively considers the influence of multiple factors. Specifically, the residuals of the dynamic equilibrium equation, the generalized constitutive equation, and the wave propagation equation are introduced into the Loss function as follows: （11）； in, For loss function, , , These are the weighting coefficients for the corresponding loss terms. For the first The measured output of each sample For the network's predicted output, The physical constraint sampling point number; , , These are the residuals obtained by substituting the model output into the dynamic equilibrium equation, the generalized constitutive equation, and the wave propagation equation, respectively. For network parameters, and These represent the number of data samples and the number of physical constraint sampling points, respectively.

5. The intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines according to claim 1, characterized in that, The method utilizes a slope stability prediction model that comprehensively considers the influence of multiple factors, which has been trained, to analyze real-time monitoring data during actual blasting construction. It automatically outputs the predicted slope safety factor, thereby obtaining the landslide risk level and generating real-time early warning information. This allows for the optimization and adjustment of the blasting construction plan. Specifically, this includes: C1: During actual blasting operations, real-time monitoring data of the slope is collected, including blasting parameters and rock physical and mechanical parameters; C2: Using a slope stability prediction model that has been trained and comprehensively considers the influence of multiple factors, the predicted slope safety factor is output in real time, and landslide risk level classification and early warning information are automatically generated based on the slope safety factor. C3: Dynamically optimize and adjust the blasting construction plan based on the landslide risk level classification and early warning information.

6. An intelligent sensing, monitoring, and early warning system for landslides caused by blasting in open-pit mines, used to implement the intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines as described in any one of claims 1-5, characterized in that, include: The data acquisition module is used to acquire geological survey data and historical case data of the mine slope in the area to be excavated; The model building module is used to construct a three-dimensional numerical model of slope stability in the area to be excavated based on the geological survey data of the mine slope in the area to be excavated. The membership function construction module is used to simulate the three-dimensional slope stability numerical model of the area to be excavated under different blasting slope stability influencing factors, clarify the influence law of blasting slope stability influencing factors on slope safety factor, and establish membership function. The model building and training module is used to build a slope stability prediction model that comprehensively considers the influence of multiple factors based on the membership function, and to train it using historical case data to obtain a trained slope stability prediction model that comprehensively considers the influence of multiple factors. The risk level prediction module is used to analyze real-time monitoring data during actual blasting construction using a slope stability prediction model that has been trained and comprehensively considers the influence of multiple factors. It automatically outputs the predicted slope safety factor, thereby obtaining the landslide risk level and generating early warning information in real time, and optimizing and adjusting the blasting construction plan.

7. An electronic device, characterized in that, include: One or more processors, and a memory for storing instructions that, when executed by the one or more processors, cause the one or more processors to perform the intelligent sensing, monitoring and early warning method for landslides caused by blasting disturbance in open-pit mines as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, It stores executable instructions that, when executed, cause the processor to perform the intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines, as described in any one of claims 1-5.

9. A computer program product, characterized in that, It includes a computer program or instructions that, when executed by a processor, implement the intelligent sensing, monitoring, and early warning method for landslides caused by blasting in open-pit mines as described in any one of claims 1-5.

Images