Deep stope structure safety evaluation method based on data and knowledge model joint driving

Abstract

The invention provides a deep stope structure safety evaluation method based on data and knowledge model combined driving, and relates to the technical field of deep stope structures, and the method specifically comprises the steps: continuously collecting geological parameters and environmental parameters related to a deep stope, building a knowledge model based on geotechnical mechanics and environmental analysis, and training the knowledge model; and meanwhile, a deformation response model is constructed and trained. A geological condition coefficient, an environment risk coefficient and a deformation response coefficient are obtained by utilizing current data, a comprehensive evaluation coefficient is generated through data processing and correlation analysis, and whether the stope structure is safe or not is judged by comparing the comprehensive evaluation coefficient with a preset threshold value. In the face of dynamic changes of geological conditions, environmental factors and deformation response in deep stope mining, data collected in real time is input into a geomechanical model, an environmental analysis model and a deformation response model which are trained in advance, dynamic updating of an evaluation result is achieved, and the evaluation accuracy is improved. And the defect of evaluation based on static data and a fixed model is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of deep stope structures, and specifically provides a safety evaluation method for deep stope structures jointly driven by data and knowledge models. Background Art

[0002] With the gradual depletion of shallow mineral resources, deep mining has become an important way to obtain resources. However, due to its special geological environment and mining conditions, the structural safety problems in deep stopes are becoming increasingly complex and severe. Accurately evaluating the structural safety status of deep stopes is crucial for ensuring the smooth progress of mining operations and personnel safety.

[0003] The existing evaluation methods have the following defects: First, only the influence of a single factor on the stope structure safety is concerned. For example, only the geological parameters are emphasized, ignoring the complex interaction between environmental factors and deformation responses. This one-sided evaluation method cannot truly reflect the actual complex situation of deep stopes, resulting in poor accuracy of evaluation results.

[0004] Second, during the mining process of deep stopes, the geological conditions, environmental factors, and the deformation responses of the stopes themselves are all in dynamic changes. However, most of the existing methods are based on static data and fixed models for evaluation, and fail to update and adapt to these dynamic changes in a timely manner. For example, as the mining depth increases and the mining scope expands, factors such as ground stress and groundwater level will change, but traditional evaluation methods cannot adjust the evaluation results in real time, thus affecting the accurate judgment of the stope structure safety.

[0005] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The purpose of the present invention is to provide a safety evaluation method for deep stope structures jointly driven by data and knowledge models to solve the problems raised in the above background art.

[0007] To achieve the above purpose, the present invention provides the following technical solutions: A safety evaluation method for deep stope structures jointly driven by data and knowledge models, the specific steps include: S1. Continuously collect various data related to the area where the deep stope is located for multiple times, including past and current geological parameters, environmental parameters, and deformation response data; S2. Construct a knowledge model based on geomechanics and environmental analysis. The knowledge model includes a geomechanical model and an environmental analysis model. Evaluate the previous geological parameters based on the expert scoring method to generate a geological condition coefficient for evaluating the quality of geological factors. Combine the previous geological parameters and the corresponding geological condition coefficients to train the geomechanical model. Evaluate the previous environmental parameters based on the expert scoring method to generate an environmental risk coefficient for evaluating the quality of environmental factors. Combine the previous environmental parameters and the corresponding environmental risk coefficients to train the environmental analysis model. Input the current geological parameters and environmental parameters into the geomechanical model and the environmental analysis model respectively to obtain the current geological condition coefficient and environmental risk coefficient. S3. Construct a deformation response model. Use the previous deformation response data as input and the deformation response coefficient of the previous deep stope as the label. The deformation response coefficient is determined by the expert group and train the deformation response model. S4. Input the current deformation response data into the trained deformation response model to obtain the deformation response coefficient of the current deep stope. S5. Perform data processing and correlation analysis on the current geological condition coefficient, environmental risk coefficient, and deformation response coefficient of the deep stope to generate a comprehensive evaluation coefficient for comprehensively evaluating the structural safety of the deep stope. S6. Compare the comprehensive evaluation coefficient with the pre-set structural safety threshold to determine whether the current deep stope structure meets the safety standards.

[0008] Furthermore, the geological parameters include rock mass density, rock mass porosity, and rock mass thickness. The environmental parameters include CO concentration, temperature, humidity, and groundwater level in the deep stope. The deformation response data includes horizontal displacement, vertical displacement, stress, and strain.

[0009] Furthermore, based on the geomechanical model, the expression of the geological condition coefficient is as follows: ; Where is the geological condition coefficient of the current deep stope; In the formula, is the current rock mass density, is the current rock mass porosity, is the current rock mass thickness; is the regression coefficient of the rock mass density, is the regression coefficient of the rock mass porosity, is the regression coefficient of the rock mass thickness. On the basis of , let .

[0010] Further, based on the environmental analysis model, the expression of the environmental risk coefficient is as follows: ; where, is the environmental risk coefficient of the current deep stope; In the formula, is the interaction term of the CO concentration and temperature in the current deep stope, is the interaction term of the humidity and groundwater level in the current deep stope; is the regression coefficient of the interaction term of the CO concentration and temperature, is the regression coefficient of the interaction term of the humidity and groundwater level. On the basis of , let .

[0011] Further, the geological condition coefficient, environmental risk coefficient and deformation response coefficient of the current deep stope are processed and analyzed for correlation to generate a comprehensive evaluation coefficient. The formula is as follows: ; where, is the comprehensive evaluation coefficient of the current deep stope, is the geological condition coefficient of the current deep stope, is the environmental risk coefficient of the current deep stope, is the deformation response coefficient of the current deep stope; In the formula, , and are the weight coefficients of the geological condition coefficient, environmental risk coefficient and deformation response coefficient respectively, and , and The specific values of are determined by the analytic hierarchy process.

[0012] Further, the specific process of step S6 is as follows: When , the structure of the current deep stope meets the safety standards; When , the structure of the current deep stope does not meet the safety standards; is the safety threshold of the deep stope structure, indicating the critical value of the safety of the deep stope structure.

[0013] Compared with the prior art, the beneficial effects of the present invention are: The present invention overcomes the limitation of only focusing on a single factor in the prior art by continuously collecting geological parameters, environmental parameters, and deformation response data covering the past and the present for multiple times. It comprehensively considers numerous key factors affecting the structural safety of deep mining stope and their complex interactions, significantly improving the evaluation accuracy. When facing the dynamic changes in geological conditions, environmental factors, and deformation response during the deep mining stope operation, the present invention utilizes the real-time collected data, inputs it into the pre-trained geomechanics model, environmental analysis model, and deformation response model, and obtains the corresponding current geological condition coefficient, environmental risk coefficient, and deformation response coefficient in real time, realizing the dynamic update of the evaluation results and avoiding the drawbacks of evaluation based on static data and fixed models.

[0014] Meanwhile, the present invention constructs a professional knowledge model and conducts targeted training. Each model is constructed based on professional theories and a large amount of data. By comparing with the coefficients determined by the expert scoring method, the model coefficients are iteratively updated using an optimization algorithm to minimize the loss function, enhancing the reliability of the evaluation.

[0015] By comparing the comprehensive evaluation coefficient with the pre-set structural safety threshold, it provides a clear decision-making basis for the deep mining stope operation, ensures the smooth progress of the operation and the safety of personnel, and effectively improves the scientificity, accuracy, and practicality of the deep mining stope structural safety evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic diagram of the overall method flow of the present invention; Figure 2 It is a fitting curve graph of the rock mass density and the geological condition coefficient of the present invention; Figure 3 It is a fitting curve graph of the rock mass thickness and the geological condition coefficient of the present invention; Figure 4 It is a fitting curve graph of the rock mass porosity and the geological condition coefficient of the present invention; Figure 5 It is a fitting curve graph of the interaction term of humidity and groundwater level and the environmental risk coefficient of the present invention; Figure 6 It is a fitting curve graph of the interaction term of CO concentration and humidity and the environmental risk coefficient of the present invention; Figure 7 It is a fitting curve graph of the geological condition coefficient and the comprehensive evaluation coefficient of the present invention; Figure 8 It is a fitting curve graph of the environmental risk coefficient and the comprehensive evaluation coefficient of the present invention; Figure 9 It is a fitting curve graph of the deformation response coefficient and the comprehensive evaluation coefficient of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0017] To make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to specific embodiments.

[0018] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention shall have the ordinary meanings understood by those of ordinary skill in the field to which the present invention pertains. The "first", "second", and similar terms used in the present invention do not denote any order, quantity, or importance, but are merely used to distinguish different components. The terms such as "comprising" or "including" mean that the elements or items appearing before this word cover the elements or items listed after this word and their equivalents, without excluding other elements or items. The terms such as "connected" or "coupled" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. The terms such as "upper", "lower", "left", "right", etc. are only used to represent relative positional relationships, and when the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0019] Embodiment 1: Please refer to Figure 1 , the present invention provides a technical solution: A method for safety assessment of deep stope structures jointly driven by data and knowledge models, the specific steps including: S1. Continuously collect various data related to the area where the deep stope is located for multiple times, including past and current geological parameters, environmental parameters, and deformation response data. The geological parameters include rock mass density, rock mass porosity, and rock mass thickness. The environmental parameters include CO concentration, temperature, humidity, and groundwater level in the deep stope. The deformation response data includes horizontal displacement, vertical displacement, stress, and strain; Based on the above embodiment, the structure of the deep stope in S1 mainly refers to various support, enclosure, and transportation systems established for resource extraction or transportation in deep engineering such as underground mines or tunnels.

[0020] Based on the above embodiment, the collection methods and equipment for geological parameters, environmental parameters, and deformation response data are as follows: Through the core sampling method, core samples are taken from underground rock formations, and the density of the rock mass in the area where the deep stope is located is determined through laboratory analysis.

[0021] Using the water saturation method, the extracted core samples are immersed in clean water to ensure that the cores are completely submerged in water. Wait for the samples to fully absorb water until their mass no longer changes. By measuring the volume of water displaced by the samples in water, the pore volume of the samples is calculated. Porosity = (pore volume / sample volume) x 100%. Ultrasonic detectors are arranged in the area where the deep stope is located. An ultrasonic transmitter is used to send ultrasonic signals to the rock mass, and the time when the signals are reflected back is recorded. According to the propagation speed of sound waves in the rock, the thickness of the rock mass is calculated.

[0022] The CO concentration in the deep stope is monitored by an electrochemical sensor.

[0023] The temperature change is continuously recorded by a temperature sensor to obtain the temperature of the deep stope.

[0024] The humidity of the deep stope is measured using a humidity sensor.

[0025] A water level gauge is set up in the area where the deep stope is located. Usually, a bubble water level gauge or an electronic water level gauge is used. The data of the water level gauge is read regularly to monitor the change of the groundwater level.

[0026] An electronic level is used to measure the horizontal and vertical displacements of the rock in the area where the deep stope is located.

[0027] A stress gauge is installed on the surface of the rock mass in the area where the deep stope is located. The stress gauge can monitor the stress change inside the rock in real time and transmit the data to the recording system.

[0028] The strain gauge is pasted on the surface of the rock. An area where the stress is relatively concentrated is selected to obtain effective data. The strain gauge can monitor the minute deformation on the surface in real time and convert the strain signal into an electrical signal. The recorded data is collected and analyzed to evaluate the strain state of the rock under different conditions.

[0029] After collecting density, porosity, thickness, CO concentration, temperature, humidity, groundwater level, horizontal displacement, vertical displacement, stress and strain, maximum-minimum normalization processing is performed on these parameters respectively. Then, the normalized data is used for the subsequent analysis and processing, so that during the subsequent analysis and processing, various data can be analyzed and processed under the same dimension, avoiding the problem that some data is ignored due to different dimensions.

[0030] Among them, for the equipment measuring the density and porosity of the rock mass, ultrasonic detectors, electrochemical sensors, temperature sensors, humidity sensors, water level gauges, electronic levels, stress gauges, and strain gauges, the models in the existing equipment can be used, and there is no limitation here.

[0031] Equipment for measuring the density and porosity of rock masses. The number of times of collecting data by ultrasonic detectors, electrochemical sensors, temperature sensors, humidity sensors, water level gauges, electronic level gauges, stress gauges, and strain gauges is multiple groups (such as 3 groups). The density, porosity, thickness, CO concentration, temperature, humidity, groundwater level, horizontal displacement, vertical displacement, stress, and strain are detected at different positions. Then, the same type of data detected at different positions is averaged, and the finally obtained average value is used as the corresponding data for density, porosity, thickness, CO concentration, temperature, humidity, groundwater level, horizontal displacement, vertical displacement, stress, and strain, so as to avoid accidental errors caused by single-point sampling.

[0032] S2. Construct a knowledge model based on geotechnical mechanics and environmental analysis. The knowledge model includes a geomechanical model and an environmental analysis model. Evaluate the past geological parameters based on the expert scoring method to generate a geological condition coefficient for evaluating the quality of geological factors. Combine the past geological parameters and the corresponding geological condition coefficients to train the geomechanical model. Evaluate the past environmental parameters based on the expert scoring method to generate an environmental risk coefficient for evaluating the quality of environmental factors. Combine the past environmental parameters and the corresponding environmental risk coefficients to train the environmental analysis model. Input the current geological parameters and environmental parameters into the geomechanical model and the environmental analysis model respectively to obtain the current geological condition coefficient and environmental risk coefficient. Table 1 shows the variation of the geological condition coefficient with the rock mass density, rock mass porosity, and rock mass thickness

[0033] Based on the data in the above embodiment, it can be seen from Table 1 that as the rock mass density gradually increases from 2.5 to 4.4, the geological condition coefficient continuously increases from 122.22 to 1448. That is, the greater the rock mass density, the greater the geological condition coefficient. Therefore, the rock mass density is positively correlated with the geological condition coefficient. The rock mass porosity gradually decreases from 0.15 to 0.01, and the geological condition coefficient increases from 122.22 to 1448, indicating that the smaller the rock mass porosity, the greater the geological condition coefficient. Therefore, the rock mass porosity is negatively correlated with the geological condition coefficient. The thickness of the rock layer gradually increases from 10 to 29, and the geological condition coefficient increases from 122.22 to 1448, meaning that the greater the thickness of the rock layer, the greater the geological condition coefficient. Therefore, the thickness of the rock layer is positively correlated with the geological condition coefficient.

[0034] From Figure 2Look, the data points of the geological condition coefficient of the black squares are distributed near the black fitting curve, showing an upward trend, and the curve is significantly bent, indicating a non-linear positive correlation between the rock mass density and the geological condition coefficient. As the rock mass density increases, the geological condition coefficient increases, and the growth rate gradually accelerates, indicating that the influence degree of the rock mass density on the geological condition coefficient is continuously increasing.

[0035] Figure 3 Among them, the black square data points rise along the black fitting curve, and the curve is non-linear, reflecting that the rock layer thickness and the geological condition coefficient are non-linearly positively correlated. When the rock layer thickness increases, the geological condition coefficient increases, and the growth trend is slow first and then fast, meaning that as the rock layer thickness increases, its promotion effect on the geological condition coefficient becomes more significant.

[0036] Observe Figure 4 , the black square data points rise along the black fitting curve as the rock mass porosity decreases, and the curve is non-linear. It shows that there is a non-linear negative correlation between the rock mass porosity and the geological condition coefficient, that is, the smaller the rock mass porosity, the larger the geological condition coefficient, and for the same reduction in porosity, the increase in the geological condition coefficient is more obvious in the early stage.

[0037] Based on the above embodiments, the expression of the geomechanical model is as follows: ; Among them, is the geological condition coefficient of the current deep stope. The geological condition coefficient is used to comprehensively evaluate the structural stability and safety of the deep stope by combining the rock mass density, porosity, and rock layer thickness. The larger the geological condition coefficient, the stronger the overall quality and bearing capacity of the rock mass, and the higher the structural safety and stability of the deep stope.

[0038] In the formula, is the current rock mass density, is the current rock mass porosity, is the current rock layer thickness; is the regression coefficient of the rock mass density, is the regression coefficient of the rock mass porosity, is the regression coefficient of the rock layer thickness, which is used to reflect the influence degree of different geological parameters on the geological condition coefficient. are all the coefficients to be optimized in the model expression. Through model training for iterative optimization, specifically: calculate the geological condition coefficient output by the geomechanical model and the geological condition coefficient determined by the expert scoring method, obtain the deviation value between the two, define the loss function as the deviation value between the two, and iteratively update through an optimization algorithm (such as gradient descent) to minimize the loss function. Use the mean square error as the loss function. When the mean square error is within When within the range, the optimization of is completed.

[0039] The expression input of the geomechanical model is the rock mass density, the rock mass porosity, and the rock mass thickness, and the output is the geological condition coefficient.

[0040] On this basis, it should be noted that: When the rock mass density increases, it means that the content of rock mass substances per unit volume increases, the interaction force between atoms or molecules strengthens, making the rock mass more dense. This directly improves the strength and bearing capacity of the rock mass itself, and can better resist various stresses brought by mining activities. For example, when bearing the overlying rock pressure above the stope and the mining disturbance stress, the high-density rock mass is less likely to deform and break, thus providing a more stable support structure for the deep stope. The higher the structural safety and stability of the deep stope, and then the higher the geological condition coefficient.

[0041] When the rock mass thickness increases, it is equivalent to increasing the thickness of the bearing layer of the stope surrounding rock. The thicker rock mass can disperse the stress generated during mining and reduce the stress concentration degree. For example, at the roof of the stope, the thicker rock layer can act like a more solid "ceiling" to better bear the pressure above the roof and reduce the risk of roof caving. At the same time, the thick rock mass also has a stronger lateral support effect on the surrounding rock of the stope, which helps to maintain the stability of the overall structure of the stope, making the structural safety and stability of the deep stope improved, so the geological condition coefficient is higher.

[0042] When the rock mass porosity increases, it indicates that there are more void spaces inside the rock mass. These voids weaken the continuity and integrity of the rock mass, making the stress transmission inside the rock mass uneven and prone to stress concentration around the pores. In addition, the existence of pores also reduces the effective bearing area of the rock mass, resulting in a decrease in the overall strength of the rock mass. When subjected to external forces such as mining stress, the rock mass containing more pores is more likely to crack and deform, leading to a reduction in the structural safety and stability of the deep stope, so the geological condition coefficient is lower.

[0043] Therefore, the geological condition coefficient and the rock mass density 、the rock mass thickness are positively correlated, and the geological condition coefficient and the rock mass porosity are negatively correlated. Therefore, the functional relationship between the geological condition coefficient and the rock mass density 、the rock mass porosity 、the rock mass thickness is expressed in the form of a fraction.

[0044] Among them, The setting standard of the size relationship: Rock mass density directly reflects the compactness and bearing capacity of the rock. A higher density usually means stronger compressive capacity and better structural stability. Therefore, in many cases, density has a greater impact on the quality of the rock mass. Thus, the regression coefficient of the rock mass density is set to be the largest; Rock mass thickness affects its overall stability and load dispersion ability. A thicker rock mass can withstand greater external pressure and provide a better safety factor. Although the influence of thickness is important, compared with density, its influence is usually smaller because the increase in thickness does not always significantly improve the strength of the rock mass, especially in the case of low density. Thus, the regression coefficient of the rock mass thickness is lower than the regression coefficient of the rock mass density ; Porosity refers to the proportion of the void volume in the total volume of the rock mass. A higher porosity will lead to a decrease in the strength and stability of the rock mass. The existence of voids weakens the overall bearing capacity of the rock mass. In the formula, porosity is in the denominator position, meaning that an increase in porosity will lead to a decrease. In the overall evaluation, although the importance of porosity cannot be ignored, its influence is negative. Thus, the regression coefficient is set relatively small.

[0045] In summary, the regression coefficient of the rock mass density is greater than the regression coefficient of the rock mass thickness, and the regression coefficient of the rock mass thickness is greater than the regression coefficient of the rock mass porosity. Therefore, the constraint condition is that on the basis of make .

[0046] Table 2 shows the variation of the environmental risk coefficient with the interaction terms of CO concentration and temperature, and the interaction terms of humidity and groundwater level

[0047] Based on the above embodiments, according to the data in Table 2, as the interaction term of CO concentration and temperature gradually increases from 10 to 105, the environmental risk coefficient continuously increases from 22 to 117, that is, the larger the value of the interaction term of CO concentration and temperature, the greater the environmental risk coefficient. Therefore, the interaction term of CO concentration and temperature and the environmental risk coefficient are positively correlated.

[0048] The interaction term of humidity and groundwater level gradually increases from 50 to 145, and the environmental risk coefficient increases from 22 to 117, indicating that the larger the value of the interaction term of humidity and groundwater level, the greater the environmental risk coefficient. Therefore, the interaction term of humidity and groundwater level and the environmental risk coefficient are positively correlated.

[0049] Observation Figure 5 , the data points of the environmental risk coefficient of the black squares are also closely distributed on the black fitting curve, showing a linear upward trend. This reflects a linear positive correlation between the interaction term of humidity and groundwater level and the environmental risk coefficient. That is, the larger the value of the interaction term of humidity and groundwater level, the environmental risk coefficient increases linearly at a fixed ratio, meaning that the influence of this interaction term on the environmental risk coefficient also has a stable linear relationship.

[0050] From Figure 6 it can be seen that the data points of the black squares represent the environmental risk coefficient, and the black curve is the fitting curve. It can be found that the data points are closely distributed near the black fitting curve and show an obvious linear upward trend. This indicates a linear positive correlation between the interaction term of CO concentration and temperature and the environmental risk coefficient. That is to say, as the value of the interaction term of CO concentration and temperature increases, the environmental risk coefficient will increase linearly.

[0051] Based on the above embodiments, the expression of the environmental analysis model is as follows: ; Among them, is the environmental risk coefficient of the current deep stope. The environmental risk coefficient is used to comprehensively evaluate the environmental risk of the deep stope by combining the interaction term of CO concentration and temperature and the interaction term of humidity and groundwater level. The greater the environmental risk coefficient, the greater the environmental risk of the deep stope.

[0052] In the formula, is the interaction term of CO concentration and temperature in the deep stope, is the interaction term of humidity and groundwater level in the deep stope; is the regression coefficient of the interaction term of CO concentration and temperature, is the regression coefficient of the interaction term of humidity and groundwater level, which is used to reflect the influence degree of different environmental parameters on the environmental risk coefficient, are all coefficients to be optimized in the model expression, and are iteratively optimized through model training. Specifically: calculate the environmental risk coefficient output by the environmental analysis model and the environmental risk coefficient determined by the expert scoring method, obtain the deviation value between the two, define the loss function as the deviation value between the two, and iteratively update by an optimization algorithm (such as gradient descent) to minimize the loss function. The mean square error is used as the loss function. When the mean square error is within When within the range, the optimization of is completed.

[0053] The input of the model expression of the environmental analysis model is CO concentration, temperature, humidity, and groundwater level, and the output is the environmental risk coefficient.

[0054] On this basis, it should be noted that: When the interaction term of CO concentration and temperature increases, it means that in a high-temperature environment, the CO concentration rises, and high temperature will accelerate the diffusion and chemical reaction activity of CO. High-concentration CO not only poses a direct threat to the life safety of underground workers and is prone to poisoning accidents, but also the reaction of CO with surrounding substances may change the physical and chemical properties of the rock mass, such as corroding minerals or support materials in the rock mass, thereby affecting the stability of the stope structure. At the same time, high temperature itself will also reduce the mechanical properties of the rock, and the interaction between the two further increases the environmental risk of the deep stope, so the environmental risk coefficient is larger.

[0055] When the interaction term of humidity and groundwater level increases, it indicates that the groundwater level rises and the humidity increases. A high-humidity environment easily causes the clay minerals in the rock mass to expand, resulting in an increase in the volume of the rock mass, and then generating additional stress on the surrounding rock of the stope, increasing the possibility of surrounding rock deformation and failure. The rise of the groundwater level will reduce the effective stress of the rock mass, weaken the shear strength of the rock mass, and increase the risk of disasters such as water inrush in the stope. In addition, high humidity and high groundwater level may also accelerate the corrosion of metal support materials and reduce the support effect. These factors combined make the environmental risk of the deep stope increase significantly, so the environmental risk coefficient is larger.

[0056] Therefore, the environmental risk coefficient is positively correlated with the interaction term of CO concentration and temperature, and the interaction term of humidity and groundwater level.

[0057] From a physical perspective, the influence of environmental factors such as CO concentration, temperature, humidity, and groundwater level on the environmental risk of the deep stope does not exist in isolation, but interacts with each other. The linear weighting function can intuitively reflect the combined influence of the interaction between CO concentration and temperature, and the interaction between humidity and groundwater level on the environmental risk coefficient.

[0058] When the interaction term of CO concentration and temperature increases, and the interaction term of humidity and groundwater level increases, their respective promotion effects on environmental risk can be reflected in a simple additive form in the linear weighting function, reflecting the characteristic that environmental risk accumulates as each factor changes.

[0059] Among them, Setting standard for the size relationship: There is a strong positive correlation between CO concentration and temperature. For example, under certain environmental conditions, an increase in temperature will promote the release of CO or an increase in its concentration. This strong correlation makes the value larger because the interaction of these two factors has a more significant impact on environmental risks.

[0060] The groundwater level is mainly affected by factors such as precipitation, soil type, and pumping activities. The change in humidity has a relatively small impact on the groundwater level. Therefore, the interaction term between humidity and the groundwater level has a relatively small impact on environmental risks.

[0061] In summary, the regression coefficient of the interaction term between CO concentration and temperature is greater than that of the interaction term between humidity and the groundwater level. Therefore, on the basis of let .

[0062] S3. Construct a deformation response model. Use the previous horizontal displacement, vertical displacement, stress, and strain as inputs, and the deformation response coefficients of previous deep stope as labels. The deformation response coefficients are determined by an expert group, and then train the deformation response model; S4. Input the current horizontal displacement, vertical displacement, stress, and strain into the trained deformation response model to obtain the current deformation response coefficients of the deep stope; On the basis of the above embodiments, the deformation response model is established based on historical data (such as previous horizontal displacement, vertical displacement, stress, and strain). By learning and training a large amount of historical deformation data, this model can capture the complex relationship between deformation response and input parameters, thereby realizing the prediction of the current deformation state of the stope. Therefore, the deformation response model is called a data model.

[0063] On the basis of the above embodiments, the deformation response model is composed of a deep learning network based on a multi-layer perceptron. The deep neural network of the multi-layer perceptron includes an input layer, a first hidden layer, a second hidden layer, a third hidden layer, and an output layer. The first hidden layer, the second hidden layer, and the third hidden layer all have at least two neurons and all use ReLU as the activation function; In this embodiment, the input features of the deep learning network of the multi-layer perceptron include: horizontal displacement, vertical displacement, stress, and strain, a total of 4 features.

[0064] The structure of the deep learning network of the multi-layer perceptron is as follows: Input layer: Receive the input of 4 features; First hidden layer: Has 64 neurons and uses ReLU as the activation function; Second hidden layer: Has 32 neurons and also uses the ReLU activation function; The third hidden layer: It has 16 neurons and uses the ReLU activation function; The output layer: It has 1 neuron and outputs the deformation response coefficient of the deep stope.

[0065] The process of training the deformation response model is as follows: The previous horizontal displacement, vertical displacement, stress, and strain are used as input quantities, and the previous deformation response coefficient of the deep stope is used as a label for training. The deformation response coefficient is determined by the expert group and is in the range of 1 to 9. The larger the deformation response coefficient, the lower the structural safety of the deep stope. The mean square error is used as the loss function. When the mean square error is within the range, the training of the deformation response model is completed.

[0066] S5. Perform data processing and correlation analysis on the geological condition coefficient, environmental risk coefficient, and deformation response coefficient of the current deep stope to generate a comprehensive evaluation coefficient for comprehensively evaluating the structural safety of the deep stope; Table 3 shows the variation of the comprehensive evaluation coefficient with the geological condition coefficient, environmental risk coefficient, and deformation response coefficient

[0067] Based on the data in the above embodiments, it can be seen from Table 3 that the geological condition coefficient is significantly positively correlated with the comprehensive evaluation coefficient: when the geological condition coefficient increases from 1.7 to 15, the comprehensive evaluation coefficient synchronously increases from 1 to 13.35, indicating that the better the geological conditions, the higher the structural safety of the deep stope.

[0068] The environmental risk coefficient and the deformation response coefficient are negatively correlated with the comprehensive evaluation coefficient: when the environmental risk coefficient decreases from 34 to 15 and the deformation response coefficient decreases from 10 to 0.5, the comprehensive evaluation coefficient gradually increases, indicating that the lower the environmental risk and the smaller the deformation, the higher the safety of the stope.

[0069] In summary, the change of the comprehensive evaluation coefficient is the result of the combined action of the improvement of geological conditions and the reduction of environmental risk and deformation. For example, when the geological condition coefficient increases by 0.7, the environmental risk coefficient decreases by 1, and the deformation response coefficient decreases by 0.5, the comprehensive evaluation coefficient increases by about 0.65, reflecting the quantitative impact of each factor on the stope safety.

[0070] From Figure 7Looking at it, the data points of the comprehensive evaluation coefficient of the black squares are closely distributed on the black fitting curve, showing a linear upward trend. This indicates that there is a linear positive correlation between the geological condition coefficient and the comprehensive evaluation coefficient. That is, as the geological condition coefficient increases, the comprehensive evaluation coefficient increases linearly, reflecting that the geological condition coefficient has a positive impact on the comprehensive evaluation coefficient. The higher the geological condition coefficient, the better the safety of the deep stope structure is reflected in the comprehensive evaluation.

[0071] Observation Figure 8 , the data points of the black squares show a linear downward trend along the black fitting curve. This reflects a linear negative correlation between the environmental risk coefficient and the comprehensive evaluation coefficient. When the environmental risk coefficient increases, the comprehensive evaluation coefficient decreases linearly, indicating that the environmental risk coefficient has a reverse impact on the comprehensive evaluation coefficient. The higher the environmental risk coefficient, the worse the safety of the deep stope structure in the comprehensive evaluation.

[0072] In Figure 9 , the data points of the comprehensive evaluation coefficient of the black squares show a linear downward trend along the black fitting curve. This indicates a linear negative correlation between the deformation response coefficient and the comprehensive evaluation coefficient. As the deformation response coefficient increases, the comprehensive evaluation coefficient decreases linearly, meaning that the deformation response coefficient has a reverse effect on the comprehensive evaluation coefficient. That is, the higher the deformation response coefficient, the lower the safety of the deep stope structure in the comprehensive evaluation.

[0073] Based on the above embodiments, the geological condition coefficient, environmental risk coefficient, and deformation response coefficient of the current deep stope are processed and analyzed for correlation to generate a comprehensive evaluation coefficient. The formula is as follows: ; Among them, is the comprehensive evaluation coefficient of the current deep stope. The comprehensive evaluation coefficient is used to comprehensively score the safety of the current deep stope structure by combining the geological condition coefficient, environmental risk coefficient, and deformation response coefficient. And the larger the comprehensive evaluation coefficient, the safer the deep stope structure; It should be noted that from the above description, it can be seen that the larger the geological condition coefficient , the higher the safety and stability of the deep stope structure, the larger the environmental risk coefficient , the greater the environmental risk of the deep stope, and the larger the deformation response coefficient , the lower the safety of the deep stope structure. Therefore, the comprehensive evaluation coefficient is positively correlated with the geological condition coefficient , and the comprehensive evaluation coefficient is negatively correlated with both the environmental risk coefficient and the deformation response coefficient ; The safety of the deep stope structure is the result of the combined action of various factors such as geological conditions, environmental factors, and deformation responses. By means of weighted summation, the geological condition coefficient, environmental risk coefficient, and deformation response coefficient can be integrated into a comprehensive evaluation coefficient to comprehensively reflect the comprehensive impact of various factors on the safety of the deep stope structure. This method can capture the interrelationships between different factors and their overall contributions to the stope structure safety, avoiding the limitations of considering only a single factor in isolation and ignoring the influence of other factors.

[0074] In the formula, , and are the weight coefficients of the geological condition coefficient, environmental risk coefficient, and deformation response coefficient respectively, and , and The specific values are determined by the analytic hierarchy process, and the specific logic is as follows: Mark the three indicators of the geological condition coefficient, environmental risk coefficient, and deformation response coefficient, determine the relative importance values between each pair through the nine-scale method, and construct a judgment matrix. Among them, mark the index of the geological condition coefficient as 1, the index of the environmental risk coefficient as 2, and the index of the deformation response coefficient as 3. The constructed judgment matrix is: ; Among them, , Both represent the indexes of the coefficients, and , It means that the coefficient with index is more important to the comprehensive evaluation coefficient than the coefficient with index v. The specific value of is determined by relevant experts using the 1-9 scoring method. It means that the coefficient with index is extremely important to the comprehensive evaluation coefficient compared to the index with index v. It means that the coefficient with index is extremely unimportant to the comprehensive evaluation coefficient compared to the coefficient with index v;

[0075] Divide each element value in the judgment matrix by the sum of its columns to obtain a normalized judgment matrix. Calculate the mean value of each row element value in the normalized judgment matrix, and take the mean value of the first row element value as the proportional coefficient of the geological condition coefficient, the mean value of the second row element value as the proportional coefficient of the environmental risk coefficient, and the mean value of the third row element value as the proportional coefficient of the deformation response coefficient. With the constraint that the sum of the scaled values is equal to 1, scale the three proportional coefficients proportionally, and take the scaled values as the weights of the corresponding coefficients.

[0075] S6. Compare the comprehensive evaluation coefficient with the pre-set structural safety threshold to determine whether the current deep stope structure meets the safety standards.

[0076] Based on the above embodiments, the specific process of step S6 is as follows: When , the current deep stope structure meets the safety standards; When , the current deep stope structure does not meet the safety standards; is the safety threshold of the deep stope structure, representing the critical value of the safety of the deep stope structure. The safety threshold can be determined by statistically analyzing a large amount of experimental data to determine the distribution of the comprehensive evaluation coefficient. Statistical indicators such as the maximum value, minimum value, mean, and standard deviation in historical data can be used to set the safety threshold. For example, can be set to a certain standard deviation higher than the mean.

[0077] The above formulas are all dimensionless and take their numerical calculations. The formulas are obtained by collecting a large amount of data for software simulation to obtain a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0078] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this article can be implemented by computer software, electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0079] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They can be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0080] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and they should all be covered within the protection scope of this application.

Claims

1. A method for safety assessment of deep stope structure jointly driven by data and knowledge model, characterized in that: The specific steps include: S1. Continuously collect various data related to the area where the deep stope is located for multiple times, including past and current geological parameters, environmental parameters, and deformation response data; S2. Construct a knowledge model based on geomechanics and environmental analysis. The knowledge model includes a geomechanics model and an environmental analysis model. Evaluate the past geological parameters using the expert scoring method to generate a geological condition coefficient for evaluating the quality of geological factors. Combine the past geological parameters and the corresponding geological condition coefficients to train the geomechanics model. Evaluate the past environmental parameters using the expert scoring method to generate an environmental risk coefficient for evaluating the quality of environmental factors. Combine the past environmental parameters and the corresponding environmental risk coefficients to train the environmental analysis model. Input the current geological parameters and environmental parameters into the geomechanics model and the environmental analysis model respectively to obtain the current geological condition coefficient and environmental risk coefficient; S3. Construct a deformation response model. Use the past deformation response data as input and the deformation response coefficient of the past deep stope as the label. The deformation response coefficient is determined by the expert group and train the deformation response model; S4. Input the current deformation response data into the trained deformation response model to obtain the deformation response coefficient of the current deep stope; S5. Perform data processing and correlation analysis on the current geological condition coefficient, environmental risk coefficient, and deformation response coefficient of the deep stope to generate a comprehensive evaluation coefficient for comprehensively evaluating the structural safety of the deep stope; S6. Compare the comprehensive evaluation coefficient with the pre-set structural safety threshold to determine whether the current deep stope structure meets the safety standard.

2. The deep stope structure safety evaluation method based on the combined drive of data and knowledge model according to claim 1, characterized in that: The geological parameters include rock mass density, rock mass porosity, and rock mass thickness. The environmental parameters include CO concentration, temperature, humidity, and groundwater level in the deep stope. The deformation response data includes horizontal displacement, vertical displacement, stress, and strain.

3. The deep stope structure safety evaluation method based on the combined drive of data and knowledge model according to claim 2, wherein: Based on the geomechanics model, the expression of the geological condition coefficient is as follows: ； Among them, is the geological condition coefficient of the current deep stope; In the formula, is the current rock mass density, is the current rock mass porosity, is the current rock mass thickness; is the regression coefficient of the rock mass density, is the regression coefficient of the rock mass porosity, is the regression coefficient of the rock mass thickness. On the basis of , let .

4. The deep stope structure safety evaluation method based on the combined drive of data and knowledge model according to claim 2, characterized in that: Based on the environmental analysis model, the expression of the environmental risk coefficient is as follows: ； Among them, is the environmental risk coefficient of the current deep stope; In the formula, is the interaction term of the CO concentration and temperature in the current deep stope, is the interaction term of the humidity and groundwater level in the current deep stope; is the regression coefficient of the interaction term of CO concentration and temperature, is the regression coefficient of the interaction term of humidity and groundwater level. On the basis of , let .

5. The deep stope structure safety evaluation method jointly driven by data and knowledge model according to claim 1, characterized in that: Perform data processing and correlation analysis on the current geological condition coefficient, environmental risk coefficient, and deformation response coefficient of the deep stope to generate a comprehensive evaluation coefficient. The formula is as follows: ； Among them, is the comprehensive evaluation coefficient of the current deep stope, is the geological condition coefficient of the current deep stope, is the environmental risk coefficient of the current deep stope, is the deformation response coefficient of the current deep stope; In the formula, , and are the weight coefficients of the geological condition coefficient, the environmental risk coefficient, and the deformation response coefficient, respectively, and , and are determined by the analytic hierarchy process.

6. The deep stope structure safety evaluation method based on the combined drive of data and knowledge model according to claim 5, characterized in that: The specific process of step S6 is as follows: When , the current deep stope structure meets the safety standards; When , the current deep stope structure does not meet the safety standards; It represents the safety threshold of the deep stope structure and indicates the critical value for the safety of the deep stope structure.

Images