Decision-making method and system for large-scale satellite emergency plan in social network environment

Abstract

The present invention provides a large-scale satellite emergency plan decision-making method, system, storage medium and electronic device in a social network environment, and relates to the field of satellite emergency plan decision-making. In the present invention, all decision makers are clustered into several clusters based on a decision preference matrix and a social network relationship diagram, and each initial cluster decision preference matrix and an initial group decision preference matrix are obtained; the cluster consensus level and the group consensus level are updated based on a preset embedded feedback adjustment mechanism to obtain a group decision preference matrix based on consensus. On the one hand, the trust problem between decision makers is taken into account in the clustering process (i.e., the trust between nodes is considered), avoiding the default of the inter-node weighting factor to 1 for a directed graph, which plays a great role in the accuracy of the large satellite emergency plan clustering results; on the other hand, an embedded feedback adjustment mechanism is designed to serve the management of non-cooperative behavior, so that the opinions of individual decision makers and clusters are both closer to the group, thereby improving the decision-making level.

Description

Technical Field

[0001] The present invention relates to the technical field of satellite emergency plan decision-making, and in particular to a large-scale satellite emergency plan decision-making method, system, storage medium and electronic equipment in a social network environment. Background Art

[0002] In recent years, emergency events such as earthquakes, floods, fires, and local wars have occurred frequently around the world, often with catastrophic consequences. With the continuous advancement of satellite technology and increasing user demand for data, imaging satellites have become an important means of obtaining ground-based information during emergencies. During actual observations, highly dynamic, time-sensitive emergency tasks arise, often accompanied by uncertainty in their timing and number. Satellite emergency mission planning is the process of rapidly generating multiple emergency plans based on user-generated emergency observation requirements. The optimal observation plan is then rapidly selected based on emergency decision-making to maximize the effectiveness of the emergency observation. Due to the complexity and uncertainty of emergency decision-making, multiple decision makers (DMs) are often required, leading to the large group decision-making problem (LGDM).

[0003] At present, the LGDM problem has attracted more and more attention in the field of decision science. On the other hand, since satellite emergency plan decision-making involves a wide range of aspects, the LGDM applicable to satellite emergency plan decision-making usually involves a large number of decision makers (DMs) from different professional backgrounds to participate in the decision-making process. Such large-group satellite emergency plan decision-making (LGSESDM) has complex large-group characteristics: (1) the group size is relatively large, generally requiring multiple DMs from different positions to participate in the decision-making process, (2) it must be guaranteed that the decision is made in a short time, (3) it is often difficult to reach a consensus among DMs, and (4) a wrong decision or too slow may lead to catastrophic losses of the emergency mission. As the LGDM problem has attracted more and more attention from scholars, some new decision-making models and methods have been proposed. These models and methods all contain three main steps: large group clustering, consensus reaching process (CRP) and plan selection process. Traditional consensus is called "hard consensus". The scalar of consensus measurement has only two indicators: 0 (no consensus reached) or 1 (consensus achieved). In actual decision-making, it is very difficult and impractical to adjust the consensus from 0 to 1.

[0004] Most of the above clustering methods use the consistency of decision makers' decision preferences as the clustering criterion, ignoring the intrinsic connections between decision makers, which in turn affects the rationality of the final selection of large-scale satellite emergency plans.

[0005] Furthermore, traditional consensus models mostly focus on traditional decision-making problems involving a small number of DMs (e.g., 3-5) without time constraints. This may not be applicable to large-group decision-making in emergency situations. Real-world large-group decision-making problems require more than just mathematical models; the psychological behavior and actual performance of decision-makers are also worthy of attention and consideration. This is because non-cooperative behavior is common in real-world decision-making processes. Considering potential non-cooperative behavior during the decision-making process helps ensure decision quality. In LGSESDM problems, decision-makers often come from different satellite scheduling positions and have different priorities. They are more likely to engage in non-cooperative behavior to promote their own personal interests, making it impossible to reach consensus. When some DMs refuse to make adjustments, the CRP process can stall. Furthermore, LGSESDM must achieve relatively satisfactory results within a short period of time. Therefore, managing non-cooperative behavior is crucial to ensuring the timeliness and quality of the CRP process. Summary of the Invention

[0006] (1) Technical problems solved

[0007] In response to the shortcomings of the existing technology, the present invention provides a large-scale satellite emergency plan decision-making method, system, storage medium and electronic device in a social network environment, which solves the technical problems that the existing clustering method uses the consistency of decision-makers' decision preferences as the clustering criterion, ignoring the intrinsic connections between decision-makers; and the decision-making process ignores non-cooperative behavior, resulting in low accuracy in the consensus-building process.

[0008] (2) Technical solution

[0009] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0010] A large-scale satellite emergency plan decision-making method in a social network environment, comprising:

[0011] S1. Obtaining a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix uses a hesitant fuzzy binary semantic set as a form of decision opinion expression;

[0012] S2. Clustering all decision makers into a number of clusters based on the decision preference matrix and the social network relationship diagram, and obtaining each initial cluster decision preference matrix and an initial group decision preference matrix;

[0013] S3. Based on each of the initial cluster decision preference matrices and the initial group decision preference matrix, the cluster consensus level and the group consensus level are updated based on a preset embedded feedback adjustment mechanism to obtain a group decision preference matrix based on consensus.

[0014] S4. Obtaining a finally selected large-group satellite emergency plan based on the group decision preference matrix.

[0015] Preferably, the Louvain clustering algorithm based on modularity is used for clustering in S2, which specifically includes:

[0016] S21, treating each node of the social network relationship graph as a cluster, and then merging neighbor nodes of the cluster into the same cluster to obtain multiple clusters;

[0017] S22. Obtaining weighting factors between different decision makers in the social network relationship graph based on the trust matrix of the social network relationship graph and the similarity between any two decision preference matrices; obtaining the modularity of each cluster and the corresponding modularity change based on the weighting factors between the different decision makers;

[0018] S23, repeating the above steps S21 to S22 until the overall modularity no longer changes, and finally obtaining several clusters;

[0019] S24. Obtaining the cluster weight and initial cluster decision preference matrix of each cluster according to the final clustering result; and obtaining the initial group decision preference matrix according to the cluster weight and initial cluster decision preference matrix of each cluster.

[0020] Preferably, the S22 specifically includes:

[0021] S221. Use a tuple of type λ=(t,d) to represent the trust function value, where t,d∈[0,1], the first component t is a trust degree, and the second component d is a distrust degree; define the trust matrix as Represents decision makers The trust between Represents decision makers The trust score between any two decision makers is expressed as

[0022] S222: Use the following formula to obtain the weighting factors between different decision makers in the social network relationship diagram:

[0023]

[0024] in, represents the similarity between any two decision preference matrices;

[0025] are the decision preference matrices of any two decision makers, i.e., the decision information of m large group satellite emergency plans on n attribute indicators;

[0026] in, They are the granularity of the corresponding hesitant fuzzy binary semantic set;

[0027]

[0028] express and Hausdorff distance between hesitant fuzzy binary semantic sets;

[0029] and It consists of several hesitant fuzzy binary semantic sets, the number of which is determined by the scheduling opinion; It is a language tag, and its value is S={s0,s1,…,s g}; represents the sign conversion value and has a range of [-0.5, 0.5), g is the potential of S;

[0030] S223, obtaining the modularity Q0 of each cluster according to the weighting factors between the different decision makers;

[0031]

[0032] Wherein, M represents the sum of weights of all edges of the social network relationship graph; Represents decision makers The weighting factor of the edge between them. When the network is not a weighted graph, the weight of the edge is 1; Represents the sum of weighted factors of all edges connected to node l1 or l2 respectively; Respectively The cluster to which it belongs, if belong to the same cluster C, then otherwise

[0033] S224, and obtaining the corresponding modularity change ΔQ0;

[0034]

[0035] Among them, ∑in represents the sum of the weighted factors of all edges in cluster C, ∑tot represents the sum of the weighted factors of the edges connected to the points in cluster C, Represents the sum of the weighted factors of the edges connecting node l1 with the points in cluster C.

[0036] Preferably, the S24 specifically includes:

[0037] based on Calculate the initial decision preference matrix of each cluster; where it is assumed that the decision maker Clustered into K clusters, the kth cluster is recorded as C k ,N k Indicates the number of members in the cluster;

[0038] based on Calculate cluster weights, where represents the weight of the decision maker, and

[0039] based on Calculate the initial group decision preference matrix.

[0040] Preferably, S3 specifically includes:

[0041] S31. Let t = 0, given the group consensus threshold

[0042] S32. Set t=t+1 and calculate the current cluster consensus level

[0043]

[0044] Among them, D e (G k(t) ,G f(t) ) is G k(t) and G f(t) The Hausfdorff distances of hesitant fuzzy binary semantic sets corresponding to the clusters;

[0045] Calculate the current group consensus level GCL t :

[0046]

[0047] judge and Then go to S35; otherwise, go to S33;

[0048] S33. Based on the current cluster consensus level and the previous cluster consensus level Get the cooperation coefficient used to measure the degree of change in the consensus level of a single cluster

[0049]

[0050] in, Assign values ​​according to actual conditions;

[0051] Determine the cooperation coefficient threshold

[0052]

[0053] Among them, ρ t Is the penalty coefficient, used to determine the cooperation coefficient How much smaller will it fall into the penalty area, satisfying 0≤ρ t ≤1;

[0054] S34, Comprehensive and as well as GCL t and The size comparison result is obtained, and the embedded feedback adjustment mechanism is introduced to adjust the cluster weight of the non-cooperative cluster and / or the decision preference matrix of its internal decision makers, update the current cluster decision preference matrix, and go to S32;

[0055] S35, let t * =t, output the final cluster decision preference matrix Get the final group decision matrix R c , and serves as the group decision preference matrix based on the consensus reached.

[0056] Preferably, the embedded feedback regulation mechanism in S3 includes:

[0057] (1) Opinion modification includes two situations:

[0058] The first one, if The clusters that indicate the need for modification cooperate with the modification decision suggestion and only use the subjective coefficient adjustment;

[0059]

[0060] Among them, G k (t) represents the cluster C that needs to be modified after t iterations k The decision preference matrix, is the subjective correction factor, determined by the cluster;

[0061] The second type, if Indicates that the cluster that needs to be modified does not cooperate with the modification of the decision suggestion, that is, there is an uncooperative cluster. The objective coefficient and subjective coefficient are used to modify the opinion, including:

[0062] First, by solving the minimum similarity between the opinions of each decision maker in the cluster that needs to be modified and all other decision makers, the position of the decision preference matrix with the largest opinion difference in the cluster is determined.

[0063]

[0064] in, Representing any two decision makers The similarity between the corresponding decision preference matrices;

[0065] Secondly, objectively adjust the decision preference matrix determined above.

[0066]

[0067] in, is the objective adjustment coefficient, and get;

[0068] Through objective modification, the cluster at this time

[0069] Finally, if subjective adjustments are still required,

[0070]

[0071] (2) Weight penalty;

[0072]

[0073] in, Represents cluster C k The cluster consensus level in the tth iteration, GCL t Represents cluster C k The group consensus level in the t-th iteration of Represents cluster C k The cluster weights at iteration t.

[0074] Preferably, in S34, adjusting the cluster weight of the non-cooperative cluster and / or the decision preference matrix of its internal decision makers specifically includes:

[0075] Area 1 is the first cooperation area: if No adjustments are made;

[0076] Area 2 is the subjective adjustment area: if Only subjective coefficient adjustment is used;

[0077] Region 3 is the semi-complete penalty region: if Only subjective coefficient adjustment is used;

[0078] Area 4 is the complete penalty area: if Fully penalize non-cooperative clusters, reduce their cluster weights, and use objective coefficient adjustments;

[0079] Area 5 is the objective adjustment area: if Only objective coefficient adjustments are used;

[0080] Area 6 is the second cooperation area: if No adjustments are made.

[0081] Preferably, the S4 uses MULTIMOORA to obtain the final selected large group satellite emergency plan, specifically including:

[0082] S41, standard group decision preference matrix R c =(x ij ) m×n , and obtain the dimensionless matrix

[0083]

[0084] S42. Calculate the utility value based on the RS model of the following arithmetic integral and Get the first ranking of the large group satellite emergency plan in descending order;

[0085]

[0086] S43. Calculate the utility value based on the following RP model and The second sorting of the large group satellite emergency plan is obtained in ascending order;

[0087] Determine the maximum value r of each attribute index j the maximum distance to the relevant alternatives;

[0088]

[0089] in

[0090] S44. Calculate the utility value based on the geometric integral formula based on the following FMF model and The third ranking of the large group satellite emergency plan is obtained in descending order;

[0091]

[0092] S45. Use the advantage theory method to integrate the above three sorting results to obtain the final selected large-scale satellite emergency plan.

[0093] A large-scale satellite emergency plan decision-making system in a social network environment, comprising:

[0094] An acquisition module is used to obtain a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix uses a hesitant fuzzy binary semantic set as a decision opinion expression form;

[0095] A clustering module is used to cluster all decision makers into a number of clusters based on the decision preference matrix and the social network relationship diagram, and obtain the initial cluster decision preference matrix and the initial group decision preference matrix;

[0096] A consensus module, configured to update the cluster consensus level and the group consensus level based on the respective initial cluster decision preference matrices and the initial group decision preference matrix, based on a preset embedded feedback adjustment mechanism, and obtain a group decision preference matrix based on consensus;

[0097] The selection module is used to obtain a final selected large-group satellite emergency plan according to the group decision preference matrix.

[0098] A storage medium stores a computer program for making emergency plan decisions for a large group of satellites in a social network environment, wherein the computer program enables a computer to execute the above-mentioned method for making emergency plan decisions for a large group of satellites in a social network environment.

[0099] An electronic device, comprising:

[0100] one or more processors;

[0101] storage; and

[0102] One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including a method for executing the large-scale satellite emergency plan decision-making method in the social network environment as described above.

[0103] (3) Beneficial effects

[0104] The present invention provides a method, system, storage medium, and electronic device for making emergency plans for large-scale satellite groups in a social network environment. Compared with existing technologies, it has the following advantages:

[0105] The present invention includes obtaining a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix adopts a hesitant fuzzy binary semantic set as a decision opinion expression form; clustering all decision makers into a plurality of clusters according to the decision preference matrix and the social network relationship diagram, obtaining each initial cluster decision preference matrix and an initial group decision preference matrix; updating the cluster consensus level and the group consensus level based on a preset embedded feedback adjustment mechanism according to each of the initial cluster decision preference matrices and the initial group decision preference matrix, and obtaining a group decision preference matrix based on consensus; and obtaining a large-scale satellite emergency plan that is finally selected according to the group decision preference matrix.

[0106] On the one hand, the trust between decision makers (i.e., the trust between nodes) is taken into account during the clustering process, avoiding the default setting of the inter-node weighting factor to 1 for directed graphs, which plays a significant role in the accuracy of the clustering results of large satellite emergency plans. On the other hand, an embedded feedback adjustment mechanism is designed to serve the management of non-cooperative behavior, so that the opinions of individual decision makers and clusters are closer to the group, thereby improving the decision-making level. BRIEF DESCRIPTION OF THE DRAWINGS

[0107] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0108] Figure 1 A flowchart of a method for making emergency plans for large-scale satellite groups in a social network environment provided by an embodiment of the present invention;

[0109] Figure 2 A different representation scheme in social network analysis is provided for the embodiment of the present invention;

[0110] Figure 3 A schematic diagram of the division of cluster non-cooperative behavior areas provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0111] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0112] The embodiments of the present application provide a method, system, storage medium and electronic device for making emergency plans for large-scale satellites in a social network environment, thereby solving the technical problems that existing clustering methods use the consistency of decision-makers' decision preferences as the clustering criterion, ignoring the intrinsic connections between decision-makers; and the decision-making process ignores non-cooperative behavior, resulting in low accuracy in the consensus-building process.

[0113] The technical solution in the embodiments of the present application is to solve the above technical problems, and the overall idea is as follows:

[0114] An embodiment of the present invention includes obtaining a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix uses a hesitant fuzzy binary semantic set as a decision opinion expression form; clustering all decision makers into a plurality of clusters based on the decision preference matrix and the social network relationship diagram, obtaining each initial cluster decision preference matrix and an initial group decision preference matrix; updating the cluster consensus level and the group consensus level based on a preset embedded feedback adjustment mechanism according to each of the initial cluster decision preference matrices and the initial group decision preference matrix, and obtaining a group decision preference matrix based on consensus; and obtaining a finally selected large-scale satellite emergency plan based on the group decision preference matrix.

[0115] On the one hand, the trust between decision makers (i.e., the trust between nodes) is taken into account during the clustering process, avoiding the default setting of the inter-node weighting factor to 1 for directed graphs, which plays a significant role in the accuracy of the clustering results of large satellite emergency plans. On the other hand, an embedded feedback adjustment mechanism is designed to serve the management of non-cooperative behavior, so that the opinions of individual decision makers and clusters are closer to the group, thereby improving the decision-making level.

[0116] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0117] Example:

[0118] like Figure 1 As shown, an embodiment of the present invention provides a large-scale satellite emergency plan decision-making method in a social network environment, comprising:

[0119] S1. Obtaining a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix uses a hesitant fuzzy binary semantic set as a form of decision opinion expression;

[0120] S2. Clustering all decision makers into a number of clusters based on the decision preference matrix and the social network relationship diagram, and obtaining each initial cluster decision preference matrix and an initial group decision preference matrix;

[0121] S3. Based on each of the initial cluster decision preference matrices and the initial group decision preference matrix, the cluster consensus level and the group consensus level are updated based on a preset embedded feedback adjustment mechanism to obtain a group decision preference matrix based on consensus.

[0122] S4. Obtaining a finally selected large-group satellite emergency plan based on the group decision preference matrix.

[0123] In the embodiments of the present invention, on the one hand, the trust between decision makers is taken into account during the clustering process (i.e., the trust between nodes is taken into account), thereby avoiding the default setting of the inter-node weighting factor to 1 for directed graphs, which greatly improves the accuracy of the clustering results of large satellite emergency plans. On the other hand, an embedded feedback adjustment mechanism is designed to serve the management of non-cooperative behavior, so that the opinions of individual decision makers and clusters are closer to the group, thereby improving the decision-making level.

[0124] The following will introduce the various steps of the above solution in detail with specific content:

[0125] In step S1, a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers are obtained. The decision preference matrix uses a hesitant fuzzy binary semantic set as a decision opinion expression form.

[0126] Each expert decision maker DM provides his / her decision opinion according to the attributes of the satellite emergency decision plan, and constructs a decision preference matrix with hesitant fuzzy binary semantic sets (HF2TLSs).

[0127] The selected attributes can be selected according to actual conditions, such as task completion efficiency, including the completion rate and completion efficiency of emergency tasks; scheme performance, focusing on the disturbance to the original observation scheme; resource utilization, such as the conflict degree of the observation time window and satellite utilization rate.

[0128] In addition, since trust relationships have always been considered a reliable source for evaluating the importance of experts, they are usually studied by social network analysis (SNA). There are three representations in SNA analysis: actor groups, relationships themselves, and actor standards (e.g. Figure 2 shown).

[0129] The embodiment of the present invention only takes one type of social network, namely, a trust network, as an example. In this network, decision makers clearly express their opinions as trust and distrust statements.

[0130] SNA is used to analyze the social network relationship of decision makers. The social network relationship graph constructed is a social structure composed of a set of nodes E and a set of edges L, where node l i Representing decision makers Directed edges represent trust relationships. For example, an edge from e5 to e3 and e6 indicates that e5 trusts e3 and e6.

[0131] In step S2, all decision makers are clustered into several clusters according to the decision preference matrix and the social network relationship diagram, and the initial cluster decision preference matrix and the initial group decision preference matrix are obtained;

[0132] This step uses the modularity-based Louvain clustering algorithm for clustering, combining it with the aforementioned SNA analysis method to form a new clustering method. In the context of a large-scale satellite emergency plan, the present invention combines the SNA method with the fast and efficient Louvain algorithm to embed the inherent connections between multiple decision makers (DMs) into the clustering process. This not only increases the accuracy and timeliness of the clustering results, but also reduces clustering complexity.

[0133] The S2 specifically includes:

[0134] S21, treating each node of the social network relationship graph as a cluster, and then merging neighbor nodes of the cluster into the same cluster to obtain multiple clusters;

[0135] S22: Obtain weighting factors between different decision makers in the social network relationship graph based on the trust matrix of the social network relationship graph and the similarity between any two decision preference matrices; obtain the modularity of each cluster and the corresponding modularity change based on the weighting factors between different decision makers; specifically including:

[0136] S221. Use a tuple of type λ=(t,d) to represent the trust function value, where t,d∈[0,1], the first component t is a trust degree, and the second component d is a distrust degree.

[0137] The set of trust function values ​​(TFs), or trust function, will be represented by Λ = {λ = (t, d) | t, d∈ [0, 1]} ≡ [0, 1] 2 (indicates that it satisfies the property of triangular norm) indicates.

[0138] Define the trust matrix as Represents decision makers The trust between Represents decision makers The trust score between any two decision makers is expressed as

[0139] S222: Use the following formula to obtain the weighting factors between different decision makers in the social network relationship diagram:

[0140]

[0141] in, represents the similarity between any two decision preference matrices;

[0142] are the decision preference matrices of any two decision makers, i.e., the decision information of m large group satellite emergency plans on n attribute indicators;

[0143] in, They are the granularity of the corresponding hesitant fuzzy binary semantic set;

[0144]

[0145] express and Hausdorff distance between hesitant fuzzy binary semantic sets;

[0146] and It consists of several hesitant fuzzy binary semantic sets, the number of which is determined by the scheduling opinion; It is a language tag, and its value is S={s0,s1,…,s g} (e.g., S = {s0 = extremely poor, s1 = very poor, s2 = poor, s3 = average, s4 = good, s5 = very good, s6 = exceptionally good}); represents the sign conversion value and has a range of [-0.5, 0.5), g is the potential of S;

[0147] S223, obtaining the modularity Q0 of each cluster according to the weighting factors between the different decision makers;

[0148]

[0149] Wherein, M represents the sum of weights of all edges of the social network relationship graph; Represents decision makers The weighting factor of the edge between them. When the network is not a weighted graph, the weight of the edge is 1; Represents the sum of weighted factors of all edges connected to node l1 or l2 respectively; Respectively The cluster to which it belongs, if belong to the same cluster C, then otherwise

[0150] S224, and obtaining the corresponding modularity change ΔQ0;

[0151]

[0152] Among them, ∑in represents the sum of the weighted factors of all edges in cluster C, ∑tot represents the sum of the weighted factors of the edges connected to the points in cluster C, Represents the sum of the weighted factors of the edges connecting node l1 with the points in cluster C.

[0153] S23, repeating the above steps S21 to S22 until the overall modularity no longer changes, and finally obtaining several clusters;

[0154] This step will ultimately lead the decision maker Clustered into K clusters; the kth cluster is recorded as C k ,N k Indicates the number of members in the cluster.

[0155] S24. Obtaining the cluster weight and initial cluster decision preference matrix of each cluster based on the final clustering result; and obtaining the initial group decision preference matrix based on the cluster weight and initial cluster decision preference matrix of each cluster; specifically including:

[0156] based on Calculate the initial decision preference matrix of each cluster; where it is assumed that the decision maker Clustered into K clusters, the kth cluster is recorded as C k ,N k Indicates the number of members in the cluster;

[0157] based on Calculate cluster weights, where represents the weight of the decision maker, and

[0158] based on Calculate the initial group decision preference matrix.

[0159] In step S3, based on each of the initial cluster decision preference matrices and the initial group decision preference matrix, the cluster consensus level and the group consensus level are updated based on a preset embedded feedback adjustment mechanism to obtain a group decision preference matrix based on consensus. Specifically, the following steps are performed:

[0160] S31. Let t = 0, given the group consensus threshold

[0161] S32. Set t=t+1 and calculate the current cluster consensus level (Obviously, when t=1, the initial decision preference matrix and the initial group decision preference matrix are solved ):

[0162]

[0163] Among them, D e (G k(t) ,G f(t) ) is G k(t) and G f(t) The Hausfdorff distances of hesitant fuzzy binary semantic sets corresponding to the clusters;

[0164] Calculate the current group consensus level GCL t :

[0165]

[0166] judge and Then go to S35; otherwise, go to S33;

[0167] S33. Based on the current cluster consensus level and the previous cluster consensus level Get the cooperation coefficient used to measure the degree of change in the consensus level of a single cluster

[0168]

[0169] in, Assign values ​​according to actual conditions.

[0170] Determine the cooperation coefficient threshold

[0171]

[0172] Among them, ρ t Is the penalty coefficient, used to determine the cooperation coefficient How much smaller will it fall into the penalty area, satisfying 0≤ρ t ≤1;

[0173] S34, Comprehensive and as well as GCL t and The size comparison result is obtained, and the embedded feedback adjustment mechanism is introduced to adjust the cluster weight of the non-cooperative cluster and / or the decision preference matrix of its internal decision makers, update the current cluster decision preference matrix, and go to S32;

[0174] In particular, the embedded feedback regulation mechanism in S34 includes:

[0175] (1) Opinion modification includes two situations:

[0176] The first one, if The clusters that indicate the need for modification cooperate with the modification decision suggestion and only use the subjective coefficient adjustment;

[0177]

[0178] Among them, G k(t) represents the decision preference matrix of cluster Ck that needs to be modified after t iterations, is the subjective correction factor, determined by the cluster;

[0179] The second type, if Indicates that the cluster that needs to be modified does not cooperate with the modification decision suggestion, that is, there is an uncooperative cluster. The objective coefficient and subjective coefficient are used to modify the opinion (forced modification mode), including:

[0180] First, by solving the minimum similarity between the opinions of each decision maker and all other decision makers in the cluster that needs to be modified, the position of the decision preference matrix with the largest opinion difference in the cluster (the decision preference matrix that needs to be modified) is determined.

[0181]

[0182] in, Representing any two decision makers The similarity between the corresponding decision preference matrices;

[0183] Secondly, objectively adjust the decision preference matrix determined above.

[0184]

[0185] in, is the objective adjustment coefficient, and get;

[0186] It is not difficult to see that the first two steps represent a two-stage objective opinion modification process; the strategy is modified objectively, and the cluster at this time

[0187] Finally, if subjective adjustments are still required,

[0188]

[0189] It's easy to understand that subjective adjustments favor the cluster decision preference matrix, while objective adjustments focus on the decision preference matrix of the individual decision maker. In this adjustment model, based on the three-tiered structure of "decision maker-cluster-group," the individual decision matrix is ​​first aligned with the cluster decision matrix, then aligned with the group decision matrix. Ultimately, both the individual decision maker and the cluster's opinions converge toward the group, improving decision-making quality.

[0190] (2) Weight penalty;

[0191]

[0192] in, Represents cluster C k The cluster consensus level in the tth iteration, GCL t Represents cluster C k The group consensus level in the t-th iteration of Represents cluster C k The cluster weights at iteration t.

[0193] In the S34, Figure 3 As shown, the cluster weights of the non-cooperative cluster and / or the decision preference matrix of its internal decision makers are adjusted, specifically including:

[0194] Area 1 is the first cooperation area: if Not only is the level of individual decision-maker consensus higher than the group consensus, but the cluster cooperation coefficient is also above the critical point. This indicates that the cluster has adjusted its opinions even though it does not need to do so. Therefore, no further adjustments are needed to respect the cluster's own opinions.

[0195] Area 2 is the subjective adjustment area: if The decision maker's individual consensus level is higher than the group consensus level but falls short of the consensus threshold. Furthermore, the cluster's cooperation coefficient is also above the critical point. This indicates that while the cluster hasn't reached the threshold, it is willing to adjust its opinions. Therefore, it's sufficient to use only the subjective coefficient to adjust opinions closer to the group consensus.

[0196] Region 3 is the semi-complete penalty region: if The cluster consensus level is lower than the group consensus level, while the cluster cooperation coefficient is above the critical point. This indicates that while the cluster is willing to adjust its opinions, its opinions differ significantly from those of the group, failing to meet the consensus threshold. Therefore, its expression level within the group, or its weight, is reduced. At the same time, adjusting opinions using only the subjective coefficient can help reach consensus.

[0197] Area 4 is the complete penalty area: if The cluster consensus level is lower than the group consensus level, and the cluster cooperation coefficient is also below the critical point. This indicates that the cluster does not meet the consensus threshold conditions, but is unwilling to adjust its opinions too much, engaging in uncooperative behavior that is detrimental to consensus. It should be fully penalized, its weight reduced, and its opinions subjected to two-stage objective revisions using the aforementioned objective coefficients.

[0198] Area 5 is the objective adjustment area: if The cluster consensus level is higher than the group consensus level but lower than the threshold, while the cluster cooperation coefficient is lower than the critical point. However, this does not reduce its weight. This is because after the consensus iteration, the cluster consensus level still exceeds the group consensus level. Therefore, the cluster made a small change, which does not indicate a lack of cooperation. This shows that its opinion and the given adjustment coefficient still have a positive impact on consensus. In other words, there is no need to penalize its weight. It is only necessary to use the above objective coefficient to objectively modify its opinion in two stages to bring it closer to the group decision.

[0199] Area 6 is the second cooperation area: if No adjustments are made. The cluster's consensus level is above the group consensus threshold, while the cluster's cooperation coefficient is below the critical point. Similar to region 5, this is because the consensus threshold has been exceeded and no further changes are necessary. This does not necessarily mean that the level of cooperation is low; rather, because the threshold has been exceeded, it is considered to be highly cooperative and respectful of group opinion. In return, the cluster is treated the same as region 1, respecting its own integrity.

[0200] In most previous studies, there are few studies on non-cooperative behavior, and the management methods of non-cooperative behavior are single, especially for large-scale emergency decision-making. No scholar has yet conducted research on this. More importantly, almost all studies are based on the premise that decision-making experts are independent of each other and do not influence each other. However, with the development of social networks, this premise has become difficult to meet. The embodiment of the present invention analyzes the degree of non-cooperation of the cluster in detail, divides the cluster non-cooperative behavior into six categories according to the cooperation coefficient and the cluster consensus level, and proposes different preference adjustment mechanisms to construct a new consensus model, so as to better manage the non-cooperative behavior of one or more clusters in LGSESDM, so as to optimize the consensus-building process of large-scale satellite emergency decision-making.

[0201] S35, let t * =t, output the final cluster decision preference matrix Get the final group decision matrix R c , and serves as the group decision preference matrix based on the consensus reached.

[0202] In step S4, the final selected large-group satellite emergency plan is obtained according to the group decision preference matrix.

[0203] In this step, MULTIMOORA is used to obtain the final selected large-scale satellite emergency solution. The MOORA (Multiobjective Optimization By Ratio Analysis) method was originally introduced by Brauers and Zavadskas and further enhanced by adding the full multiplication method (FMF) to generate the multiplication MOORA (MULTIMOORA) method.

[0204] MULTIMOORA is an emerging multi-attribute decision-making method. Compared to other methods, it is simple, effective, and easy to rank and optimize. MULTIMOORA is based on three subordinate methods: the ratio system (RS), the reference point (RP), and the fully multiplicative method (FMF). It also uses dominance theory to calculate the final ranking. Therefore, the MULTIMOORA method can improve the accuracy of the final selection process for satellite emergency options.

[0205] The S4 specifically includes:

[0206] S41, standard group decision preference matrix R c =(x ij ) m×n , and obtain the dimensionless matrix

[0207]

[0208] S42. Calculate the utility value based on the RS model of the following arithmetic integral and Get the first ranking of the large group satellite emergency plan in descending order;

[0209]

[0210] S43. Calculate the utility value based on the following RP model and The second sorting of the large group satellite emergency plan is obtained in ascending order;

[0211] Determine the maximum value r of each attribute index j the maximum distance to the relevant alternatives;

[0212]

[0213] in

[0214] S44. Calculate the utility value based on the geometric integral formula based on the following FMF model and The third ranking of the large group satellite emergency plan is obtained in descending order;

[0215]

[0216] S45. Use the advantage theory method to integrate the above three sorting results to obtain the final selected large-scale satellite emergency plan.

[0217] An embodiment of the present invention provides a large-scale satellite emergency plan decision-making system in a social network environment, comprising:

[0218] An acquisition module is used to obtain a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix uses a hesitant fuzzy binary semantic set as a decision opinion expression form;

[0219] A clustering module is used to cluster all decision makers into a number of clusters based on the decision preference matrix and the social network relationship diagram, and obtain the initial cluster decision preference matrix and the initial group decision preference matrix;

[0220] A consensus module, configured to update the cluster consensus level and the group consensus level based on the respective initial cluster decision preference matrices and the initial group decision preference matrix, based on a preset embedded feedback adjustment mechanism, and obtain a group decision preference matrix based on consensus;

[0221] The selection module is used to obtain a final selected large-group satellite emergency plan according to the group decision preference matrix.

[0222] An embodiment of the present invention provides a storage medium storing a computer program for making emergency plan decisions for a large group of satellites in a social network environment, wherein the computer program enables a computer to execute the above-mentioned method for making emergency plan decisions for a large group of satellites in a social network environment.

[0223] An embodiment of the present invention further provides an electronic device, including:

[0224] one or more processors;

[0225] storage; and

[0226] One or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including a method for executing the large-scale satellite emergency plan decision-making method in the social network environment as described above.

[0227] It can be understood that the large-scale satellite emergency plan decision-making system, storage medium and electronic device in a social network environment provided by the embodiments of the present invention correspond to the large-scale satellite emergency plan decision-making method in a social network environment provided by the embodiments of the present invention. The explanation, examples and beneficial effects of the relevant contents can refer to the corresponding parts of the large-scale satellite emergency plan decision-making method in a social network environment based on blockchain, and will not be repeated here.

[0228] In summary, compared with the existing technology, the present invention has the following beneficial effects:

[0229] 1. In the embodiments of the present invention, on the one hand, the trust between decision makers is taken into account during the clustering process (i.e., the trust between nodes is taken into account), thereby avoiding the default setting of the inter-node weighting factor to 1 for directed graphs, which greatly improves the accuracy of the clustering results of large satellite emergency plans. On the other hand, an embedded feedback adjustment mechanism is designed to serve the management of non-cooperative behavior, so that the opinions of individual decision makers and clusters are closer to the group, thereby improving the decision-making level.

[0230] 2. In the context of the large-scale satellite emergency plan of the embodiment of the present invention, the SNA method is combined with the Louvain algorithm with fast clustering speed and high efficiency to embed the inherent connections between multiple decision makers (DMs) into the clustering process; on the one hand, it increases the accuracy and timeliness of the clustering results, and on the other hand, it reduces the clustering complexity.

[0231] 3. The embodiment of the present invention analyzes the degree of non-cooperation of clusters in detail, divides the non-cooperative behaviors of clusters into six categories according to the cooperation coefficient and cluster consensus level, proposes different preference adjustment mechanisms, constructs a new consensus model, and better manages the non-cooperative behaviors of one or more clusters in LGSESDM to optimize the consensus-building process of large-scale satellite emergency decision-making.

[0232] 4. The embodiment of the present invention uses the MULITMOORA method to improve the accuracy and robustness of the solution selection result.

[0233] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0234] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A large-scale satellite emergency plan decision-making method in a social network environment, characterized by: include: S1. Obtaining a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix uses a hesitant fuzzy binary semantic set as a form of decision opinion expression; The acquisition process of the decision preference matrix includes: Obtaining the decision opinions provided by each decision maker regarding the attributes of the satellite emergency decision plan; the attributes are mission completion benefit, plan performance, or resource utilization, wherein the mission completion benefit includes the completion rate or completion benefit of the emergency mission, the plan performance is the focus on the disturbance to the original observation plan, and the resource utilization is the conflict degree of the observation time window or the satellite utilization rate; Using hesitant fuzzy binary semantic sets as the decision opinion expression form to obtain the decision preference matrix; S2. Clustering all decision makers into a number of clusters based on the decision preference matrix and the social network relationship diagram, and obtaining each initial cluster decision preference matrix and an initial group decision preference matrix; S3. Based on each of the initial cluster decision preference matrices and the initial group decision preference matrix, the cluster consensus level and the group consensus level are updated based on a preset embedded feedback adjustment mechanism to obtain a group decision preference matrix based on consensus. S4. Obtaining a final selected large-group satellite emergency plan based on the group decision preference matrix; In S2, the Louvain clustering algorithm based on modularity is used for clustering; specifically, the following steps are performed: S21, treating each node of the social network relationship graph as a cluster, and then merging neighbor nodes of the cluster into the same cluster to obtain multiple clusters; S22. Obtaining weighting factors between different decision makers in the social network relationship graph based on the trust matrix of the social network relationship graph and the similarity between any two decision preference matrices; obtaining the modularity of each cluster and the corresponding modularity change based on the weighting factors between different decision makers; including: S221. Use a tuple of type λ=(t, d) to represent the trust function value, where t, d∈[0, 1], the first component t is a trust degree, and the second component d is a distrust degree; define the trust matrix as Represents decision makers The trust between Represents decision makers The trust score between any two decision makers is expressed as S222: Use the following formula to obtain the weighting factors between different decision makers in the social network relationship diagram: in, represents the similarity between any two decision preference matrices; are the decision preference matrices of any two decision makers, i.e., the decision information of m large group satellite emergency plans on n attribute indicators; in, They are the granularity of the corresponding hesitant fuzzy binary semantic set; express and Hausdorff distance between hesitant fuzzy binary semantic sets; and It consists of several hesitant fuzzy binary semantic sets, the number of which is determined by the scheduling opinion; is a language tag, whose value is S={s0, s1, ..., s g }; represents the sign conversion value and has a range of [-0.5, 0.5), g is the potential of S; S223, obtaining the modularity Q0 of each cluster according to the weighting factors between the different decision makers; Wherein, M represents the sum of weights of all edges of the social network relationship graph; Represents decision makers The weighting factor of the edge between them. When the network is not a weighted graph, the weight of the edge is 1; Represents the sum of weighted factors of all edges connected to node l1 or l2 respectively; Respectively The cluster to which it belongs, if belong to the same cluster C, then otherwise S224, and obtaining the corresponding modularity change ΔQ0; Among them, ∑in represents the sum of the weighted factors of all edges in cluster C, ∑tot represents the sum of the weighted factors of the edges connected to the points in cluster C, represents the sum of the weighted factors of the edges connecting node l1 with the points in cluster C; S23, repeating the above steps S21 to S22 until the overall modularity no longer changes, and finally obtaining several clusters; S24. Obtaining the cluster weight and initial cluster decision preference matrix of each cluster according to the final clustering result; and obtaining the initial group decision preference matrix according to the cluster weight and initial cluster decision preference matrix of each cluster.

2. The method for making emergency plans for large-scale satellite groups in a social network environment as claimed in claim 1, wherein: The S24 specifically includes: based on Calculate the initial decision preference matrix for each cluster; where it is assumed that the decision maker Clustered into K clusters, the kth cluster is recorded as C k , N k Indicates the number of members in the cluster; based on Calculate cluster weights, where represents the weight of the decision maker, and based on The initial group decision preference matrix is ​​calculated.

3. The method for making emergency plans for large-scale satellite groups in a social network environment as claimed in claim 2, wherein: S3 specifically includes: S31. Let t = 0, given the group consensus threshold S32. Set t=t+1 and calculate the current cluster consensus level Among them, D e (G k(t) , G f(t) ) is G k(t) and G f(t) The Hausfdorff distances of hesitant fuzzy binary semantic sets corresponding to the clusters; Calculate the current group consensus level GCL t : judge and Then go to S35; otherwise, go to S33; S33. Based on the current cluster consensus level and the previous cluster consensus level Get the cooperation coefficient used to measure the degree of change in the consensus level of a single cluster in, Assign values ​​according to actual conditions; Determine the cooperation coefficient threshold Among them, ρ t Is the penalty coefficient, used to determine the cooperation coefficient How much smaller will it fall into the penalty area, satisfying 0≤ρ t ≤1; S34, Comprehensive and as well as GCL t and The size comparison result is obtained, and the embedded feedback adjustment mechanism is introduced to adjust the cluster weight of the non-cooperative cluster and / or the decision preference matrix of its internal decision makers, update the current cluster decision preference matrix, and go to S32; S35, let t * =t, output the final cluster decision preference matrix Get the final group decision matrix R c , and serves as the group decision preference matrix based on the consensus reached.

4. The method for making emergency plans for large-scale satellite groups in a social network environment as claimed in claim 3, wherein: The embedded feedback regulation mechanism in S3 includes: (1) Opinion modification includes two situations: The first one, if Clusters that indicate the need for modification cooperate with modification decision suggestions, and only subjective coefficient adjustments are used; Among them, G k(t) Indicates the cluster C that needs to be modified after t iterations k The decision preference matrix, is the subjective correction factor, determined by the cluster; The second type, if Indicates that the cluster that needs to be modified does not cooperate with the modification of the decision suggestion, that is, there is an uncooperative cluster. The objective coefficient and subjective coefficient are used to modify the opinion, including: First, by solving the minimum similarity between the opinions of each decision maker in the cluster that needs to be modified and all other decision makers, the position of the decision preference matrix with the largest opinion difference in the cluster is determined. in, Representing any two decision makers The similarity between the corresponding decision preference matrices; Secondly, objectively adjust the decision preference matrix determined above. in, is the objective adjustment coefficient, and get; Through objective modification, the cluster at this time Finally, if subjective adjustments are still required, (2) Weight penalty; in, Represents cluster C k The cluster consensus level in the tth iteration, GCL t Represents cluster C k The group consensus level in the t-th iteration of Represents cluster C k The cluster weights at iteration t.

5. The method for making emergency plans for large-scale satellite groups in a social network environment as claimed in claim 4, characterized in that: In S34, adjusting the cluster weight of the non-cooperative cluster and / or the decision preference matrix of its internal decision makers specifically includes: Area 1 is the first cooperation area: if No adjustments are made; Area 2 is the subjective adjustment area: if Only subjective coefficient adjustment is used; Region 3 is the semi-complete penalty region: if Only subjective coefficient adjustment is used; Area 4 is the complete penalty area: if Fully penalize non-cooperative clusters, reduce their cluster weights, and use objective coefficient adjustments; Area 5 is the objective adjustment area: if Only objective coefficient adjustments are used; Area 6 is the second cooperation area: if No adjustments are made.

6. The method for making emergency plans for large-scale satellite groups in a social network environment according to any one of claims 1 to 5, characterized in that: In S4, MULTIMOORA is used to obtain the final selected large group satellite emergency plan, which specifically includes: S41, standard group decision preference matrix R c =(x ij ) m×n , and obtain the dimensionless matrix S42. Calculate the utility value based on the RS model of the following arithmetic integral and Get the first ranking of the large group satellite emergency plan in descending order; S43. Calculate the utility value based on the following RP model and The second sorting of the large group satellite emergency plan is obtained in ascending order; Determine the maximum value r of each attribute index j the maximum distance to the relevant alternatives; in S44. Calculate the utility value based on the geometric integral formula based on the following FMF model and The third ranking of the large group satellite emergency plan is obtained in descending order; S45. Use the advantage theory method to integrate the above three sorting results to obtain the final selected large-scale satellite emergency plan.

7. A large-scale satellite emergency plan decision-making system in a social network environment, characterized by: The method for executing a large-scale satellite emergency plan decision-making method in a social network environment as claimed in claim 1 comprises: An acquisition module is used to obtain a decision preference matrix of multiple satellite emergency decision makers and a social network relationship diagram between the decision makers, wherein the decision preference matrix uses a hesitant fuzzy binary semantic set as a decision opinion expression form; A clustering module is used to cluster all decision makers into a number of clusters based on the decision preference matrix and the social network relationship diagram, and obtain the initial cluster decision preference matrix and the initial group decision preference matrix; A consensus module, configured to update the cluster consensus level and the group consensus level based on the respective initial cluster decision preference matrices and the initial group decision preference matrix, based on a preset embedded feedback adjustment mechanism, and obtain a group decision preference matrix based on consensus; The selection module is used to obtain a final selected large-group satellite emergency plan according to the group decision preference matrix.

8. A storage medium, characterized in that: The computer program for making emergency plan decisions for a large group of satellites in a social network environment is stored therein, wherein the computer program enables a computer to execute the emergency plan decision method for a large group of satellites in a social network environment according to any one of claims 1 to 6.

9. An electronic device, characterized in that: include: one or more processors; Memory; and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the programs including a method for executing the large-scale satellite emergency plan decision-making method in a social network environment according to any one of claims 1 to 6.

Images