Cloud model-based mine deep underground engineering safety resilience evaluation method and system

Abstract

This invention provides a method and system for evaluating the safety resilience of deep underground mining engineering based on a cloud model, belonging to the field of safety management technology. The method includes: constructing an evaluation index system for the safety resilience of deep underground mining engineering, comprising basic layer indicators, system layer indicators, and target layer indicators; determining the weights of each indicator through system dynamics; using a cloud model to process the uncertainty of each indicator data to generate an indicator cloud map; obtaining a comprehensive evaluation cloud through the synthesis rules of the cloud model; and combining it with evaluation level standards to achieve a quantitative measurement and evaluation of the safety resilience of deep underground mining engineering. The system includes: an index system construction module, a weight calculation module, a cloud model generation module, a cloud distance measurement module, and a safety resilience level determination module. This invention solves the problems of insufficient handling of uncertainty and strong subjectivity in traditional evaluation methods, achieving accurate quantitative determination of resilience levels and improving the accuracy and reliability of evaluation results.

Description

Technical Field

[0001] This invention relates to the field of safety management technology, and in particular to a method and system for evaluating the safety resilience of deep underground engineering projects in mines based on a cloud model. Background Technology

[0002] Deep underground engineering in mines faces a complex environment characterized by high stress, high ground temperature, high osmotic pressure, and strong mining disturbances, leading to frequent dynamic disasters such as rock bursts and rock bursts. Traditional safety assessment theories, primarily based on static prevention and control, are increasingly showing limitations in addressing these sudden and systemic engineering risks. Therefore, introducing the concept of "safety resilience" into the field of deep mining engineering and constructing a systematic evaluation index system to quantitatively assess the post-disaster absorption, adaptation, and rapid recovery capabilities of engineering systems has significant theoretical and practical value.

[0003] Existing methods for evaluating system safety resilience are mainly divided into qualitative and quantitative categories. Qualitative methods, such as the Delphi method, rely on expert experience, are highly subjective, and lack objectivity and consistency. Quantitative methods, such as linear weighting and TOPSIS, while having clear structures, often treat uncertain index values ​​as definite points, ignoring data acquisition errors and the inherent ambiguity of concepts, leading to overly absolute evaluation results. Furthermore, although cloud models have been introduced to handle uncertainty, existing cloud model distance measurement methods mainly rely on visual observation of cloud map positions, lacking precise quantitative calculations, especially when cloud map shapes are complex or distances are close, making it difficult to accurately determine resilience levels. Summary of the Invention

[0004] The purpose of this invention is to provide a cloud model-based method and system for evaluating the safety and toughness of deep underground engineering in mines. This invention addresses the core technical problems in existing technologies, such as the strong subjectivity in weight determination, insufficient handling of uncertainty in evaluation results, and reliance on visual observation for grade determination, resulting in low accuracy. This improves the accuracy and reliability of safety and toughness evaluation for deep underground engineering in mines.

[0005] To achieve the above objectives, this invention proposes a method for evaluating the safety and toughness of deep underground engineering in mines based on a cloud model, comprising the following steps:

[0006] Step S1: Construct a safety resilience evaluation index system for deep underground engineering in mines. The index system includes three levels of indicators: target layer, criterion layer, and indicator layer. The target layer is the comprehensive evaluation result of the safety resilience of deep underground engineering in mines. The criterion layer includes the main structure subsystem, spatial condition subsystem, infrastructure subsystem, emergency management subsystem, and human factors subsystem. The indicator layer is the specific evaluation indicators under the system layer contained in the criterion layer.

[0007] Step S2: By analyzing the mutual influence relationship between evaluation indicators, construct a system dynamics model, conduct sensitivity analysis on the indicators, and determine the weight of the evaluation indicators. The weight is determined based on the change in safety resilience value caused by a 20% increase or decrease in the value of the evaluation indicator.

[0008] Step S3: Obtain the indicator data of the indicator layer, and use the reverse cloud generator BGG to convert the indicator data into indicator cloud digital features characterized by expectation, entropy and hyperentropy; based on the weight of the indicator, perform layer-by-layer weighted aggregation of the cloud digital features of the indicator layer to calculate the cloud digital features of the criterion layer and the target layer, and use the forward cloud generator FCG to generate a comprehensive evaluation cloud graph.

[0009] Step S4: Propose a cloud distance measurement method to calculate the cloud distance between the comprehensively evaluated cloud and the preset standard-level clouds. The cloud distance is calculated based on the Euclidean distance of cloud droplets in a two-dimensional feature space and the topological similarity of the cloud droplet distribution. The Euclidean distance measurement algorithm for clouds can be expressed as:

[0010] Input: Two cloud models , ;

[0011] Output: Distance between the two cloud models ;

[0012] in: For cloud models Expectations For cloud models The expectation represents the central value of a certain qualitative concept, that is, the centroid of the cloud droplet distribution in the domain space; For cloud models entropy, For cloud models Entropy, representing a measure of the uncertainty of a qualitative concept, reflects the degree of dispersion of cloud droplets. The greater the entropy, the more vague the concept and the wider its scope. For cloud models hyperentropy, For cloud models Hyperentropy, representing a measure of the uncertainty of entropy, also known as "entropy of entropy," reflects the degree of cloud droplet cohesion.

[0013] Step S5: Based on the calculated cloud distance between the empirical project cloud and the standard level clouds, determine the safety toughness level of the deep underground engineering in the mine according to the principle of minimum distance.

[0014] Preferably, in step S2, the initial values ​​of each index layer in the system dynamics model fluctuate within a range of ±20%; the changing trend of the target layer safety toughness value is observed, the absolute value of the slope of the index sensitivity curve is calculated, and the absolute value of the slope is used as the basis for weighting. The larger the absolute value of the slope, the greater the weight of the corresponding index.

[0015] Preferably, in step S3, the inverse cloud generator (BGG) is used to convert the index data into index cloud digital features characterized by expectation, entropy, and hyperentropy. The BGG algorithm can be expressed as:

[0016] Input: Sample size ;

[0017] Output: Expected ,entropy hyperentropy ;

[0018] Preferably, in step S3, a comprehensive evaluation cloud image is generated using a forward cloud generator (FCG). The forward cloud generator (FCG) algorithm can be expressed as:

[0019] Input: Expectation ,entropy hyperentropy ;

[0020] Output: A cloud droplet and Corresponding degree of certainty , For sample number, .

[0021] Preferably, step S4 specifically includes the following steps:

[0022] Step S41: Input two cloud models , Two cloud models are generated by the forward cloud generator FCG. Each cloud droplet forms two cloud droplet sets;

[0023] Step S42: Sort the cloud droplets in the two cloud droplet sets in ascending order of their x-coordinates;

[0024] Step S43: Filter and retain those falling within the interval Cloud droplets inside;

[0025] Step S44: Let the number of cloud droplets for the two filtered cloud models be respectively and To unify the droplet count between two droplet sets, the droplet count with the smaller count is used as the unified droplet count. The droplets are then sorted by their x-coordinate from smallest to largest and stored in a set. and middle;

[0026] Step S45: Calculate the two sets in sequence. and The average distance between cloud droplets is given by the formula:

[0027] ;

[0028] in, Let be the average distance between corresponding cloud droplets in the two sets. For set The Middle The x-coordinate of each cloud droplet For set The Middle The x-coordinate of each cloud droplet For set The Middle The membership degree value of each cloud droplet. For set The Middle The membership degree value of each cloud droplet. The number of the cloud droplet. To unify the number of cloud droplets after two cloud droplet sets.

[0029] Preferably, in step S5, each standard level cloud is generated based on the safety resilience evaluation level and its corresponding score range, and the corresponding score ranges are (0,30], [30,60), [60,75), [75,90), [90,100].

[0030] Preferably, the expected value, entropy, and hyperentropy of clouds at each standard level are obtained by converting the score interval, as shown in the formula:

[0031] ;

[0032] in, For the first The expectation of a cloud with a security resilience standard level. For the first The minimum value within the range of safety resilience evaluation score levels. For the first The maximum value within the range of safety resilience evaluation score. For the first The entropy of a cloud at a security resilience standard level. For the first The hyperentropy of a cloud with a security resilience standard level. This is a preset constant for hyperentropy.

[0033] This invention also proposes a cloud model-based safety resilience evaluation system for deep underground engineering in mines, comprising: an index system construction module, a weight calculation module, a cloud model generation module, a cloud distance measurement module, and a safety resilience level determination module, wherein:

[0034] The indicator system construction module includes a three-level indicator configuration unit, an indicator association definition unit, and an indicator library storage unit;

[0035] The weight calculation module includes a system dynamics modeling unit, a sensitivity analysis unit, and a weight normalization unit;

[0036] The cloud model generation module includes a standard cloud generation unit, an indicator cloud generation unit, and a comprehensive cloud fusion unit;

[0037] The cloud distance measurement module includes a cloud droplet filtering unit, a cloud droplet matching unit, and an improved Euclidean distance calculation unit;

[0038] The safety resilience level determination module includes a minimum distance matching unit.

[0039] Preferably, in the weight calculation module, the sensitivity analysis unit is configured as follows: the initial values ​​of each index layer in the system dynamics model are set to fluctuate within a range of ±20%, the changing trend of the target layer safety toughness value is observed, the absolute value of the slope of the sensitivity curve of each index is calculated, and the absolute value of the slope is sent to the weight normalization unit as the basis for weighting.

[0040] Preferably, in the cloud distance measurement module, the cloud droplet filtering unit is configured to: filter and retain droplets falling within the interval based on the 3σ rule. The cloud droplets within the set are configured as follows: The number of cloud droplets in two sets is unified, the smaller number of droplets is used as the unified number, and the droplets in the two sets are sorted by their x-coordinates from smallest to largest before being paired one-to-one; The improved Euclidean distance calculation unit is configured to calculate the first... The average distance between cloud droplets.

[0041] Therefore, this invention proposes a method and system for evaluating the safety and toughness of deep underground engineering in mines based on a cloud model, with the following beneficial effects:

[0042] (1) This invention unifies the randomness and fuzziness in the safety resilience evaluation by using the expectation, entropy and hyperentropy of the cloud model. It utilizes the reverse cloud generator to integrate expert knowledge and the forward cloud generator to soften the monitoring data, so that subjective and objective information are organically integrated under the same framework, overcoming the shortcomings of traditional methods that are highly subjective and ignore data fluctuations.

[0043] (2) This invention constructs a system dynamics model and determines the weights through single-factor sensitivity analysis. The index values ​​are set to fluctuate within a range of ±20%, and the absolute value of the slope of the sensitivity curve is used as the basis for weighting, so that the weights truly reflect the sensitivity of the indexes under the interaction of system risks, and the determination of weights is more objective and scientific.

[0044] (3) This invention proposes an improved cloud distance measure based on the comparison of two clouds. It considers the Euclidean distance and membership information of cloud droplets. By filtering cloud droplets through the 3σ rule, sorting and pairing, and calculating the weighted average distance, it achieves accurate quantification of cloud similarity, overcomes the limitations of visual observation, and significantly improves the accuracy and robustness of hierarchical classification. Attached Figure Description

[0045] Figure 1 A flowchart of a cloud model-based method for evaluating the safety and resilience of deep underground engineering in mines;

[0046] Figure 2 A dynamic stock flow diagram for a safety toughness evaluation system of deep underground engineering in mines;

[0047] Figure 3 An evaluation cloud map is provided for a cloud model-based method to evaluate the safety and resilience of deep underground engineering in mines.

[0048] Figure 4 This provides a comprehensive cloud map for a cloud model-based method of evaluating the safety and resilience of deep underground engineering in mines. Detailed Implementation

[0049] The technical solution of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0050] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0051] Example 1

[0052] like Figure 1 As shown, this invention provides a method for evaluating the safety and toughness of deep underground engineering in mines based on a cloud model. This method is used to handle uncertain information and determine the toughness level, and includes the following steps:

[0053] Step S1: Construct a safety resilience evaluation index system for deep underground engineering in mines. The index system includes three levels of indicators: target layer, criterion layer, and indicator layer. The target layer is the comprehensive evaluation result of the safety resilience of deep underground engineering in mines. The criterion layer includes the main structure subsystem, spatial condition subsystem, infrastructure subsystem, emergency management subsystem, and human factors subsystem. The indicator layer consists of specific evaluation indicators under the system layer contained in the criterion layer. The evaluation system, through five indicator subsystems, including the main structure subsystem, spatial condition subsystem, infrastructure subsystem, emergency management subsystem, and human factors subsystem, forces evaluators to consider both disaster control and personnel response dimensions simultaneously. This allows for a more systematic and comprehensive characterization of the overall resilience level of mining engineering in fire scenarios. A space with strong fire resistance but poor evacuation capacity will inevitably have a low overall resilience level, and vice versa.

[0054] Step S2: By analyzing the mutual influence relationship between evaluation indicators, construct a system dynamics model, conduct sensitivity analysis on the indicators, and determine the weights of the evaluation indicators. The weights are determined based on the change in safety resilience value caused by a 20% increase or decrease in the value of the evaluation indicator. Observe the changing trend of the safety resilience value of the target layer, calculate the absolute value of the slope of the indicator sensitivity curve, and use the absolute value of the slope as the basis for weighting. The larger the absolute value of the slope, the greater the weight of the corresponding indicator.

[0055] Step S3: Obtain the indicator data of the indicator layer, and use the Inverse Cloud Generator (BGG) to convert the indicator data into indicator cloud digital features characterized by expectation, entropy, and hyperentropy. The BGG algorithm can be expressed as:

[0056] Input: Sample size ;

[0057] Output: Expected ,entropy hyperentropy ;

[0058] The steps of the BGG algorithm are as follows:

[0059] Step W1: Calculate the sample mean, first-order sample absolute central moment, and sample variance of a set of sample data. The formula is:

[0060] ;

[0061] ;

[0062] ;

[0063] in, The sample mean of a set of sample data. The number of cloud droplet samples. For cloud droplets, The first-order sample absolute central moments, This represents the sample variance.

[0064] Step W2: Calculate the expected value, using the following formula:

[0065] ;

[0066] Step W3: Calculate the entropy, using the following formula:

[0067] ;

[0068] Step W4: Calculate the hyperentropy, using the following formula:

[0069] ;

[0070] Based on the weights of the indicators, the cloud digital features of the indicator layer are weighted and aggregated layer by layer to calculate the cloud digital features of the criterion layer and the target layer. A comprehensive evaluation cloud image is then generated using a forward cloud generator (FCG). The forward cloud generator (FCG) algorithm can be expressed as:

[0071] Input: Expectation ,entropy hyperentropy ;

[0072] Output: A cloud droplet and Corresponding degree of certainty , For sample number, ;

[0073] The steps of the FCG algorithm are as follows:

[0074] Step Z1: Generate with For the expected value, The formula for a normally distributed random number with variance is:

[0075] ;

[0076] in, For temporary entropy, It is a function for generating normally distributed random numbers;

[0077] Step Z2: Generate with For the expected value, The formula for a normally distributed random number with variance is:

[0078] ;

[0079] Step Z3: Calculate the certainty of the cloud droplets using the following formula:

[0080] ;

[0081] in, For cloud droplets The corresponding degree of certainty (membership);

[0082] Step Z4: Has determinism of Generate a cloud droplet within the domain;

[0083] Step Z5: Repeat the above steps until a certain amount of cloud droplets are generated and converge to form a cloud pattern;

[0084] Step S4: Propose a cloud distance measurement method to calculate the cloud distance between the comprehensively evaluated cloud and the preset standard-level clouds. The cloud distance is calculated based on the Euclidean distance of cloud droplets in a two-dimensional feature space and the topological similarity of the cloud droplet distribution. The Euclidean distance measurement algorithm for clouds can be expressed as:

[0085] Input: Two cloud models , ;

[0086] Output: Distance between the two cloud models ;

[0087] Specifically, the following steps are included:

[0088] Step S41: Input two cloud models , Two cloud models are generated by the forward cloud generator FCG. Each cloud droplet forms two cloud droplet sets;

[0089] Step S42: Sort the cloud droplets in the two cloud droplet sets in ascending order of their x-coordinates;

[0090] Step S43: Filter and retain those falling within the interval Cloud droplets inside;

[0091] Step S44: Let the number of cloud droplets for the two filtered cloud models be respectively and To unify the droplet count between two droplet sets, the droplet count with the smaller count is used as the unified droplet count. The droplets are then sorted by their x-coordinate from smallest to largest and stored in a set. and middle;

[0092] Step S45: Calculate the two sets in sequence. and The average distance between cloud droplets is given by the formula:

[0093] ;

[0094] in, Let be the average distance between corresponding cloud droplets in the two sets. For set The Middle The x-coordinate of each cloud droplet For set The Middle The x-coordinate of each cloud droplet For set The Middle The membership degree value of each cloud droplet. For set The Middle The membership degree value of each cloud droplet. The number of the cloud droplet. To unify the number of cloud droplets after two cloud droplet sets.

[0095] Step S5: Based on the calculated cloud distances between the empirical project cloud and the standard-level clouds, and according to the principle of minimum distance, determine the safety resilience level of the deep underground engineering in the mine. Each standard-level cloud is generated based on its safety resilience evaluation level and its corresponding score range, which are (0,30], [30,60), [60,75), [75,90), and [90,100], respectively. The expectation, entropy, and hyperentropy of each standard-level cloud are obtained by converting the score ranges, using the following formula:

[0096] ;

[0097] in, For the first The expectation of a cloud with a security resilience standard level. For the first The minimum value within the range of safety resilience evaluation score levels. For the first The maximum value within the range of safety resilience evaluation score. For the first The entropy of a cloud at a security resilience standard level. For the first The hyperentropy of a cloud with a security resilience standard level. This is a preset constant for hyperentropy.

[0098] Example 2

[0099] This invention provides a cloud model-based safety and resilience evaluation system for deep underground engineering in mines, comprising: an index system construction module, a weight calculation module, a cloud model generation module, a cloud distance measurement module, and a safety and resilience level determination module, wherein:

[0100] The indicator system construction module is used to build an indicator system for evaluating the safety and resilience of deep underground mining engineering, which includes a target layer, a criterion layer, and an indicator layer. It outputs the indicator system to the weight calculation module and includes a three-level indicator configuration unit, an indicator association definition unit, and an indicator library storage unit. The three-level indicator configuration unit loads or edits the target layer, criterion layer, and indicator layer; the indicator association definition unit defines the causal relationships and hierarchical relationships between indicators; and the indicator library storage unit stores the set indicator names, measurement methods, and standard thresholds.

[0101] The weight calculation module is used to construct the system dynamics model and determine the weights of each indicator through single-factor sensitivity analysis, eliminating subjective weighting bias. It outputs the weight vector to the cloud model generation module, and includes a system dynamics modeling unit, a sensitivity analysis unit, and a weight normalization unit. The system dynamics modeling unit constructs stock / rate / auxiliary variable equations and draws causal feedback loops; the sensitivity analysis unit calculates the sensitivity slope for ±20% fluctuation of the indicators; and the weight normalization unit normalizes the slope to obtain the final weights of the indicators between 0 and 1.

[0102] The cloud model generation module uses a reverse cloud generator to convert indicator data into cloud digital features, and a forward cloud generator to generate comprehensive evaluation cloud graphics. It outputs all cloud model digital features and cloud droplet data, and outputs cloud data to the cloud distance measurement module. It includes a standard cloud generation unit, an indicator cloud generation unit, and a comprehensive cloud fusion unit. The standard cloud generation unit generates five levels of standard clouds (poor, weak, medium, good, and excellent) based on the 3σ rule; the indicator cloud generation unit generates indicator-level cloud digital features through forward and reverse cloud generators; and the comprehensive cloud fusion unit performs hierarchical weighted fusion to generate comprehensive clouds at the criterion and target levels.

[0103] The cloud distance measurement module is used to calculate the cloud distance between the comprehensive evaluation cloud and the standard-level cloud using cloud distance measurement methods, quantify the similarity between the empirical cloud and each standard cloud, and output the distance value to the safety resilience level determination module. It includes a cloud droplet screening unit, a cloud droplet matching unit, and an improved Euclidean distance calculation unit. The cloud droplet screening unit filters valid cloud droplets according to the 3σ rule and standardizes the sample size; the cloud droplet matching unit pairs cloud droplets from the empirical cloud and standard clouds according to their x-coordinates; and the improved Euclidean distance calculation unit calculates the weighted average distance of the cloud droplets based on their x-coordinates and membership degrees.

[0104] The safety toughness level determination module is used to determine the safety toughness level of deep underground engineering in mines based on the minimum distance principle, and outputs the final safety toughness level and risk warning, including a minimum distance matching unit. The minimum distance matching unit determines the toughness level according to the minimum distance principle.

[0105] The invention will be further illustrated below through specific implementation examples.

[0106] Taking a kilometer-deep mine as an example, a safety and toughness evaluation method for deep underground engineering in mines based on a cloud model is implemented. The specific steps are as follows:

[0107] Step 1: Construct an indicator system and calculate indicator weights. This system is developed from five aspects: main structure, spatial conditions, infrastructure, emergency management, and human factors. This forms a complete logical framework covering "hardware foundation - environmental constraints - system functions - management mechanisms - human core." The main structure directly bears the pressure of the surrounding rock and determines the overall stability of the project. Spatial conditions define the physical boundaries of the working environment, affecting ventilation, access, and personnel safety. Infrastructure ensures the continuous operation of lifeline systems such as ventilation, drainage, and power supply. Emergency management focuses on post-disaster response speed and recovery capabilities, which is the key difference between resilience and traditional safety assessments. Human factors are present throughout, including daily operations and crisis decision-making; human awareness and behavior are core variables that trigger or interrupt the accident chain. These five aspects are interconnected and progressively enhance each other, jointly depicting the mine system's ability to resist impact and recover quickly in complex environments such as deep high stress and strong disturbances. Therefore, they become a necessary dimension for evaluating its safety resilience. The established indicator system is shown in Table 1.

[0108] Table 1 Indicator System

[0109]

[0110]

[0111] Step 2: Determine the weights of each indicator using the system dynamics model:

[0112] The interaction between various evaluation indicators is analyzed, and a system dynamics model is constructed. The system dynamics model describes the driving relationship between variables through causal chains to reveal the dynamic behavior inside the system. Inevitably, there is a special type of variable in the model. This special variable only changes due to the influence of the external environment, but does not form causal feedback with other variables. This type of variable is usually called a constant.

[0113] To avoid numerical overflow caused by large differences in the values ​​of different constants, the stock flow diagram is as follows: Figure 2As shown in the figure, the values ​​of the five constants are all derived from the indicator data of the actual project. Among them, the rate variable is the basis for driving the change of the level variable, and each level variable corresponds to a rate variable. The equation of the rate variable is composed of the combination of the value of the indicator or criterion level indicator and its corresponding standard value. The value of the level variable is affected by two factors: the initial value of the level variable itself and the effect of other variables on it. The method for determining the initial value of the level variable is the same as the method for estimating the initial value of the constant.

[0114] Apart from the constants, rate variables, and level variables that require precise definition, all other variables are auxiliary variables. This model sets up a total of 18 auxiliary variables, corresponding to the 18 indicator layers in the indicator system. The equations for the auxiliary variables are expressed in a weighted summation form.

[0115] To quantify the impact of various indicators on safety resilience under the interaction of risks, this study uses single-factor sensitivity analysis to determine the indicator weights. By setting the initial values ​​of the criterion layer and indicator layer to fluctuate within ±20%, the changing trend of the safety resilience value at the target layer was observed. Considering that changes in all indicators lead to a positive increase in resilience value, the absolute value of the slope of each indicator's sensitivity curve was used as the basis for weighting. The larger the absolute value of the slope, the higher the sensitivity of the indicator under the interaction of system risks, and the greater its weight. The constructed system dynamics stock-flow diagram is shown below. Figure 2 As shown, by constructing a stock-flow diagram and a causal feedback loop, system dynamics can characterize the nonlinear, multivariable, high-order, and time-delay features of the system, reveal the dynamic change pattern of system behavior over time, and make up for the shortcomings of static analysis methods.

[0116] Step 3: Generate an evaluation cloud:

[0117] The evaluation of the safety resilience of deep underground engineering in mines is conducted through actual measurements or expert scoring, thus exhibiting a degree of randomness and ambiguity. The first manifestation of this is the randomness of scoring, as different survey subjects consider different dimensions, resulting in different scores. The second manifestation is the ambiguity of scoring, with evaluation scores fluctuating within a certain range. Therefore, the safety resilience evaluation is graded, and corresponding score ranges are set.

[0118] The safety resilience evaluation of deep underground engineering in mines is divided into five levels: Excellent, Good, Medium, Weak, and Poor, with corresponding score ranges shown in Table 2. The "Poor" and "Weak" ranges are relatively wide (both spanning 30 points): This is because, while accurately distinguishing between "extremely poor" and "very poor" is important for evaluators, the more crucial goal is to differentiate between systems above and below the passing grade. 60 points is a significant threshold, thus allowing more leeway for the two levels below. The ranges for "Medium," "Good," and "Excellent" are narrower (spanning 15 or 10 points): This is because for systems that have already met basic qualification requirements, a more refined differentiation of their performance levels is needed. In particular, the 10-point difference between "Excellent" and "Good" signifies a significant performance improvement and investment difference, requiring a smaller range to sensitively capture this difference, reflecting the evaluation system's ability to identify high-level performance.

[0119] Table 2 Safety toughness evaluation levels and corresponding score ranges

[0120]

[0121] According to normal cloud The formula for converting the numerical characteristics of the safety resilience evaluation level is as follows:

[0122] ;

[0123] in, and These represent the minimum and maximum values ​​of the interval, respectively. Since the cloud model exhibits both fuzziness and randomness, the parts with a membership degree greater than 50% should be clearer, while those with a membership degree less than 50% should be intersecting and fuzzy. After multiple trials, the optimal choice was determined. The cloud model has moderate fuzziness and relatively accurate membership.

[0124] Therefore, based on the above formula, the score range is converted into the digital features of the cloud model, as shown in Table 3:

[0125] Table 3 Safety resilience evaluation levels and corresponding score ranges and numerical characteristics

[0126]

[0127] Using a reverse cloud generator to transform quantitative data into data generated by the desired cloud. ,entropy and hyperentropy Descriptive language used to express meaning.

[0128] Based on the measurement method, expert scoring, and standardization approach, the cloud digital characteristics and weights of the indicators at the indicator layer are obtained through a system dynamics model, as shown in Table 4:

[0129] Table 4 Cloud Digital Characteristics of Indicator Layers

[0130]

[0131] The comprehensive cloud of the empirical project is solved in the order of "indicator layer → criterion layer → target layer", and the expected value is calculated sequentially. ,entropy and hyperentropy The three numerical features are aggregated layer by layer from bottom to top using a unique weighted fusion method. Traditional linear weighting can lead to the averaging of uncertainty, thus overestimating security. In the final generated cloud map, if a system's cloud map becomes very thick and blurry due to a high-risk, high-uncertainty indicator, the evaluator can immediately see that the system's reliability is questionable, or that the confidence level of its evaluation results is low, serving as a strong risk warning.

[0132] expect ,entropy and hyperentropy The specific calculation formula is as follows:

[0133] ;

[0134] in, For the first The overall weight of each evaluation indicator.

[0135] The cloud digital characteristics of the indicator layer can be used to calculate the criteria layer indicators, and then the cloud digital characteristics of the target layer can be further calculated. The specific results are shown in Table 5.

[0136] Table 5. Cloud digital characteristics of the target layer

[0137]

[0138] A safety resilience assessment cloud graph was generated using a forward cloud generator, consisting of 1000 cloud droplets. Cloud droplets with scores ranging from (0 to 100) were retained. Figure 3 As shown.

[0139] Step 4, Cloud Distance Measurement:

[0140] An improved cloud distance measurement method is used to calculate the cloud distance between the comprehensive evaluation cloud and clouds of each standard level. This method overcomes the limitations of traditional visual observation and includes the following steps:

[0141] Step 4.1: Input two cloud models , Two cloud models are generated by the forward cloud generator FCG. Each cloud droplet forms two cloud droplet sets;

[0142] Step 4.2: Sort the cloud droplets in the two cloud droplet sets in ascending order of their x-coordinates;

[0143] Step 4.3: Filter and retain those falling within the interval Cloud droplets inside;

[0144] Step 4.4: Let the number of cloud droplets for the two filtered cloud models be respectively... and To unify the droplet count between two droplet sets, the droplet count with the smaller count is used as the unified droplet count. The droplets are then sorted by their x-coordinate from smallest to largest and stored in a set. and middle;

[0145] Step 4.5: Calculate the two sets in sequence. and The average distance between cloud droplets is given by the formula:

[0146] ;

[0147] in, Let be the average distance between corresponding cloud droplets in the two sets. For set The Middle The x-coordinate of each cloud droplet For set The Middle The x-coordinate of each cloud droplet For set The Middle The membership degree value of each cloud droplet. For set The Middle The membership degree value of each cloud droplet. The number of the cloud droplet. To unify the number of cloud droplets after two cloud droplet sets.

[0148] Step 5: Determining the safety toughness level:

[0149] Based on the calculated cloud distances between the empirical project cloud and the clouds of each standard level, the safety resilience level is determined according to the principle of minimum distance. That is, the distance between the cloud model representing the actual performance of the project and the predefined cloud models representing different resilience level standards are calculated respectively. The distance with the smallest value points to the most similar standard level, and therefore this level is determined as the final safety resilience level of the empirical project.

[0150] Based on the cloud digital characteristics in Table 5, a security resilience cloud graph of the empirical project was drawn. Figure 4 The entropy of the cloud droplets formed by the cloud model of this project is... A score of 5.8655 indicates that the respondents' ratings were relatively consistent, with little difference in their understanding of the issues; Expectations The value is 82.1724. Analysis of the empirical project cloud graph shows that its safety resilience level is "Good". The distance between it and the "Medium" and "Good" safety resilience evaluation cloud graphs is calculated using the formula:

[0151] ;

[0152] because Based on the principle of minimum distance, the overall safety resilience of this mining project is at a good level.

[0153] This invention also provides an intelligent measurement system for the safety resilience cloud model of deep underground engineering in mines based on MATLAB software, which can be used to conveniently realize intelligent safety resilience assessment of deep underground engineering in mines.

[0154] It is worth noting that all contents not described in detail in this invention are existing technologies and are well known to those skilled in the art.

[0155] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for evaluating the safety and resilience of deep underground engineering in mines based on a cloud model, characterized in that, Includes the following steps: Step S1: Construct a safety resilience evaluation index system for deep underground engineering in mines. The index system includes three levels of indicators: target layer, criterion layer, and indicator layer. The target layer is the comprehensive evaluation result of the safety resilience of deep underground engineering in mines. The criterion layer includes the main structure subsystem, spatial condition subsystem, infrastructure subsystem, emergency management subsystem, and human factors subsystem. The indicator layer is the specific evaluation indicators under the system layer contained in the criterion layer. Step S2: By analyzing the mutual influence relationship between evaluation indicators, construct a system dynamics model, conduct sensitivity analysis on the indicators, and determine the weight of the evaluation indicators. The weight is determined based on the change in safety resilience value caused by a 20% increase or decrease in the value of the evaluation indicator. Step S3: Obtain the indicator data of the indicator layer, and use the reverse cloud generator BGG to convert the indicator data into indicator cloud digital features characterized by expectation, entropy and hyperentropy; based on the weight of the indicator, perform layer-by-layer weighted aggregation of the cloud digital features of the indicator layer to calculate the cloud digital features of the criterion layer and the target layer, and use the forward cloud generator FCG to generate a comprehensive evaluation cloud graph. Step S4: Based on the 3σ rule, cloud droplets are screened. Combining the Euclidean distance of cloud droplets in the two-dimensional feature space with the topological similarity of cloud droplet distribution, the cloud distance between the comprehensive evaluation cloud and each preset standard level cloud is calculated. The Euclidean distance measurement algorithm for clouds can be expressed as: Input: Two cloud models , ; Output: Distance between the two cloud models ; in, For cloud models Expectations For cloud models entropy, For cloud models hyperentropy, For cloud models Expectations For cloud models entropy, For cloud models hyperentropy; Step S5: Based on the calculated cloud distance between the empirical project cloud and the standard level clouds, determine the safety toughness level of the deep underground engineering in the mine according to the principle of minimum distance.

2. The method for evaluating the safety and resilience of deep underground engineering in mines based on a cloud model according to claim 1, characterized in that: In step S2, the initial values ​​of each index layer in the system dynamics model fluctuate within a range of ±20%; the changing trend of the target layer safety toughness value is observed, the absolute value of the slope of the index sensitivity curve is calculated, and the absolute value of the slope is used as the basis for weighting. The larger the absolute value of the slope, the greater the weight of the corresponding index.

3. The method for evaluating the safety and resilience of deep underground engineering in mines based on a cloud model according to claim 1, characterized in that: In step S3, the inverse cloud generator (BGG) is used to convert the index data into index cloud digital features characterized by expectation, entropy, and hyperentropy. The BGG algorithm is expressed as follows: Input: Sample size ; Output: Expected ,entropy hyperentropy .

4. The method for evaluating the safety and resilience of deep underground engineering in mines based on a cloud model according to claim 1, characterized in that: In step S3, a comprehensive evaluation cloud image is generated using a forward cloud generator (FCG). The forward cloud generator (FCG) algorithm is expressed as follows: Input: Expectation ,entropy hyperentropy ; Output: A cloud droplet and Corresponding degree of certainty , For sample number, ; The formula for calculating the certainty of cloud droplets is: ； in, For cloud droplets The corresponding degree of certainty, For cloud droplets, This is the temporary entropy.

5. The method for evaluating the safety and resilience of deep underground engineering in mines based on a cloud model according to claim 1, characterized in that: Step S4 specifically includes the following steps: Step S41: Input two cloud models , Two cloud models are generated by the forward cloud generator FCG. Each cloud droplet forms two cloud droplet sets; Step S42: Sort the cloud droplets in the two cloud droplet sets in ascending order of their x-coordinates; Step S43: Filter and retain those falling within the interval Cloud droplets inside; Step S44: Let the number of cloud droplets for the two filtered cloud models be respectively and To unify the droplet count between two droplet sets, the droplet count with the smaller count is used as the unified droplet count. The droplets are then sorted by their x-coordinate from smallest to largest and stored in a set. and middle; Step S45: Calculate the two sets in sequence. and The average distance between cloud droplets is given by the formula: ； in, Let be the average distance between corresponding cloud droplets in the two sets. For set The Middle The x-coordinate of each cloud droplet For set The Middle The x-coordinate of each cloud droplet For set The Middle The membership degree value of each cloud droplet. For set The Middle The membership degree value of each cloud droplet. The number of the cloud droplet. To unify the number of cloud droplets after the two cloud droplet sets.

6. The method for evaluating the safety and resilience of deep underground engineering in mines based on a cloud model according to claim 1, characterized in that: In step S5, each standard level cloud is generated based on the safety resilience evaluation level and its corresponding score range, which are (0,30], [30,60), [60,75), [75,90), and [90,100].

7. The method for evaluating the safety and resilience of deep underground engineering in mines based on a cloud model, as described in claim 6, is characterized in that: The expected value, entropy, and hyperentropy of clouds at each standard level are obtained by converting the score interval, as shown in the formula: ； in, For the first The expectation of a cloud with a security resilience standard level. For the first The minimum value within the range of safety resilience evaluation score. For the first The maximum value within the range of safety resilience evaluation score. For the first The entropy of a cloud at a security resilience standard level. For the first The hyperentropy of a cloud with a security resilience standard level. This is a preset constant for hyperentropy.

8. A cloud model-based safety resilience evaluation system for deep underground mining engineering, used to implement the cloud model-based safety resilience evaluation method for deep underground mining engineering as described in claim 1, characterized in that, include: The system comprises an indicator system construction module, a weight calculation module, a cloud model generation module, a cloud distance measurement module, and a security resilience level determination module, among which: The indicator system construction module includes a three-level indicator configuration unit, an indicator association definition unit, and an indicator library storage unit; The weight calculation module includes a system dynamics modeling unit, a sensitivity analysis unit, and a weight normalization unit; The cloud model generation module includes a standard cloud generation unit, an indicator cloud generation unit, and a comprehensive cloud fusion unit; The cloud distance measurement module includes a cloud droplet filtering unit, a cloud droplet matching unit, and an improved Euclidean distance calculation unit; The safety resilience level determination module includes a minimum distance matching unit.

9. The cloud model-based safety resilience evaluation system for deep underground engineering in mines according to claim 8, characterized in that: In the weight calculation module, the sensitivity analysis unit is configured as follows: set the initial values ​​of each index layer in the system dynamics model to fluctuate within the range of ±20%, observe the changing trend of the target layer safety toughness value, calculate the absolute value of the slope of the sensitivity curve of each index, and send the absolute value of the slope as the basis for weighting to the weight normalization unit.

10. The cloud model-based safety and resilience evaluation system for deep underground engineering in mines according to claim 8, characterized in that: In the cloud distance measurement module, the cloud droplet filtering unit is configured to: filter and retain droplets falling within the interval based on the 3σ rule. The cloud droplets within the set are configured as follows: The number of cloud droplets in two sets is unified, the smaller number of droplets is used as the unified number, and the droplets in the two sets are sorted by their x-coordinates from smallest to largest before being paired one-to-one; The improved Euclidean distance calculation unit is configured to calculate the first... The average distance between cloud droplets.

Images