Metal hydride safety risk multi-level fuzzy evaluation method based on game equilibrium clustering neutralization

Abstract

The invention discloses a metal hydride safety risk multi-level fuzzy evaluation method based on game equilibrium clustering neutralization. The method comprises the following steps: firstly, constructing a natural environment, production process and production personnel management criterion layer system according to a multi-process flow risk, and subdividing specific risk indexes to construct a hierarchical structure; thirdly, organizing experts to compare every two level factors according to a scoring rule to construct a fuzzy judgment matrix, then calculating index weights by using a game theory equilibrium utility function, a clustering neutralization algorithm and a fuzzy analytic hierarchy process, performing normalization, defuzzification by using a gravity center method, solving an average value to obtain a weight vector, and solving a maximum characteristic value; and the weight is ensured to be reasonable through consistency check. And finally, risk index coefficients are calculated and summed to obtain a safety risk score by combining the index weights of all levels and expert scores, and the risk level is determined according to a score interval. The method is systematic, comprehensive, scientific and objective, risk assessment accuracy is effectively improved, and powerful guarantee is provided for industrial safety production.

Description

Technical Field

[0001] This invention relates to the field of safe production technology for metal hydrides, and in particular to a multi-level fuzzy assessment method for the safety risks of metal hydrides based on game equilibrium clustering neutralization. Background Technology

[0002] In the field of safe production of metal hydrides, safety has always been a crucial aspect. With the continuous advancement of hydrogen energy technology and the widespread application of metal hydrides in hydrogen storage and energy storage, the complexity of their production processes is constantly increasing, and various safety risk factors are becoming more diversified and complex. Traditional risk assessment methods often struggle to comprehensively and accurately consider the combined effects of various factors in chemical production, including natural environmental factors, production process factors, and personnel management factors. For example, some factories focus solely on whether equipment is operating normally, neglecting the indirect impact of natural environmental factors such as extreme weather on equipment stability, and personnel management factors such as the increased probability of operational errors due to fatigue. These factors may have limited impact individually, but their interaction in the actual production environment can lead to serious safety accidents, causing huge economic losses to the company and even threatening the lives of employees and the surrounding environment.

[0003] Moreover, existing assessment methods rely on multiple experts to jointly determine the weights of assessment indicators within the theoretical framework when determining the weights of various risk factors. This inevitably leads to inaccurate assessment results due to subjective factors, creating potential risks for companies in developing relevant safety precautions.

[0004] Therefore, there is an urgent need for a comprehensive and scientific method to assess the risks of metal hydride safety production systems, thereby enabling effective risk assessment for industrial safety production systems. Summary of the Invention

[0005] This invention aims to provide a multi-level fuzzy assessment method for the safety risks of metal hydrides based on game equilibrium clustering neutralization. It mainly addresses the problem of insufficient accuracy in the safety production risk assessment results of metal hydrides. It uses a combination of game equilibrium clustering neutralization algorithm and fuzzy hierarchical analysis to reduce the impact of errors generated during the risk assessment process on the results, thereby enabling effective risk assessment for the safety production system.

[0006] To achieve the above objectives, the technical solution adopted by this invention is: a multi-level fuzzy assessment method for the safety risks of metal hydrides based on game equilibrium clustering neutralization, comprising the following steps:

[0007] S1: Environmental risk, production process risk, and production personnel management risk are taken as primary indicators;

[0008] S2: Based on step S1, for each primary indicator, specific risk indicators are subdivided as secondary indicators, and further subdivided to obtain tertiary indicators; these specific indicators constitute the hierarchical indicator layer.

[0009] S3: Integrate experience and professional knowledge to formulate evaluation rules and obtain expert scoring data. Optimize the expert scoring data by integrating experience and professional knowledge based on a clustering neutralization algorithm to initially reduce errors caused by subjective human factors. Then, use a fuzzy evaluation method with cross-correlation analysis to judge the relative importance of factors at each level, and then construct a fuzzy judgment matrix. This evaluation scale, which is combined with the actual situation, fully reflects the relative importance of factors at each level.

[0010] S4: Defuzzify the fuzzy judgment matrix to obtain the judgment matrix. The judgment matrix is ​​then standardized using the Fuzzy Analytic Hierarchy Process (FAHP). By calculating the weights of each indicator in the standardized judgment matrix, the maximum eigenvalue of the standardized judgment matrix is ​​obtained. Finally, a consistency check is performed to ensure the rationality of the standardized judgment matrix. If the consistency ratio is less than the threshold, the weight result is accepted; otherwise, the fuzzy judgment matrix is ​​readjusted. Finally, the weights of each indicator are determined.

[0011] S5: Based on the weights of each level of indicators, let the weight of the first-level indicator be... The weight of the secondary indicators is The weight of the third-level indicators is Calculate the coefficients of n risk indicators. Then, the actual score for the corresponding indicator is recorded as follows: Finally, the quantitative results of each risk indicator are multiplied by their corresponding weights and summed to obtain the safety risk score of the metal hydride safety production system. Based on the range of the safety risk score, the risk level of the metal hydride safety production system is determined, thereby enabling effective risk assessment for industrial safety production systems.

[0012] Preferably, step S1 specifically includes: considering the safety risk characteristics of discrete and continuous mixing, multi-process, batch production and explosive materials in the production of metal hydrides, environmental risk, production process risk and production personnel management risk are taken as primary indicators, and a three-level criterion system of environmental factors, production process factors and production personnel management factors is constructed.

[0013] Preferably, environmental factors include meteorological conditions and geological risks as secondary indicators; production process factors include production process risks, logistics and storage risks, equipment and facility risks, and supporting engineering service risks as secondary indicators; and production personnel management factors include employee skills training factors, personnel operating procedures factors, and personnel health risks as secondary indicators.

[0014] Preferably, meteorological conditions include extreme temperatures as a tertiary indicator; geological risk factors include adverse geological conditions and vibration risk as tertiary indicators; production process risk factors include fire and explosion risk, toxic substance leakage risk, electric shock risk, and reaction indicator control deviation risk as tertiary indicators; logistics and storage risk factors include unsuitable land transportation and storage conditions as tertiary indicators; equipment and facility risk factors include equipment aging risk and maintenance quality risk as tertiary indicators; supporting engineering service risk factors include emission reduction project risks, unstable power supply, and ventilation system failure as tertiary indicators; employee skills training factors include lack of training plans, improper emergency handling, and short-sighted behavior as tertiary indicators; personnel operation standard factors include violations of operating procedures, improper use of protective equipment, and fatigue as tertiary indicators; and personnel health risks include inadequate regular physical examinations as tertiary indicators.

[0015] Preferably, step S3 specifically includes: for the risk indicators in each level of the structure, scoring is performed based on experience and professional knowledge according to the established evaluation rules; the scoring results are initially optimized using a clustering neutralization algorithm; and then the relative importance between factors at each level is evaluated using a fuzzy evaluation method based on cross-correlation analysis. In the set 1-9 scale rule, 1 indicates that two factors are equally important, 9 indicates that one factor is extremely more important than the other, and the intermediate values ​​represent different degrees of relative importance.

[0016] S3-1: The goal of the K-Means clustering neutralization algorithm is to divide the expert rating data into k clusters, and calculate the mean of the k cluster centers to initially reduce the subjective differences among the experts; calculate the expert rating data... To the cluster center European distance The formula is:

[0017]

[0018] in, This represents the score data of the i-th expert. Let m represent the cluster center, m represent the total number of dimensions of the risk indicators, and l represent the l-th risk indicator.

[0019] Will Assigned to the nearest cluster center The cluster it belongs to, namely:

[0020]

[0021] This represents finding the cluster index that minimizes the distance based on the calculated Euclidean distance range, which means dividing the expert data points into clusters according to the calculated distance range; j represents the index of the j-th cluster center, and k represents the total number of cluster centers;

[0022] S3-2: Construct a fuzzy judgment matrix. Each element in the fuzzy judgment matrix is ​​evaluated using a fuzzy evaluation method based on cross-correlation analysis of the corresponding factor. The matrix is ​​reciprocal, with all elements on the main diagonal being 1, to accurately reflect the relative importance of each factor. For an element in the criterion layer, the elements in the index layer below it are... Construct the fuzzy judgment matrix R:

[0023]

[0024] In the formula: As an indicator Compared to The importance of fuzzy membership degree, m=n.

[0025] Preferably, in step S4, fuzzy hierarchical analysis is used to calculate the weights of each indicator by performing specific operations on the fuzzy judgment matrix, so as to obtain a quantitative weight value that reflects the relative importance of each indicator, specifically including:

[0026] S4-1: First, normalize each fuzzy element:

[0027]

[0028]

[0029] in, Triples representing fuzzy membership degrees of importance; , , This represents the three elements of a triplet: the minimum value, the most likely value, and the maximum value. This represents the normalized importance fuzzy weight vector for each fuzzy element;

[0030] Calculate the sum of the fuzzy weights of all factors. :

[0031]

[0032] S4-2: The normalized fuzzy elements are calculated from S4-1. The centroid method is used to defuzzify the fuzzy numbers, and the defuzzified value of each fuzzy number is obtained. :

[0033]

[0034] S4-3: The judgment matrix is ​​composed of the defuzzified values ​​calculated in step S4-2. Then, based on the entropy-value fuzzy hierarchical analysis method, the standardized judgment matrix is ​​calculated and denoted as follows: Then, the weight vector, i.e., the weights, is obtained by calculating the average of each row of the standardized judgment matrix. This is the arithmetic mean of each row; specifically, it includes first obtaining the judgment matrix. :

[0035]

[0036] In the formula: As an indicator Compared to The importance of fuzzy membership degree is defuzzified; then... After standardizing the elements in the matrix using positive indices, a standardized judgment matrix is ​​constructed, denoted as . Positive indicators are standardized using the following formula:

[0037]

[0038] After standardization using positive indices, a standardized judgment matrix is ​​obtained. :

[0039]

[0040] Finally, based on the standardized judgment matrix Further calculation of index weights:

[0041]

[0042] in, This represents the weight of index i. This represents the standardized judgment matrix obtained after normalization using the entropy-based fuzzy hierarchical analysis method. The element in the i-th row and j-th column; n represents the standardized judgment matrix. number of rows;

[0043] S4-4: To address the multi-source risk factor correlation and coupling characteristics and error amplification effect in the primary indicator system, a non-cooperative game optimization framework for collaborative decision-making among multiple stakeholders is constructed. Each relevant decision-maker is modeled as a game participant. By designing multi-objective optimization functions and constraints, the Nash equilibrium solution is calculated to determine the optimal weight combination for the primary indicators, thereby reducing the multi-source risk factor correlation and coupling characteristics and error amplification effect.

[0044]

[0045] in, This represents the game-theoretic utility function of expert j for indicator i. This refers to the attention coefficient of expert j towards indicator i. >0; It is the natural logarithm of the weight of indicator i; It is the penalty coefficient for expert j; It is the weight of indicator i; The initial weights of expert j on index i are obtained through FAHP calculation;

[0046] Optimization of Nash equilibrium solution:

[0047]

[0048] Representing utility function Treatment of Optimization Weights The first-order partial derivatives were calculated. The value of is the optimized weight value obtained after the primary index is calculated using FAHP and the game theory utility function, and it must satisfy ∑ =1, otherwise not applicable;

[0049] S4-5: Based on the standardized judgment matrix The weight of index i is calculated. Then, we can find the largest eigenvalue of the judgment matrix; let the eigenvector of the matrix be... :

[0050]

[0051] Find the largest eigenvalue :

[0052]

[0053] Where R represents the n-order fuzzy judgment matrix constructed by experts, and w is the indicator weight eigenvector, (Rw) i Let be the i-th component of the vector obtained by multiplying matrix R by the weight vector w, where n is the order of the judgment matrix and w is the weight vector. i Let be the weight of the i-th indicator;

[0054] S4-6: Perform a consistency check, which includes calculating the consistency index (CI) and the consistency ratio (CR). If CR is less than 0.1, the weighted result is accepted. If CR is not less than 0.1, the fuzzy judgment matrix is ​​readjusted by adjusting the element values ​​in the fuzzy judgment matrix provided by experts, and the above calculation and check process is repeated until the consistency requirements are met. This ensures the reliability and rationality of the determined weights of each indicator, thereby guaranteeing the scientific validity and effectiveness of the entire risk assessment process. Relevant indicator calculation formulas:

[0055]

[0056] .

[0058] Preferably, the RI values ​​are as follows: RI = 0 for a judgment matrix of order 1; RI = 0 for a judgment matrix of order 2; RI = 0.52 for a judgment matrix of order 3; RI = 0.89 for a judgment matrix of order 4; RI = 1.12 for a judgment matrix of order 5; RI = 1.26 for a judgment matrix of order 6; RI = 1.36 for a judgment matrix of order 7; RI = 1.41 for a judgment matrix of order 8; RI = 1.46 for a judgment matrix of order 9; RI = 1.49 for a judgment matrix of order 10; RI = 1.52 for a judgment matrix of order 11; RI = 1.54 for a judgment matrix of order 12; and RI = 1.56 for a judgment matrix of order 13.

[0059] Preferably, step S5 includes:

[0060] S5-1: Based on the entropy-value fuzzy hierarchical analysis method, the weights of each level of indicators are obtained. The weight of the first-level indicator to which indicator i belongs is: The weight of the secondary indicator to which indicator i belongs is The weight of the third-level indicator corresponding to indicator i is Calculate the coefficients of n risk indicators. :

[0061]

[0062] S5-2: Finally, record the actual scores for the corresponding indicators as follows: The safety risk score of the metal hydride safety production system is obtained by multiplying the quantitative result of each risk indicator by its corresponding weight and summing the results.

[0063]

[0064] in, This represents the index coefficient corresponding to each risk factor; This indicates the actual score of the indicator corresponding to each risk factor; This represents the overall safety risk score for the entire production process;

[0065] S5-3: Assess the safety production risk level of metal hydrides according to the risk range of the calculated safety risk score; determine the risk level of the metal hydride safety production system based on the range of the safety risk score, thereby achieving effective risk assessment for industrial safety production systems.

[0066] Preferably, the risk level of the metal hydride safety production system is determined based on the range of the safety risk score:

[0067] The score range is 0-50, and the risk level is: dangerous, indicating a high safety risk.

[0068] The score range is 50-70; risk level: relatively dangerous, with safety risks present.

[0069] The score range is 70-80; risk level: safety risk exists.

[0070] The score range is 80-90; the risk level is relatively safe, with a small safety risk.

[0071] The score range is 90-100, and the risk level is safe.

[0072] Beneficial Effects: This invention first categorizes the risk factors in the safe production of metal hydrides, classifying them from multiple dimensions such as the natural environment, production process, and personnel management. Second, it constructs a fuzzy judgment matrix using fuzzy evaluation and clustering neutralization algorithms, and invites experts to conduct pairwise comparisons based on their experience and professional knowledge. This combination of qualitative and quantitative methods fully considers the complex relative importance of various factors in actual situations. Compared to existing technologies that rely solely on subjective experience, the analysis results of this invention are more scientific, objective, and accurate, better reflecting the true risk status of the metal hydride safe production system. Furthermore, it calculates the index weights using rigorous fuzzy hierarchical analysis and performs consistency checks, ensuring the rationality and reliability of the weighting results. This effectively avoids risk assessment bias caused by unreasonable weight settings, making the final risk assessment results more convincing. Attached Figure Description

[0073] Figure 1 A flowchart illustrating the overall process of the risk assessment technology for the metal hydride safe production system of the present invention is shown.

[0074] Figure 2 The flowchart for calculating the first-level index FAHP+ game theory utility function of the present invention is shown. Detailed Implementation

[0075] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0076] like Figure 1 A multi-level fuzzy assessment method for the safety risk of metal hydrides based on game equilibrium clustering neutralization includes the following steps:

[0077] S1: To address the safety risks associated with discrete and continuous mixing, multi-process, batch production, and the explosive nature of materials in the production of metal hydrides, an industrial production line safety risk assessment is conducted. Through the identification and analysis of major hazard factors, a three-tiered criterion system is constructed, comprising environmental factors, production process factors, and production personnel management factors, to achieve structured risk analysis.

[0078] S2: Based on the steps in S1, develop a three-tiered risk indicator system:

[0079] S2-1: The fault tree classification method, which integrates correlation analysis, subdivides specific risk indicators into three aspects: environmental risk, production process risk, and personnel risk, as primary indicators. Then, based on the specific risk points they cover, they are extended two levels down to form a total of three levels of indicators.

[0080] The hierarchical structure of risk indicators for the safe production system of metal hydrides is shown in Table 1:

[0081] Table 1: Hierarchical Structure of Risk Indicators for Safe Production Systems Using Metal Hydrides

[0082]

[0083] S3: Based on the risk indicator hierarchy constructed in S2, a team of experts in relevant fields is invited to evaluate the risk factors in each hierarchy. The experts, based on established judgment rules, experience, professional knowledge, and pre-defined scaling rules, construct a fuzzy judgment matrix.

[0084] S3-1: Expert consultation and scoring. A team of experts in relevant fields is organized. Team members must possess professional knowledge and rich practical experience in industrial safety production to ensure the comprehensiveness and professionalism of the evaluation. For risk factors at each level, evaluations are conducted based on established judgment rules and experience and professional knowledge. A clustering neutralization algorithm is initially used to optimize the evaluation results. Then, a fuzzy evaluation method using cross-correlation analysis is applied to judge the factors at each level. In the established 1-9 scale rules, 1 indicates that two factors are equally important, 9 indicates that one factor is extremely more important than the other, and intermediate values ​​represent different degrees of relative importance. Table 2 shows the numerical rules for the expert scoring scale of the first-level indicators for levels 1, 3, 5, 7, and 9 in this embodiment.

[0085] Table 2: Expert Scoring Scale Rules for Primary Indicators

[0086]

[0087] The goal of the K-Means clustering neutralization algorithm is to divide expert rating data into k cluster centers, and then calculate the mean of the k cluster centers to initially reduce the subjective differences. For each data point, the Euclidean distance to the k cluster centers is calculated. The formula is:

[0088]

[0089] in, This represents the score data of the i-th expert. Let m represent the cluster center, m represent the total number of dimensions of the risk indicators, and l represent the l-th risk indicator.

[0090] Will Assigned to the nearest cluster center The cluster it belongs to, namely:

[0091]

[0092] This represents finding the cluster index that minimizes the calculated Euclidean distance range; that is, dividing the expert data points into clusters according to the calculated distance range. j represents the index of the j-th cluster center, and k represents the total number of cluster centers.

[0093] S3-2: Construct a fuzzy judgment matrix. Each element in the fuzzy judgment matrix is ​​an evaluation of the corresponding factor using cross-correlation analysis. The matrix is ​​reciprocal, with all elements on the main diagonal being 1, to accurately reflect the relative importance of each factor. Assume that for an element in the criterion layer, the elements in the indicator layer below it are... Construct the fuzzy judgment matrix R:

[0094]

[0095] In the formula: As an indicator Compared to The importance of fuzzy membership degree, m=n.

[0096] S4: Combination Figure 2 Based on the constructed fuzzy judgment matrix, the weights of risk indicators are calculated and consistency checks are performed to ensure the reliability and rationality of the determined indicator weights, thereby guaranteeing the scientific nature and effectiveness of the entire risk assessment process. Simultaneously, addressing the multi-source risk factor correlation and coupling characteristics and error amplification effects existing in the primary indicator system, a non-cooperative game optimization framework for multi-stakeholder collaborative decision-making is constructed to reduce these characteristics and amplification effects.

[0097] S4-1: Using the fuzzy hierarchical analysis method, the weights of each indicator are calculated by performing the following operation steps on the fuzzy judgment matrix to obtain a quantitative weight value that reflects the relative importance of each indicator.

[0098] First, for each fuzzy element, it can be normalized:

[0099]

[0100]

[0101] in, Triples representing fuzzy membership degrees of importance; , , This represents the three elements of a triplet: the minimum value, the most likely value, and the maximum value. This represents the normalized importance fuzzy weight vector for each fuzzy element; S is the sum of the fuzzy weights of all factors, i.e., the sum of all medians.

[0102]

[0103] S4-2: The normalized fuzzy elements are calculated from S4-1. The centroid method is used to defuzzify the fuzzy numbers, and the defuzzified value of each fuzzy number is obtained. :

[0104]

[0105] S4-3: The judgment matrix is ​​composed of the defuzzified values ​​calculated in step S4-2. Then, based on the entropy-value fuzzy hierarchical analysis method, the standardized judgment matrix is ​​calculated and denoted as follows: Then, the weight vector, i.e., the weights, is obtained by calculating the average of each row of the standardized judgment matrix. The arithmetic mean of each row:

[0106]

[0107] In the formula: As an indicator Compared to Importance of fuzzy membership degree and defuzzification value; Further standardization of positive indicators:

[0108]

[0109] After standardization using positive indices, the resulting standardized judgment matrix is ​​denoted as: :

[0110]

[0111] In information theory, entropy refers to the degree of disorder or chaos in a system (entropy is a measure of uncertainty). The larger the entropy value of an indicator, the smaller its weight.

[0112] Based on standardized judgment matrix Further calculation of index weights:

[0113]

[0114] in, It is the weight of risk indicator i. This represents the judgment matrix obtained after normalization based on the entropy-based fuzzy hierarchical analysis method. The element in the i-th row and j-th column; n represents the judgment matrix. number of rows.

[0115] S4-4: To address the multi-source risk factor correlation and coupling characteristics and error amplification effect in the primary indicator system, a non-cooperative game optimization framework for collaborative decision-making among multiple stakeholders is constructed. Each relevant decision-maker is modeled as a game participant. By designing multi-objective optimization functions and constraints, the Nash equilibrium solution is calculated to determine the optimal weight combination for the primary indicators, thereby reducing the multi-source risk factor correlation and coupling characteristics and error amplification effect.

[0116]

[0117] in, This represents the game-theoretic utility function of expert j for indicator i. The attention coefficient of expert j to indicator i ( >0); It is the natural logarithm of the weight of indicator i; It is the penalty coefficient for expert j; It is the weight of risk indicator i; The initial weights of expert j on index i are calculated using the entropy-value fuzzy hierarchical analysis method.

[0118] Optimization of Nash equilibrium solution:

[0119]

[0120] Representing utility function Treatment of Optimization Weights The first-order partial derivatives were calculated. The value of is the optimized weight value obtained after the primary index is calculated using the entropy-value fuzzy hierarchical analysis method and the game theory utility function, and it must satisfy ∑ =1, otherwise it is not applicable.

[0121] The comparison of the technical effects, based on calculations, is shown in Table 3:

[0122] Table 3: Comparison of Technical Effects

[0123]

[0124] S4-5: Based on the calculated weights Then, we can find the largest eigenvalue of the judgment matrix. Let the judgment matrix be... The feature vector is :

[0125]

[0126] Find the largest eigenvalue :

[0127]

[0128] Where R represents the n-order fuzzy judgment matrix constructed by experts, and w is the indicator weight eigenvector, (Rw) i Let be the i-th component of the vector obtained by multiplying matrix R by the weight vector w, where n is the order of the judgment matrix and w is the weight vector. i Let be the weight of the i-th indicator.

[0129] S4-6: Perform a consistency check, which includes calculating the consistency index (CI) and the consistency ratio (CR). If CR is less than 0.1, the weighted result is accepted. If CR is not less than 0.1, the fuzzy judgment matrix is ​​readjusted by adjusting the element values ​​in the fuzzy judgment matrix provided by experts, and the above calculation and check process is repeated until the consistency requirements are met. This ensures the reliability and rationality of the determined weights of each indicator, thereby guaranteeing the scientific validity and effectiveness of the entire risk assessment process. Relevant indicator calculation formulas:

[0130]

[0131]

[0132] The RI values ​​are as follows: RI = 0 for order 1; 0 for order 2; 0.52 for order 3; 0.89 for order 4; 1.12 for order 5; 1.26 for order 6; 1.36 for order 7; 1.41 for order 8; 1.46 for order 9; 1.49 for order 10; 1.52 for order 11; 1.54 for order 12; and 1.56 for order 13.

[0133] S5: Based on the calculated reasonable risk indicator weights and expert team scores, conduct an overall risk assessment of the metal hydride safety production system:

[0134] S5-1: The weights of each level of indicators are obtained using the entropy-based fuzzy hierarchical analysis method. Based on the weights of each level of indicators, the weight of the first-level indicator to which indicator i belongs is: The weight of the secondary indicator to which indicator i belongs is The weight of the third-level indicator corresponding to indicator i is Calculate the coefficients of n risk indicators. :

[0135]

[0136] S5-2: Finally, record the actual scores for the corresponding indicators as follows: The safety risk score of the metal hydride safety production system is obtained by multiplying the quantitative result of each risk indicator by its corresponding weight and then summing the results.

[0137]

[0138] in, This represents the index coefficient corresponding to each risk factor; This indicates the actual score of the indicator corresponding to each risk factor; This represents the overall safety risk score for the entire production process;

[0139] S5-3: Based on the range of the safety risk score, determine the risk level of the metal hydride safety production system, thereby enabling effective risk assessment for industrial safety production systems.

[0140] Based on the range of the safety risk score, the risk level of the metal hydride safety production system is determined:

[0141] The score range is 0-50, and the risk level is: dangerous, indicating a high safety risk.

[0142] The score range is 50-70; risk level: relatively dangerous, with safety risks present.

[0143] The score range is 70-80; risk level: safety risk exists.

[0144] The score range is 80-90; the risk level is relatively safe, with a small safety risk.

[0145] The score range is 90-100, and the risk level is safe.

[0146] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the scope of protection of the present invention in any way, and all technical solutions obtained by equivalent substitution or other means fall within the scope of protection of the present invention. Parts not covered in this invention are the same as or can be implemented using existing technology.

Claims

1. A multi-level fuzzy assessment method for the safety risk of metal hydrides based on game equilibrium clustering neutralization, characterized in that... Includes the following steps: S1: Environmental risk, production process risk, and production personnel management risk are taken as primary indicators; S2: Based on step S1, for each primary indicator, specific risk indicators are subdivided as secondary indicators, and further subdivided to obtain tertiary indicators; these specific indicators constitute the hierarchical indicator layer. S3: Formulate evaluation rules and obtain expert scoring data. Optimize the expert scoring data based on the clustering neutralization algorithm to initially reduce errors caused by subjective human factors. Then, use the fuzzy evaluation method of cross-correlation analysis to evaluate the relative importance of factors at each level, and finally construct a fuzzy judgment matrix. S4: Defuzzify the fuzzy judgment matrix to obtain the judgment matrix. The judgment matrix is ​​then standardized using fuzzy hierarchical analysis. By calculating the weights of each indicator in the standardized judgment matrix, the maximum eigenvalue of the standardized judgment matrix is ​​obtained. Finally, a consistency check is performed to ensure the rationality of the standardized judgment matrix. If the consistency ratio is less than the threshold, the weight result is accepted; otherwise, the fuzzy judgment matrix is ​​readjusted. Finally, the weights of each indicator are determined. S5: Calculate the risk indicator coefficient based on the weights of each level of indicators corresponding to the i-th indicator. Actual scores for all indicators Combined with risk index coefficient Obtain the safety risk score of the metal hydride safety production system; determine the risk level of the metal hydride safety production system based on the range of the safety risk score.

2. The method according to claim 1, characterized in that, Step S1 specifically includes: In view of the safety risk characteristics of discrete and continuous mixing, multi-process, batch production and explosive materials in the production of metal hydrides, environmental risk, production process risk and production personnel management risk are used as primary indicators to construct a three-level criterion system of environmental factors, production process factors and production personnel management factors.

3. The method according to claim 1, characterized in that, Environmental factors include meteorological conditions and geological risks as secondary indicators; production process factors include production process risks, logistics and storage risks, equipment and facility risks, and supporting engineering service risks as secondary indicators; production personnel management factors include employee skills training factors, personnel operating procedures factors, and personnel health risks as secondary indicators.

4. The method according to claim 3, characterized in that, Meteorological factors include extreme temperatures as a level 3 indicator; geological risk factors include adverse geological conditions and vibration risk as level 3 indicators; production process risk factors include fire and explosion risk, toxic substance leakage risk, electric shock risk, and reaction indicator control deviation risk as level 3 indicators; logistics and storage risk factors include unsuitable land transportation and storage conditions as level 3 indicators; equipment and facility risk factors include equipment aging risk and maintenance quality risk as level 3 indicators. Risk factors for supporting engineering services include emission reduction project risks, unstable power supply, and ventilation system failures as tertiary indicators; employee skills training factors include lack of training plans, improper emergency handling, and short-sighted behavior as tertiary indicators; personnel operation standard factors include violations of operating procedures, improper use of protective equipment, and fatigue as tertiary indicators; and personnel health risks include inadequate regular physical examinations as tertiary indicators.

5. The method according to claim 1, characterized in that, Step S3 specifically includes: S3-1: Calculate expert scoring data To the cluster center European distance The formula is: in, This represents the score data of the i-th expert. Let m represent the cluster center, m represent the total number of dimensions of the risk indicators, and l represent the l-th risk indicator. Will Assigned to the nearest cluster center The cluster it belongs to, namely: This represents finding the cluster index that minimizes the distance based on the calculated Euclidean distance range, which means dividing the expert data points into clusters according to the calculated distance range; j represents the index of the j-th cluster center, and k represents the total number of cluster centers; S3-2: Construct a fuzzy judgment matrix. Each element in the fuzzy judgment matrix is ​​an evaluation of the corresponding factor using a fuzzy evaluation method based on cross-correlation analysis. The matrix is ​​reciprocal, with all elements on the main diagonal being 1, to accurately reflect the relative importance relationships between factors. For an element in the criterion layer, the elements in the index layer below it are... Construct the fuzzy judgment matrix R: In the formula: As an indicator Compared to The importance of fuzzy membership degree, m=n.

6. The method according to claim 1, characterized in that, Step S4 specifically includes: S4-1: First, normalize each fuzzy element: in, Triples representing fuzzy membership degrees of importance; , , This represents the three elements of a triplet: the minimum value, the most likely value, and the maximum value. This represents the normalized importance fuzzy weight vector for each fuzzy element; Calculate the sum of the fuzzy weights of all factors. : S4-2: Calculate the defuzzification value for each fuzzy number. : S4-3: From the deblurred value Construct a judgment matrix : In the formula: As an indicator Compared to The importance of fuzzy membership degree is defuzzified; then... After standardization using positive indices, the standardized judgment matrix is ​​denoted as follows: : In the formula: This represents the standardized judgment matrix obtained after normalization using the entropy-based fuzzy hierarchical analysis method. The element in the m-th row and n-th column; Finally, based on the standardized judgment matrix Further calculation of index weights: in, This represents the weight of index i. n represents the standardized judgment matrix. number of rows; S4-4: Design the game utility function for the primary indicator: in, This represents the game-theoretic utility function of expert j for indicator i. This refers to the attention coefficient of expert j towards indicator i. >0; It is the natural logarithm of the weight of indicator i; It is the penalty coefficient for expert j; It is the weight of indicator i; The initial weights of expert j on index i are obtained through FAHP calculation; Optimization of Nash equilibrium solution: Representing utility function Treatment of Optimization Weights The first-order partial derivatives were calculated. The value of is the optimized weight value obtained after the primary index is calculated using FAHP and the game theory utility function, and it must satisfy ∑ =1, otherwise not applicable; S4-5: Based on the standardized judgment matrix The weight of index i is calculated. Then, we can find the largest eigenvalue of the judgment matrix; let the eigenvector of the matrix be... : Find the largest eigenvalue : Where R represents the n-order fuzzy judgment matrix constructed by experts, and w is the indicator weight eigenvector, (Rw) i Let be the i-th component of the vector obtained by multiplying matrix R by the weight vector w, where n is the order of the judgment matrix and w is the weight vector. i Let be the weight of the i-th indicator; S4-6: Perform a consistency check, which includes calculating the consistency index CI and the consistency ratio CR. If CR is less than 0.1, the weight result is accepted. If CR is not less than 0.1, the fuzzy judgment matrix is ​​readjusted, the element values ​​in the fuzzy judgment matrix given by the experts are adjusted, and the consistency check is performed again until the consistency requirements are met. This ensures the reliability and rationality of the determined weights of each index, thereby guaranteeing the scientificity and effectiveness of the entire risk assessment process.

7. The method according to claim 1, characterized in that, Step S5 includes: S5-1: Based on the entropy-value fuzzy hierarchical analysis method, the weights of each level of indicators are obtained. The weight of the first-level indicator to which indicator i belongs is: The weight of the secondary indicator to which indicator i belongs is The weight of the third-level indicator corresponding to indicator i is ; Calculate index coefficients : S5-2: Safety Risk Score for Metal Hydride Safety Production Systems : ； S5-3: Based on the calculated safety risk score The risk level of metal hydride production is assessed based on the risk range it is located in.

Images