Coal mine production network vulnerability assessment method based on complex network theory

Abstract

The invention discloses a coal mine production network vulnerability assessment method based on a complex network theory. The method comprises the following steps: establishing a coal mine accident risk set and an accident network model; integrating an entropy weight Topsis and a complex network theory, and calculating to obtain an accident network centrality index weight and a comprehensive importance degree of each node; the co-occurrence rate of the two nodes is obtained through QCFS index evaluation, and the risk degree of the accident network connection edge is obtained through calculation in combination with the node comprehensive importance degree. And the complex network vulnerability evaluates the node comprehensive importance and the risk degree of the connection edge, and identifies the vulnerable link of the coal mine production system. The method can effectively identify key nodes and connecting edges of a coal mine production network, and effectively improves the safety level of a coal mine production system.

Description

Technical Field

[0001] The present invention relates to the technical field of vulnerability assessment of coal mine production systems, and in particular to a vulnerability assessment method for coal mine production networks based on complex network theory. Background Art

[0002] With the increasing mechanization of coal mine production and the improvement of production monitoring systems, the entire daily operation involves many system elements, and the coal mine system is increasingly showing the characteristics of a complex social and technical system. In addition, the high degree of uncertainty and complex environment increase the dynamics of the coal mine production site, making system monitoring more difficult. Coal mine accident analysis found that the safety problems of coal mine production mainly come from the coupling effect of various functions in the entire system, and the factors affecting coal mine safety need to be systematically thought about and analyzed.

[0003] Complex networks can show the internal connections between risk factors and have considerable advantages in visualizing risk nodes and interactions, which can make up for the lack of guidance of traditional methods in risk management. In addition, the system science analysis method provides a new idea for mining important links in complex networks, that is, to find key links in the network from the perspective of network connectivity. The removed nodes are more important because they damage the connectivity of the network.

[0004] In the prior art, the system structure and indicator changes are analyzed from the perspective of complex network theory or attack methods, thereby performing vulnerability analysis on the evaluation network.

[0005] For example, Gong Gangjun, Xie Hengji, etc. A method for assessing the vulnerable structure of a substation communication network based on complex network theory (CN117155799A) designed a method for assessing the vulnerable structure of a substation communication network based on complex network theory combined with link weights; Wang Xinglong, Wang Jin. A method for identifying key points of conflict in flight zone operation based on complex networks (CN117273527A) used two methods, improved entropy weight and grey correlation analysis, to identify the key points of conflict in the flight zone complex network model. However, in the analysis of network vulnerable structure, the evaluation methods of nodes and edges in coal mine production networks are often described separately, without linking them together, resulting in insufficient in-depth research on vulnerability assessment of coal mine production networks. Summary of the invention

[0006] In order to overcome the above technical problems, the purpose of the present invention is to provide a coal mine production network vulnerability assessment method based on complex network theory, which can effectively identify key nodes and edges of the coal mine production network and effectively improve the safety level of the coal mine production system.

[0007] In order to achieve the above object, the technical solution adopted by the present invention is:

[0008] A method for evaluating the vulnerability of coal mine production network based on complex network theory includes the following steps:

[0009] Step 1: Establish a coal mine accident risk set; use the accident risk as a network node, connect the two related points, and build a coal mine production network model;

[0010] Step 2: Integrate entropy weight Topsis and complex network theory to calculate the weight of the accident network centrality index and the comprehensive importance of each network node;

[0011] Step 3: Use the QCFS index to evaluate the co-occurrence rate of the two nodes, and calculate the risk of the accident network connection edge in combination with the comprehensive importance of the network nodes;

[0012] Step 4: Evaluate the vulnerable links of the coal mine production system based on the comprehensive importance of the network nodes and the risk of the connection edges.

[0013] The process of constructing the coal mine production network model in step 1 includes:

[0014] Taking accident reports as samples, the accident risk was counted from 0 to 1, the data was sorted, and the Apriori algorithm was added to obtain three indicator data: support, confidence and gain. The most reasonable association rules were obtained through trial and error. According to the obtained association rules, the antecedent and the consequent were used as network nodes, and the association rules were used as edges to construct a network model of coal mine production system.

[0015] Step 11: Based on the data obtained from the accident samples, the Apriori algorithm is added to obtain three indicator data: support, confidence and gain;

[0016] Step 12: Obtain the most reasonable association rules through trial and error;

[0017] Step 13: Construct a coal mine production system network model based on the obtained association rules.

[0018] In step 11, the process of arranging data and adding the Apriori algorithm includes:

[0019] Taking the accident report as a sample, we perform 0-1 statistics on the risk of each accident, organize the data, and then add the Apriori algorithm. The process is as follows:

[0020] 1) Set the threshold, set the minimum support and minimum confidence threshold;

[0021] 2) Generate candidate item set C1. First, scan the database, read all transactions, filter item sets, and the combination of item sets is the candidate item set;

[0022] 3) Generate frequent item sets L1, record the number of occurrences of each candidate item set, that is, calculate the support of each candidate item set. If it is greater than or equal to the minimum support threshold, the candidate item set is a frequent item set;

[0023] 4) Connect L1 to generate candidate item set C2. Connect the item sets of L1 to obtain candidate item set C2. C2 is equivalent to the combination of two item sets of L1.

[0024] 5) Generate frequent item set L2. If the support of C2 is greater than or equal to the minimum support threshold, the candidate item set is the frequent item set L2. Otherwise, prune the infrequent item set.

[0025] 6) Obtain candidate item sets Ck and frequent item sets Lk by recursive method; repeat steps 4) and 5), and the candidate item sets generated by each recursion have one more transaction than the previous one, and all frequent item sets are obtained by recursive method layer by layer;

[0026] 7) Obtain association rules and determine whether the confidence of all frequent item sets is greater than or equal to the minimum confidence threshold. If so, generate association rules; if not, prune the item sets.

[0027] The process of obtaining the most reasonable association rule by trial and error in step 12 includes:

[0028] By setting the threshold, setting different minimum support and minimum confidence, the number of association rules mined after calculation by the Apriori algorithm is different. Through continuous trial and error, the minimum support and minimum confidence are corrected until a suitable number of strong association rules appear.

[0029] In step 13, the process of constructing the coal mine production system network model is as follows:

[0030] All accident risks collected from accident reports are used as network nodes. The most reasonable association rules are obtained through trial and error as edges, and all nodes are linked to build a coal mine production system network model.

[0031] The process of calculating the centrality index weight and the comprehensive importance of each node in step 2 includes:

[0032] Step 21: constructing a risk factor decision matrix based on the coal mine production system network model;

[0033] Step 22: Use the entropy weight method to calculate the weights of the four indicators: degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality;

[0034] Step 23: Substitute the weights of each centrality index into the Topsis method to calculate the comprehensive importance of each node.

[0035] In step 21, the process of constructing the risk factor decision matrix includes:

[0036] Calculate the centrality index of each node. First, calculate the degree centrality that reflects the power status of the node:

[0037]

[0038] Next, we calculate the closeness centrality, which represents the average distance between a node and other nodes in the network:

[0039]

[0040] Secondly, the betweenness centrality is calculated, which represents the number of shortest paths passing through a node in the shortest paths connecting all nodes in the network:

[0041]

[0042] Finally, the eigenvector centrality that measures the importance of a node while taking into account the importance of neighboring nodes is calculated:

[0043]

[0044] K i is the degree of node i, n is the total number of nodes; D i is the degree centrality of node i; and are the inward node centrality and outward node centrality of node i, respectively, and d ji is the shortest path distance from node j to node i, d ij is the shortest path distance from node i to node j, n' is the number of nodes to reach i, and n" is the number of nodes that i can reach; σ(s,t|i) is the number of shortest paths from node s to node t, and σ(s,t) is the number of paths passing through node i; λ is the neighbor matrix A ij The maximum eigenvalue of .

[0045] The process of calculating the centrality index weight using the entropy weight method in step 22 includes:

[0046] Step 221: Create a decision matrix, assuming that the TOPSIS decision matrix consists of n nodes and m indicators, and the m indicators are degree centrality, closeness centrality, betweenness centrality and eigenvector centrality;

[0047]

[0048] Step 222: Calculate the ratio;

[0049]

[0050] Step 223: e jis the entropy value of j, the entropy value;

[0051]

[0052] Step 224: w j is the weight of j, fixed weight;

[0053]

[0054] The process of calculating the comprehensive importance of each node in step 23 includes:

[0055] Step 231: Standardize the decision matrix r ij ;

[0056]

[0057] Step 232: Obtain weighted decision matrix a ij ;

[0058] a ij =w j × ij

[0059] Step 233: Positive ideal solution Negative ideal solution

[0060]

[0061] Step 234: Use Euclidean distance to calculate the distance between the evaluation object and the positive and negative ideal solutions respectively

[0062]

[0063] Step 235: Calculate relative proximity C i ; C i Higher nodes have higher influence;

[0064]

[0065] The process of calculating the risk of the accident network connection edge in step 3 includes:

[0066] Step 31: Use the QCFS index to evaluate the co-occurrence rate of the two nodes;

[0067] Step 32: Calculate the risk of the accident network connection edge in combination with the comprehensive importance of the nodes.

[0068] The process of calculating the co-occurrence rate of two nodes in step 31 includes:

[0069] The QCFS index is used to express the co-occurrence rate of risks i and j, where P ij The value of is between 0 and 1. The probability of the risk node connection edge is the probability that the failure of node i causes the failure of node j, or the possibility that node i and node j fail together, that is, the co-occurrence rate of the two nodes.

[0070]

[0071] In the formula, Indicated by V i the number of incidents that were contributing factors; It refers to the number of accidents with these two factors.

[0072] The process of calculating the risk of the accident network connection edge in step 32 includes:

[0073] Combined with the comprehensive importance of the node C i and the co-occurrence rate P of nodes i and j ij To calculate the risk of the connecting edge

[0074] Q ij =C i ×P ij .

[0075] The process of evaluating the vulnerable links of the coal mine production system in step 4 includes:

[0076] The comprehensive importance of the network nodes of the coal mine is C i and the risk of the connecting edge Q ij Arrange in descending order. The larger the link vulnerability value, the greater the comprehensive importance of the network node C. i and the risk of the connecting edge Q ij The more vulnerable the coal mine accident network is, the overall network efficiency is used as the network vulnerability measurement indicator, and the nodes and edges of the network are deliberately and randomly attacked, and the network efficiency reduction rate under different importance measurement schemes is compared;

[0077]

[0078] η o is the efficiency value of the original network, η represents the network efficiency value after some nodes are removed, and the faster the μ value decreases, the greater the impact of node removal on network efficiency.

[0079] Beneficial effects of the present invention:

[0080] The present invention can effectively improve the accuracy of vulnerability assessment of coal mine production systems. The entropy weight Topsis and complex network theory are integrated to calculate the weight of the accident network centrality index and the comprehensive importance of each node; the QCFS index is used to evaluate the co-occurrence rate of two nodes, and the risk of the accident network connection edge is calculated in combination with the comprehensive importance of the nodes. When analyzing the network vulnerability structure, the evaluation method of describing the nodes and edges in the coal mine production network is closely linked to make the evaluation result more scientific, reasonable and accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 The figure is a flow chart of the vulnerability method of the coal mine production system of the present invention.

[0082] Figure 2 It is a schematic diagram of the network efficiency reduction rate due to node failure of the present invention.

[0083] Figure 3 This is a schematic diagram of the efficiency drop rate of the edge failure network of the present invention. DETAILED DESCRIPTION

[0084] The present invention will be further described in detail below in conjunction with the accompanying drawings.

[0085] In order to make the technical solutions and advantages of the embodiments of the present invention more clearly understood, the exemplary embodiments of the present invention are further described in detail below with reference to the accompanying drawings. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0086] Example 1: Reference Figure 1-Figure 3 To illustrate this implementation, a coal mine production network vulnerability assessment method based on complex network theory includes the following steps:

[0087] Step 1: Establish coal mine accident risk set and coal mine production network:

[0088] The risk in the present invention refers to the severity of the consequences of coal mine accidents and other risks. In this embodiment, the important influencing factors of coal mine accidents are linked to the coal mine production system process, and 48 accident risks are identified in a 6-level structure. The coal mine accident risk set can be expressed as:

[0089] U={A,B,C,D,e,F}

[0090] A={A1,A2} .......

[0092] F={F1,F2,F3,F4}

[0093] U represents the risk set of the coal mine production system; A, B.....F represent the risk sets at different levels of the coal mine production system; A1, A2....F3,F4 represent the sub-level risk set of the coal mine production system.

[0094] Using accident reports as samples, we performed 0-1 statistics on accident risks and sorted the data. We added the Apriori algorithm to obtain three indicator data: support, confidence, and gain. We obtained the most reasonable association rules through trial and error.

[0095] Table 1 Top 10 association rules by confidence

[0096]

[0097] From Table 1, we can see the association rules ranked by confidence. According to the obtained association rules, the accident risk is taken as the network node and the association rule is taken as the edge to construct the network model of the coal mine production system.

[0098] Step 2: Integrate entropy weight Topsis and complex network theory to calculate the accident network centrality index weight and the comprehensive importance of each node:

[0099] The TOPSIS (Similarity to Ideal Solution) method is one of the most effective methods for multi-attribute decision-making problems that combines multiple evaluation indicators to obtain the optimal solution. The TOPSIS method takes each importance evaluation indicator as a node attribute, processes the multi-attribute problem through the TOPSIS method, and forms a comprehensive importance evaluation indicator. Since TOPSIS needs to multiply the weight corresponding to each indicator separately, the entropy weight method is used to obtain the weight of the indicator, and a weight is added to TOPSIS. By sorting the distance between the evaluation object and the optimal solution and the worst solution, the one close to the optimal solution is the best, and the opposite is the worst.

[0100] Step 21, constructing a risk factor decision matrix based on the coal mine production system network model;

[0101] Four indicators, namely degree centrality reflecting the local characteristics of complex networks and betweenness centrality, closeness centrality and eigenvector centrality reflecting the global characteristics of complex networks, are selected to quantify the node characteristics of complex networks.

[0102] Calculate the centrality index of each node. First, calculate the degree centrality that reflects the power status of the node:

[0103]

[0104] Next, we calculate the closeness centrality, which represents the average distance between a node and other nodes in the network:

[0105]

[0106] Secondly, the betweenness centrality is calculated, which represents the number of shortest paths passing through a node in the shortest paths connecting all nodes in the network:

[0107]

[0108] Finally, the eigenvector centrality that measures the importance of a node while taking into account the importance of neighboring nodes is calculated:

[0109]

[0110] K i is the degree of node i, n is the total number of nodes; D i is the degree centrality of node i; and are the inward node centrality and outward node centrality of node i, respectively, and d ji is the shortest path distance from node j to node i, d ij is the shortest path distance from node i to node j, n' is the number of nodes to reach i, and n" is the number of nodes that i can reach; σ(s,t|i) is the number of shortest paths from node s to node t, and σ(s,t) is the number of paths passing through node i; λ is the neighbor matrix A ij The maximum eigenvalue of .

[0111] Create a decision matrix. The TOPSIS decision matrix consists of 48 nodes and 4 indicators, which are degree centrality, closeness centrality, betweenness centrality, and eigenvector centrality.

[0112]

[0113] The rows represent the four indicators of degree centrality, closeness centrality, betweenness centrality and eigenvector centrality; the columns represent the specific values ​​of the centrality of the 48 nodes.

[0114] Step 22: Use the entropy weight method to calculate the weights of the four indicators: degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality;

[0115] Furthermore, step 22 specifically includes:

[0116] Step 221: Create a decision matrix.

[0117] Step 222, calculate the ratio.

[0118]

[0119] Step 223: Calculate the entropy value.

[0120]

[0121] Step 224: Determine the weight.

[0122]

[0123] Thus, the weights of the four indicators, namely, degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality, are obtained. The weights of the four indicators can be seen from Table 2.

[0124] Table 2 Centrality weights

[0125]

[0126] Step 23: Substitute the weights of each centrality index into the Topsis method to calculate the comprehensive importance of each node.

[0127] Step 231: Standardize the decision matrix.

[0128]

[0129] Step 232: Obtain a weighted decision matrix.

[0130] a ij =w j × ij

[0131] Step 233: Positive and negative ideal solutions.

[0132]

[0133] Step 234: Use the Euclidean distance to calculate the distance between the evaluation object and the positive and negative ideal solutions respectively.

[0134]

[0135] Step 235: Calculate relative proximity. i Higher nodes have higher influence.

[0136]

[0137] The relative closeness C i This represents the influence of risk nodes in the coal mine production system. Arranging the influence of all nodes from large to small can determine the vulnerability of coal mine network nodes.

[0138] Step 3: Use the QCFS index to evaluate the co-occurrence rate of the two nodes, and calculate the risk of the accident network connection edge in combination with the comprehensive importance of the nodes.

[0139] Furthermore, step three specifically includes:

[0140] Step 31, using the QCFS index to evaluate and obtain the co-occurrence rate of the two nodes;

[0141] QCFS index (quantitative comparison between factors based on the frequency statistics). In this study, the probability of occurrence of risk node connection edge refers to the probability that the failure of node i causes the failure of node j, which can also be considered as the possibility of common failure of node i and node j, that is, the co-occurrence rate of the two nodes. The present invention introduces the QCFS method to determine the probability of connection edge occurrence, so as to better dynamically evolve the complex network of coal mine accidents and make the risk assessment more convincing.

[0142] This embodiment is based on the investigation report on 105 lifting and major accidents in underground coal mines in my country from 2010 to 2023, and counts the frequency of node failure and the frequency of co-occurrence of risks. The calculation formula is as follows:

[0143]

[0144] In the formula, Indicated by V i the number of incidents that were contributing factors; It refers to the number of accidents with these two factors. The probability distribution is shown in Table 3, which shows the risk transfer probability of the connecting edges of the accident evolution network.

[0145] Table 3. Probability statistics of accident evolution network connection edges

[0146]

[0147] Step 32: Calculate the risk of the accident network connection edge in combination with the comprehensive importance of the node;

[0148] Combined with the comprehensive importance of the node C i and the co-occurrence rate P of nodes i and j ij To calculate the risk of the connecting edge.

[0149] Q ij =C i ×P ij

[0150] Step 4: Evaluate the vulnerable links of the coal mine production system based on the comprehensive importance of the nodes and the risk of the connecting edges:

[0151] The comprehensive importance C of the coal mine network node i and the risk Q of the connecting edge ij Arrange in descending order. The larger the link vulnerability value is, the greater the comprehensive importance of the node C is. i and the risk of the connecting edge Q ijThe more vulnerable the network is in the coal mine accident network. Using the overall network efficiency as the network vulnerability measurement indicator, the nodes and edges of the network are deliberately and randomly attacked. Comparing the network efficiency reduction rate under different importance measurement schemes, it provides decision support for improving the safety level of coal mine production. For details, see Figure 2 , Figure 3 .

[0152]

[0153] η o is the efficiency value of the original network, η represents the network efficiency value after some nodes are removed, and the faster the μ value decreases, the greater the impact of node removal on network efficiency.

[0154] The network vulnerability changes differently under different attack methods. In order to select the scheme with the greatest impact of node sorting on network vulnerability, the node importance measured under each strategy is sorted from large to small. A total of 6 strategies, including deliberate attacks and random attacks based on entropy weight Topsis comprehensive method, PageRank, betweenness centrality, K-core, CI, are used to attack network nodes one by one, as shown in the attached figure. Figure 2 Among the different strategies for removing network nodes, deliberate attack is significantly better than random attack, and its network efficiency decrease rate is higher, indicating that targeted prevention strategies are beneficial to improving the safety level of coal mine production; in the first half of the network attack, the network efficiency of the attack strategy based on the entropy weight Topsis comprehensive method decreases faster, and is the best throughout the whole process compared with other attacks, indicating that the deliberate attack based on the Topsis comprehensive method is more suitable for the complex network of coal mine production system.

[0155] Five strategies, including deliberate attack based on QCFS, gain, support, edge weight, and random attack, were used to attack the network connection edges one by one. By comparison, different attack strategies have different degrees of damage to the network. The results are shown in the attached figure. Figure 3 When the number of node attacks does not reach 100, the attack strategy based on QCFS is significantly better than other attack strategies; when the number of node attacks reaches 100, there is an overlap between the lines of the two attack strategies, QCFS and support, indicating that the two edge importance ranking schemes are roughly the same, and the deliberate attack strategy based on QCFS is more suitable for the complex network of coal mine production system.

[0156] In summary, the entropy weight Topsis comprehensive method and QCFS analysis can effectively identify the key nodes and edges of the complex network of coal mine production system. By preventing and controlling key nodes and edges, the safety level of coal mine production system can be effectively improved to ensure the safety and efficiency of coal mine production.

[0157] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A coal mine production network vulnerability assessment method based on complex network theory, characterized in that: The steps include: Step 1: Establish a coal mine accident risk set; use the accident risk as a network node, connect the two related points, and build a coal mine production network model; Step 2: Integrate entropy weight Topsis and complex network theory to calculate the weight of the accident network centrality index and the comprehensive importance of each network node; Step 3: Use the QCFS index to evaluate the co-occurrence rate of the two nodes, and calculate the risk of the accident network connection edge in combination with the comprehensive importance of the network nodes; Step 4: Evaluate the vulnerable links of the coal mine production system based on the comprehensive importance of the network nodes and the risk of the connection edges.

2. A coal mine production network vulnerability assessment method based on complex network theory according to claim 1, characterized in that: The process of constructing the coal mine production network model in step 1 includes: Taking accident reports as samples, the accident risk is counted from 0 to 1, the data is sorted, and the Apriori algorithm is added to obtain three indicator data: support, confidence and gain. The most reasonable association rules are obtained through trial and error. According to the obtained association rules, the antecedent and the consequent are used as network nodes, and the association rules are used as edges to construct a network model of coal mine production system. Step 11: Based on the data obtained from the accident samples, the Apriori algorithm is added to obtain three indicator data: support, confidence and gain; Step 12: Obtain the most reasonable association rules through trial and error; Step 13: Based on the obtained association rules, a coal mine production system network model is constructed.

3. A coal mine production network vulnerability assessment method based on complex network theory according to claim 2, characterized in that: In step 11, the process of arranging data and adding the Apriori algorithm includes: Taking the accident report as a sample, we perform 0-1 statistics on the risk of each accident, organize the data, and then add the Apriori algorithm. The process is as follows: 1) Set the threshold, set the minimum support and minimum confidence threshold; 2) Generate candidate item set C1. First, scan the database, read all transactions, filter item sets, and the combination of item sets is the candidate item set; 3) Generate frequent item sets L1, record the number of occurrences of each candidate item set, that is, calculate the support of each candidate item set. If it is greater than or equal to the minimum support threshold, the candidate item set is a frequent item set; 4) Connect L1 to generate candidate item set C2. Connect the item sets of L1 to obtain candidate item set C2. C2 is equivalent to the combination of two item sets of L1. 5) Generate frequent item set L2. If the support of C2 is greater than or equal to the minimum support threshold, the candidate item set is the frequent item set L2. Otherwise, prune the infrequent item set. 6) Obtain candidate item sets Ck and frequent item sets Lk by recursive method; repeat steps 4) and 5), and the candidate item sets generated by each recursion have one more transaction than the previous one, and all frequent item sets are obtained by recursive method layer by layer; 7) Obtain association rules and determine whether the confidence of all frequent item sets is greater than or equal to the minimum confidence threshold. If so, generate association rules; if not, prune the item sets.

4. A method for evaluating the vulnerability of coal mine production network based on complex network theory according to claim 2, characterized in that: The process of obtaining the most reasonable association rule by trial and error in step 12 includes: By setting the threshold, setting different minimum support and minimum confidence, the number of association rules mined after calculation by the Apriori algorithm is different. Through continuous trial and error, the minimum support and minimum confidence are corrected until a suitable number of strong association rules appear. In step 13, the process of constructing the coal mine production system network model is as follows: All accident risks collected from accident reports are used as network nodes. The most reasonable association rules are obtained through trial and error as edges, and all nodes are linked to build a coal mine production system network model.

5. A method for evaluating the vulnerability of coal mine production network based on complex network theory according to claim 4, characterized in that: The process of calculating the centrality index weight and the comprehensive importance of each node in step 2 includes: Step 21: constructing a risk factor decision matrix based on the coal mine production system network model; Step 22: Use the entropy weight method to calculate the weights of the four indicators: degree centrality, betweenness centrality, closeness centrality, and eigenvector centrality; Step 23: Substitute the weights of each centrality indicator into the Topsis method to calculate the comprehensive importance of each node.

6. A method for evaluating the vulnerability of coal mine production network based on complex network theory according to claim 5, characterized in that: In step 21, the process of constructing the risk factor decision matrix includes: Calculate the centrality index of each node. First, calculate the degree centrality that reflects the power status of the node: Next, we calculate the closeness centrality, which represents the average distance between a node and other nodes in the network: Secondly, the betweenness centrality is calculated, which represents the number of shortest paths passing through a node in the shortest paths connecting all nodes in the network: Finally, the eigenvector centrality that measures the importance of a node while taking into account the importance of neighboring nodes is calculated: K i is the degree of node i, n is the total number of nodes; D i is the degree centrality of node i; and are the inward node centrality and outward node centrality of node i, respectively, and d ji is the shortest path distance from node j to node i, d ij is the shortest path distance from node i to node j, n' is the number of nodes to reach i, and n" is the number of nodes that i can reach; σ(s,t|i) is the number of shortest paths from node s to node t, and σ(s,t) is the number of paths passing through node i; λ is the neighbor matrix A ij The maximum eigenvalue of .

7. A method for evaluating the vulnerability of coal mine production network based on complex network theory according to claim 5, characterized in that: The process of calculating the centrality index weight using the entropy weight method in step 22 includes: Step 221: Create a decision matrix, assuming that the TOPSIS decision matrix consists of n nodes and m indicators, and the m indicators are degree centrality, closeness centrality, betweenness centrality and eigenvector centrality; Step 222: Calculate the ratio; Step 223: e j is the entropy value of j, the entropy value; Step 224: w j is the weight of j, fixed weight; 8. The method for evaluating the vulnerability of coal mine production network based on complex network theory according to claim 5 is characterized in that: The process of calculating the comprehensive importance of each node in step 23 includes: Step 231: Standardize the decision matrix r ij ; Step 232: Obtain weighted decision matrix a ij ; a ij =w j xr ij Step 233: Positive ideal solution Negative ideal solution Step 234: Use Euclidean distance to calculate the distance between the evaluation object and the positive and negative ideal solutions respectively Step 235: Calculate relative proximity C i ; C i Higher nodes have higher influence; 9. The method for evaluating the vulnerability of coal mine production network based on complex network theory according to claim 1 is characterized in that: The process of calculating the risk of the accident network connection edge in step 3 includes: Step 31: Use the QCFS index to evaluate the co-occurrence rate of the two nodes; Step 32: Calculate the risk of the accident network connection edge in combination with the comprehensive importance of the node; The process of calculating the co-occurrence rate of two nodes in step 31 includes: The QCFS index is used to express the co-occurrence rate of risks i and j, where P ij The value of is between 0 and 1. The probability of the risk node connection edge is the probability that the failure of node i causes the failure of node j, or the possibility that node i and node j fail together, that is, the co-occurrence rate of the two nodes. In the formula, Indicated by V i the number of incidents that were contributing factors; refers to the number of accidents with these two factors; The process of calculating the risk of the accident network connection edge in step 32 includes: Combined with the comprehensive importance of the node C i and the co-occurrence rate P of nodes i and j ij To calculate the risk of the connecting edge Q ij =C i ×P ij 。 10. A coal mine production network vulnerability assessment method based on complex network theory according to claim 1, characterized in that: The process of evaluating the vulnerable links of the coal mine production system in step 4 includes: The comprehensive importance of the network nodes of the coal mine is C i and the risk Q of the connecting edge ij Arrange in descending order. The larger the link vulnerability value, the greater the comprehensive importance of the network node C. i and the risk Q of the connecting edge ij The more vulnerable the coal mine accident network is, the overall network efficiency is used as the network vulnerability measurement indicator, and the nodes and edges of the network are deliberately and randomly attacked, and the network efficiency reduction rate under different importance measurement schemes is compared; η o is the efficiency value of the original network, η represents the network efficiency value after some nodes are removed, and the faster the μ value decreases, the greater the impact of node removal on network efficiency.

Images