CN117436282B - Architecture optimization method based on dual variable weight and TOPSIS-gray correlation - Google Patents
Architecture optimization method based on dual variable weight and TOPSIS-gray correlation Download PDFInfo
- Publication number
- CN117436282B CN117436282B CN202311669373.3A CN202311669373A CN117436282B CN 117436282 B CN117436282 B CN 117436282B CN 202311669373 A CN202311669373 A CN 202311669373A CN 117436282 B CN117436282 B CN 117436282B
- Authority
- CN
- China
- Prior art keywords
- index
- weight
- command control
- network architecture
- weights
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 186
- 238000005457 optimization Methods 0.000 title claims abstract description 23
- 230000009977 dual effect Effects 0.000 title claims description 5
- 238000011156 evaluation Methods 0.000 claims abstract description 27
- 230000008859 change Effects 0.000 claims abstract description 19
- 238000012163 sequencing technique Methods 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 8
- 238000004458 analytical method Methods 0.000 claims abstract description 5
- 230000008569 process Effects 0.000 claims description 36
- 239000011159 matrix material Substances 0.000 claims description 35
- 230000007246 mechanism Effects 0.000 claims description 13
- 230000002776 aggregation Effects 0.000 claims description 8
- 238000004220 aggregation Methods 0.000 claims description 8
- 230000008901 benefit Effects 0.000 claims description 8
- 238000004891 communication Methods 0.000 claims description 8
- 239000013598 vector Substances 0.000 claims description 8
- 238000012216 screening Methods 0.000 claims description 7
- 230000005540 biological transmission Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000006870 function Effects 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 5
- 239000004020 conductor Substances 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 102000002274 Matrix Metalloproteinases Human genes 0.000 claims description 2
- 108010000684 Matrix Metalloproteinases Proteins 0.000 claims description 2
- 238000012937 correction Methods 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 108010015046 cell aggregation factors Proteins 0.000 claims 1
- 238000004422 calculation algorithm Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 239000006185 dispersion Substances 0.000 description 4
- 238000012847 principal component analysis method Methods 0.000 description 4
- 238000004088 simulation Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000000739 chaotic effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 239000011888 foil Substances 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000005312 nonlinear dynamic Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/18—Network design, e.g. design based on topological or interconnect aspects of utility systems, piping, heating ventilation air conditioning [HVAC] or cabling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Geometry (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a framework optimization method based on double variable weights and TOPSIS-gray correlation, which comprises the steps of constructing a command control network framework, constructing a command control network framework index system, determining subjective weights of indexes by using a hierarchical analysis method, determining objective weights of the indexes by using an improved entropy weight method, and combining the subjective and objective index weights according to subjective and objective weight factors to obtain a first layer of index weights; carrying out variable weight processing on the first layer index weight by utilizing a variable weight theory to obtain a second layer weight; and respectively solving target evaluation values of index variable weights by using a TOPSIS method and a gray correlation method, and giving out a generated network architecture sequencing result. The method synthesizes the subjective and objective factors of the indexes and the influence of the state on the weight change, the sequencing result synthesizes the distance and shape change among the indexes, and the selected command control network architecture is more in line with the actual battlefield environment.
Description
Technical Field
The invention relates to the technical field of information engineering, in particular to a framework optimization method based on double variable weights and TOPSIS-gray correlation.
Background
In modern informatization war, the two parties of the battle are no longer the fight of single army or mode, but all-round fight is carried out between the battle system consisting of the battle elements of scout detection, command control, fire striking, and the like. In the system countermeasure, the composition architecture of the battle system is complex and various, and screening of different composition architectures can directly influence the battle win or lose. Therefore, how to prefer the command control architecture network architecture becomes an important issue under the informatization condition.
In the aspect of network architecture optimization of a command control system, a plurality of methods exist in the prior art, such as complex network theory is applied to build an optimization model and an algorithm framework; by optimizing indexes such as average path length and aggregation coefficient of a network structure, further optimizing a command control system network architecture, a command control network structure optimizing method based on a hierarchical-entropy weight method is provided in the prior art, and the algorithm has a higher network reconstruction effect and a lower network reconstruction cost, so that the algorithm is more suitable for being used in a C2 network, however, the algorithm considers local and global information of the network and hierarchical characteristics and command relations, and the complexity of the algorithm is increased.
Currently, the following three main methods exist in the prior art: the method solves the problem of multi-objective architecture optimization in military aspects by using a dispersion maximization method, a principal component analysis method and a projection algorithm, but the three methods respectively have the defects of too single weight, incapability of reflecting the contribution of each single evaluation method in a combined evaluation method, information loss problem in index normal standardization, incapability of effectively distinguishing schemes with similar ideal projection values and the like.
Disclosure of Invention
The purpose of the invention is that: aiming at the problem that a plurality of command control network architectures generated for a specific combat mission are difficult to screen out an optimal architecture in a battlefield environment, the invention provides an architecture optimization method based on double variable weights and TOPSIS-grey association, and aims to sort a series of generated command control network architectures and screen out the optimal command control network architecture.
In order to realize the functions, the invention designs a framework optimization method based on double variable weights and TOPSIS-gray correlation, and the following steps S1-S7 are executed to finish the screening of an optimal command control network framework aiming at the combat mission of the battlefield environment:
step S1: constructing each command control network architecture according to a battlefield environment, wherein each command control network architecture is divided into three command control layers, namely an information network, a command control network and a fire striking network;
step S2: according to a complex network theory, abstracting information exchange relation among command control layers, communication transmission paths, interaction relation among command mechanisms among the command control layers and membership relation among the command mechanisms of a command control network architecture, and describing by adopting an undirected communication graph to establish a command control network architecture model;
step S3: constructing a command network architecture index system, wherein the command network architecture index system comprises a plurality of indexes for evaluating the performance of the established command control network architecture model, and optimizing the command control network architecture according to the indexes;
step S4: weighting each index in the index system of the index control network architecture respectively, wherein the weights comprise subjective weights and objective weights, subjective weights are assigned to each index by using a hierarchical analysis method, objective weights are assigned to each index by using an improved entropy weight method, and the subjective weights and the objective weights of each index are combined according to subjective and objective factors to obtain a first-layer index weight;
step S5: carrying out variable weight processing on the first layer of index weights by utilizing a variable weight theory to obtain variable weights which change according to the change of the target state values, and obtaining second layer of index weights according to the variable weights;
step S6: calculating a target evaluation value of the index weight of the second layer by using a TOPSIS method and a gray correlation method respectively;
step S7: and determining a target comprehensive evaluation value by combining target evaluation values of the second-layer index weights calculated by the TOPSIS method and the gray correlation method, sequencing each constructed command control network architecture, and completing screening of the optimal command control network architecture by taking the command control network architecture with the highest sequencing as the optimal command control network architecture.
The beneficial effects are that: the advantages of the present invention over the prior art include:
the invention designs a framework optimization method based on double variable weights and TOPSIS-gray correlation, firstly, respectively solving subjective and objective weights of target indexes by using a hierarchical analysis method and an improved entropy weight method, and combining to obtain a first layer weight; constructing a comprehensive equalization function, and fusing a variable weight theory and the first layer weight to form a final weight; finally, comprehensively obtaining a multi-target architecture optimization result according to the TOPSIS method and the gray correlation method. Simulation shows that the method provided by the invention integrates the influence of various factors and states of the index on weight change, and is more in line with the actual battlefield environment; the sequencing result integrates the distance and shape change among indexes, and the comparison result with other methods also shows the effectiveness and superiority of the method.
Drawings
FIG. 1 is a flow chart of a preferred method of architecture based on dual variant weights and TOPSIS-gray correlation provided in accordance with an embodiment of the present invention;
fig. 2 is a schematic diagram of a physical structure of a command control network architecture according to an embodiment of the present invention;
fig. 3 is a diagram of a command and control network architecture provided in accordance with an embodiment of the present invention;
FIG. 4 is a schematic diagram of a finger control network architecture index system according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a conventional infrastructure provided in accordance with an embodiment of the invention;
fig. 6 is a schematic diagram of a command control network architecture provided in accordance with an embodiment of the present invention;
FIG. 7 is a comparison of variable weight distribution provided in accordance with an embodiment of the present invention;
FIG. 8 is a diagram provided in accordance with an embodiment of the present inventionμEach target command control network architecture ordering result when=0;
FIG. 9 is a diagram provided in accordance with an embodiment of the present inventionμEach target command control network architecture ordering result when=1;
FIG. 10 is a diagram provided in accordance with an embodiment of the present inventionμEach target directs control network architecture ordering results when=0.5.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present invention, and are not intended to limit the scope of the present invention.
The architecture optimization method based on double variable weights and TOPSIS-gray correlation provided by the embodiment of the invention, referring to FIG. 1, performs the following steps S1-S7, and finishes the screening of the optimal command control network architecture aiming at the combat mission of the battlefield environment:
step S1: constructing each command control network architecture according to a battlefield environment, wherein each command control network architecture is divided into three command control layers, namely an information network, a command control network and a fire striking network;
referring to fig. 2, under modern informatization war, the command control network architecture comprises land, sea, air and other multidimensional space combat forces, and integrated combined combat can be realized by overall command. From the networked perspective, the command control network architecture has the characteristics of huge node quantity, compact internal connection and complex network structure, and the whole network consists of an information network, a command control network and a fire striking network.
Referring to fig. 3, the command mechanism in the information network includes a sensing unit and an information center unit, and mainly includes a space reconnaissance satellite, a beyond-view radar, an unmanned reconnaissance plane, and an information center processing vehicle; in fig. 3, the sensing unit is denoted by O, and the information center unit is denoted by P; the command mechanism in the command control network comprises a command control unit, and mainly comprises a combat command post and a communication command vehicle; the command and control unit is denoted by C in fig. 3; the command mechanism in the fire striking network comprises a fire striking unit and a soft killing unit, and mainly comprises an air defense missile, an intelligent ammunition system, a foil strip bomb, a high-power microwave weapon and an infrared bait; in fig. 3, F represents a fire striking unit, and S represents a soft killer unit;
the information network is mainly used for providing information such as battlefield situation and the like, carrying out real-time evaluation on recognition tracking and striking effects on a battlefield target, transmitting corresponding data to the information center unit for data processing and fusion, providing battlefield perception function, and transmitting the result to a battlefield command post or a communication command vehicle. The command control network is used for receiving battlefield information provided by the information center unit, so as to conduct command decision, transmit target indication information to the fire striking network, conduct fire distribution and the like. The fire striking network receives target indication data provided by the battle command post and the sensing unit and completes the fire striking task.
Step S2: according to a complex network theory, abstracting information exchange relation among command control layers, communication transmission paths, interaction relation among command mechanisms among the command control layers and membership relation among the command mechanisms of a command control network architecture, and describing by adopting an undirected communication graph to establish a command control network architecture model;
step S2, constructing node sets in each finger control hierarchyWherein the nodeFor abstract representations of command authorities in command hierarchies,nthe total number of nodes in the command hierarchy; constructing a set of edges in each finger hierarchy +.>Wherein the edge->Representing the edges of the links between the nodes,mto refer to the total number of edges in the hierarchy; according to node setVSum edge collectionEConstructing an undirected connected graphG=（V,E）。
Step S3: constructing a command network architecture index system, wherein the command network architecture index system comprises a plurality of indexes for evaluating the performance of the established command control network architecture model, and optimizing the command control network architecture according to the indexes;
the command control network architecture is preferably in the generated command control network architecture, and aims at improving the stability and flexibility of the command control network combat, and a multi-objective optimization model is established. Because the complex battlefield environment has the characteristics of strong countermeasure and high maneuver, the system is required to be stable and flexible, so as to refer to fig. 4, each index contained in the index system of the command network architecture in step S3 is respectively combat capability, average path length, aggregation coefficient and network density;
the combat capability of the command network architecture is described by introducing information entropy theory, and the influence of subjective factors can be well overcome by introducing information entropy. The information entropy is uncertainty of quantized information by borrowing the chaotic degree of the description molecular state in chemistry, and is obtained by a calculation method of the information entropy in a command control network architecture, and the probability error of the information source is provided in the capability attribute model of each nodep i Meanwhile, uncertainty is brought to the edge connecting process, the information entropy is used for quantifying the uncertainty, the combat capability is calculated based on the information entropy, the method belongs to benefit indexes, and the calculation of the information entropy is as follows:
；
in the method, in the process of the invention,p i for co-operating combat nodesv i Is a function of the information source probability error of (a),H i for co-operating combat nodesv i Is an information entropy of (a);
the uncertainty of describing the information source in the node edge process is quantified through the concept of information entropy, and the information entropy of a single command control network architecture is as follows:
；
in the method, in the process of the invention,qfor co-operating combat nodesv i Is used in the number of (a) and (b),H j for a single nodev j Is used for the information entropy of (a),bas a number of individual nodes,H l controlling a network for a single conductorArchitecture for a computer systemlIs an information entropy of (a);
the combat ability is as follows:
；
in the method, in the process of the invention,Cthe ability to perform a battle is indicated,C l controlling network architecture for single conductorlIs a combat capability of (1);
the instantaneous change of battlefield situation, the timeliness requirement of battlefield information transmission is higher and higher. The average path length is one of important indexes for measuring the command efficiency of the command control network architecture, and can effectively reflect the transmission efficiency of information in the command control network architecture. The smaller the average path length is, the higher the information transmission efficiency is, the faster the command issuing speed is, and the method belongs to the cost index. The average path is as follows:
；
in the method, in the process of the invention,Lthe average path length is indicated as such,d ij representing arbitrary nodes in command control network architecturev i Andv j the shortest distance between;Nrepresenting the total number of nodes of the command control network architecture;
the aggregate coefficient of a node represents the probability of being interconnected between two nodes in the network that are connected to the same node. Are commonly used to characterize the local structural nature of a network, describing the degree of close association, i.e., aggregation, of the network. The requirements of future war on the interconnection and intercommunication capability between the command control network architecture fight nodes are higher and higher, the cooperation relationship between the nodes is tighter and tighter, and the method belongs to benefit indexes. Task interoperability between the units of action may be measured using an aggregate factor, which is defined as:
；
in the method, in the process of the invention,representing nodesv i Is used for the aggregation coefficient of (a),w ij representing nodesv i Sum nodev j The edge weight of the middle part is calculated,w jk representing nodesv j Sum nodev k The edge weight of the middle part is calculated,w ki representing nodesv k Sum nodev i Edge weight between nodesv i Sum nodev k Is a nodev j Is a neighbor node of (a);
the network density belongs to benefit indexes, and is used for describing the degree of density of interconnected edges among nodes in a network, and the higher the network density is, the more closely the nodes are connected, and the following formula is calculated:
；
in the method, in the process of the invention,NandMrespectively representing the total number of nodes of the command control network architecture and the actual connecting edges,D i representing nodesv i Is a network density of (a).
Step S4: weighting each index in the index system of the index control network architecture respectively, wherein the weights comprise subjective weights and objective weights, subjective weights are assigned to each index by using a hierarchical analysis method, objective weights are assigned to each index by using an improved entropy weight method, and the subjective weights and the objective weights of each index are combined according to subjective and objective factors to obtain a first-layer index weight;
the subjective weight determining method of each index in step S4 is as follows:
assigning values to the important relations between every two indexes respectively, and constructing a judgment matrixIn a matrixa ij Representation matrixAMiddle (f)iIndividual elements and the firstjComparing the importance of the individual elements with the result value; in one embodiment, two indexes are usedThe important relationship between the two is given in table 1 below:
determination matrixAMaximum feature root of (2)λ max Corresponding feature vectorwThe following formula is shown:
；
for characteristic vectorwAnd carrying out consistency detection, wherein the consistency detection is shown in the following formula:
；
；
in the method, in the process of the invention,CIto judge matrixAIs used for the general consistency index of (c),RIas an index of the average random consistency,CRif the consistency is detected as an indexCR<0.1, then judge the judgment matrixAIf the consistency requirement is met, otherwise, returning to adjust the judgment matrixAUp to the judgment matrixAMeets the consistency requirement, and the feature vectorwSubjective weight as index after normalizationw s The method comprises the steps of carrying out a first treatment on the surface of the In one embodiment, the average random uniformity indexRIThe values are given in table 2 below:
the objective weight determining method of each preset index in the step S4 is as follows:
the entropy weight method determines objective weight according to the variability of the data index, and the larger the index discrete degree is, the more the implicit information quantity is, and the information entropy isE j The smaller the weight, the greater the weight. The solution is provided by an improved entropy weight methodx ij Is the firstiFirst command control network architecturejThe value of the individual index(s),i=1,2,…，m，j=1,2,…，nthe method comprises the steps of carrying out a first treatment on the surface of the Normalizing each index:
；
in the method, in the process of the invention,y ij is thatx ij Normalized values;
calculate the firstiFirst command control network architecturejInformation entropy of individual indexz ij The formula is as follows:
；
；
in the method, in the process of the invention,E j represent the firstjInformation entropy of individual index, wheny ij When=0, use 10 -4 Correctiony ij ；
The calculation formula of the original entropy weight method is as follows:
；
W e (j) For calculating by the original entropy weight methodjObjective weights of the individual indicators; since the entropy weight varies exponentially with the slight variation of the entropy value in the original entropy weight calculation formula, it is unreasonable to improve the entropy weight as follows:
according to the firstjInformation entropy of individual indexE j The objective weight is calculated as follows:
；
in the method, in the process of the invention,W ae (j) Watch (watch)Show the firstjObjective weights of the individual indicators;
combining subjective weightsw s Objective weightw ae Obtaining a first layer of index weightw pre The formula is as follows:
；
in the method, in the process of the invention,βrepresenting the ratio coefficient of subjective weight to first layer index weight, 1-βAnd the scale coefficient of the objective weight to the index weight of the first layer is represented. In order to obtain the optimal combination weight, the obtained subjective weight and objective weight are subjected to equalization treatment, so that the deviation between the subjective weight and the optimal weight is minimized, namely, an objective function is established as follows:
；
in the middle ofzRepresenting the best weight bias, further can be obtained:
；
for the above typeβAfter deriving, let it be 0, obtain the optimal weight adjustment factorβ=0.5。
Step S5: carrying out variable weight processing on the first layer of index weights by utilizing a variable weight theory to obtain variable weights which change according to the change of the target state values, and obtaining second layer of index weights according to the variable weights;
in step S5, the first layer index weight is subjected to variable weight processing by utilizing a variable weight theory, and a second layer index weight obtaining method is as follows:
the equalization sub-function is constructed as follows:
；
；
；
in the method, in the process of the invention,F(P j )、F(P 1 ，…，P 4 ) In order to equalize the sub-functions,P j represent the firstjThe number of the indexes is equal to the number of the indexes,j=1 indicates the ability to combat,j=2 represents the average path length,j=3 represents an aggregation coefficient,j=4 denotes network density;αto adjust the factor, in one embodiment,α=2；
according to the weight-changing theory, deriving the equalization subfunction to obtain a state weight-changing vectorS(P 1 ，…，P 4 ) The following formula:
；
since the final weight sum is 1,S(P 1 ) When the index weight is smaller, the other index weights are correspondingly increased, and the index weight is shown as the following formula:
；
in the method, in the process of the invention,is the firstjThe first layer of index weights of the individual indices,W(P j ) Is the firstjIndividual indexP j Is a second level of index weight.
Step S6: calculating a target evaluation value of the index weight of the second layer by using a TOPSIS method and a gray correlation method respectively;
in step S6, the method for calculating the target evaluation value of the index weight of the second layer by using the TOPSIS method is as follows:
for containingmPersonal command control network architecturenDecision-making problems to be ordered of individual indexes, and establishing a decision matrixNormalized decision matrix->Get matrix->Whereinr ij The formula is as follows:
；
pair matrixRWeighting to obtain weighted normalized matrix：
；
In the method, in the process of the invention,w j is the firstjThe weight value of the individual attribute(s),v ij is a weighted normalized matrixVMiddle (f)iLine 1jElements of a column;
then normalize the matrix according to the weightingVDetermining positive and negative ideal solutionsAndthe following formula is shown:
；
；
in the method, in the process of the invention,i=1,2,…，m，V j + to get an ideal understandingV + Is the first of (2)jThe number of attributes that can be used in the method,V j - is a negative ideal solutionV - Is the first of (2)jA plurality of attributes; the larger the benefit index is the attribute value, the better the target architecture is; the smaller the cost index refers to the attribute value, the better the target architecture;
calculating the distance from each scheme to positive and negative ideal solutionsAnd->Each scheme represents each generated command control network architecture; the following formula is shown:
；
；
the relative closeness is calculated and aligned according to the following:
；
in the method, in the process of the invention,C i is the relative closeness of the scheme, is the distance degree of the relative negative ideal solution, and the larger the value is, the better the scheme is, and the more optimal the scheme isC i A target evaluation value of the second layer index weight calculated by the TOPSIS method;
in step S6, the method for calculating the target evaluation value of the second layer index weight by using the gray correlation method is as follows:
the essence of gray association is to judge the association degree between the comparison number sequence to be processed and the reference number sequence according to the geometric similarity between the two sequences, and establish a normalized decision matrixXThe following formula is shown:
；
wherein the method comprises the steps ofx mn Representing decision matrixXFirst, themLine 1nColumn elements, determining a reference number column characterizing a data featureComparison series of values with attributed factors +.>Wherein->The method comprises the steps of carrying out a first treatment on the surface of the Ash correlation coefficient->The following formula is shown:
；
；
in the middle ofFor the resolution coefficient, the smaller the value, the stronger the resolution capability, and the value is takenρ=0.5; taking out grey related coefficient->As a degree of association with a reference seriesr i And ordered as shown in the following formula:
；
in the middle ofr i The approach to 1 represents a closer correlation of the comparison series to the reference series. Finally according to the association degreer i The magnitude of the values are ordered; degree of associationr i And a target evaluation value of the second-layer index weight calculated as a gray correlation method.
Step S7: and determining a target comprehensive evaluation value by combining target evaluation values of the second-layer index weights calculated by the TOPSIS method and the gray correlation method, sequencing each constructed command control network architecture, and completing screening of the optimal command control network architecture by taking the command control network architecture with the highest sequencing as the optimal command control network architecture.
The specific method of step S7 is as follows:
the correlation degree is as followsr i The calculation is improved:
；
；
in the middle ofAnd->Is positive and negative gray correlation coefficient->And->Respectively the improved positive and negative gray correlation degrees,W(P k ) Is the firstkIndividual indexP k Is a second layer index weight of (2);
comprehensively calculating the change in the index multidimensional space distance and shape by using a linear weighting method:
；
in the method, in the process of the invention,representing the decision maker's preference degree for the distance and shape change between the indexes, < >>Representing the selected firstiThe distance of the value of the individual scheme from the optimal scheme, < >>Representing the selected firstiThe value of each scheme is the distance from the worst scheme;
calculating relative closenessAdThe formula is as follows:
；
in relative proximity toAdAs a target comprehensive evaluation value, according to the relative closeness of each command control network architectureAdThe command control network architecture is ordered from a big order to a small order.
The following is one embodiment of a battle mission for a battlefield environment based on a dual variant weight and TOPSIS-gray correlation architecture optimization method designed by the present invention:
taking a army synthesized travel command control network system as an example, the number of network nodes is set asN=46, wherein the sensing unitOInformation center unit =10P=7, command control unitC=7, fire striking unitF=12, soft-killer cellS=10；
The conventional basic network distribution is shown in fig. 5, and the command control network architecture is constructed on the basis of the conventional basic network, as shown in fig. 6. Fig. 7 depicts a variable weight distribution scenario.
According to a specific combat task, 4 command control network architectures are randomly generatedAs a preferable simulation case of the command control network architecture, the indexes of the simulation case are shown in table 3:
constructing a judgment matrix A, and obtaining subjective weight after consistency check:
；
；
calculating state index to obtain objective weight；
Synthesizing subjective and objective weights to obtain first layer weightsw pre ：
；
The index of the finger control network architecture is analyzed through the variable weight thought, and the variable weight is obtained by combining the first layer weightW i (P j ) The target individual index weights are as follows, as shown in fig. 7.
；
Performing architecture sorting by using a TOPSIS method and a gray correlation method:
(1) TOPSIS method evaluation
When the TOPSIS method is used for evaluation, target state indexes are quantized to obtain a normalized matrixWeighting matrixVThe results were as follows:
；
；
from the following componentsVCan obtain a positive ideal solutionV + And negative ideal solutionV - ：
；
；
Re-calculating the distance of each target to the ideal solutionAnd->The following are provided:
；
；
(2) Ash correlation evaluation
Evaluating by gray correlation method according to normalized matrixCalculating to obtain positive and negative reference number seriesr + Andr - ：
；
；
then calculate positive and negative correlation coefficient matrixAnd->The following are provided:
；
；
finally, positive and negative gray correlation degrees are respectively obtained as follows:
；
；
FIG. 8 depictsμEach target command control network architecture ordering result when=0.
If it isμThe comprehensive ordering method is degenerated to be a gray correlation ordering method when the number is=0, namely, the optimal command control network architecture is judged through deformation;μwhen the method is=1, the method is degenerated into a TOPSIS method, and the optimal command control network architecture is determined through the distance change;it is a comprehensive TOPSIS and ash-related ranking method. Now pairμThe values of the three conditions are specifically analyzed:
when (when)μWhen=0, the relative closeness of each target to the positive and negative ideal targets is as follows in table 4:
o 1 >o 2 >o 4 >o 3 the generating architecture ordering is: architecture 1>Architecture 2>Architecture 4>Architecture 3, representing the best of the 4 generated command control network architectures, network architecture 1 is the worst compared to network architecture 3.
FIG. 9 depictsμEach target command control network architecture ordering result when=1. When (when)μWhen=1, the relative closeness of each target to the positive and negative ideal targets is shown in table 5:
o 1 >o 2 >o 4 >o 3 the generating command control network architecture ordering is as follows: architecture 1>Architecture 2>Architecture 4>Architecture 3, representing the best of the 4 finger network architectures generated, network architecture 1 is the worst compared to network architecture 3.
FIG. 10 depictsμEach target directs control network architecture ordering results when=0.5.μWhen=0.5, the relative closeness of each target to the positive and negative ideal targets is shown in table 6:
o 1 >o 2 >o 4 >o 3 the generating command control network architecture ordering is as follows: architecture 1>Architecture 2>Architecture 4>Architecture 3, representing the best of the 4 finger network architectures generated, network architecture 1 is the worst compared to network architecture 3.
The four command control network architectures generated at random are comprehensively available, the architecture optimization sorting is carried out by utilizing a method based on a double weight change theory and TOPSIS-gray association according to index values in each command control network architecture, and experimental results show that the network architecture 1 is optimal in the generated network architecture and can be most suitable for the formulated combat task requirements. To further verify the effectiveness of the method of the present invention, a dispersion maximization method, a principal component analysis method, and an improved projection algorithm were selected for comparison, the results are shown in table 7:
as can be seen from table 7, the method of the present invention is identical to the improved projection algorithm sequencing result, proving the preferred effectiveness of the proposed method for commanding the control network architecture. Slightly different from the dispersion maximization method and the principal component analysis method,but is provided witho 4 Ando 3 the ordering result is unchanged. In addition, when the dispersion maximization method weights the indexes in the index system, the membership between the indexes cannot be fully considered, so that the compatibility of the algorithm is poor when the algorithm is evaluated, and a decision maker is not facilitated to select a correct weighting method; the principal component analysis method is greatly restricted by subjective factors, is suitable for the situation that the linear correlation among indexes is very strong, and cannot well reflect the contribution of each single method in combination evaluation in the nonlinear problem; the improved normal standardization of the projection method is easy to cause information loss, the ideal projection value similar scheme cannot be effectively distinguished, the calculated amount is large, and the problems of no solution, singularity and the like are easy to exist, so that the final result is inaccurate. Compared with the nonlinear dynamic relation existing between indexes, the method provided by the invention combines the TOPSIS and ash correlation methods to give the final sorting result, combines subjectively and objectively, and utilizes double weights to enable the result to be more practical, so that the method is more scientific and reasonable.
In summary, a network architecture optimization method based on a double-varying-weight theory and TOPSIS-gray association is provided for scientifically and objectively constructing command control network architectures in modern war, a series of command control network architectures are generated according to a specific combat mission, and the generated architectures are ordered according to the provided method, so that the optimal command control network architecture is screened out. The method can effectively solve the problems that the prior preferred method cannot balance subjective preference of commanders and objective uncertainty factors of indexes, is not easy to cause information loss, and has small calculated amount. Simulation results show that the method can more accurately reflect the influence of subjective and objective factors of indexes and state value changes on weight changes, and meanwhile comprehensively considers the distance and shape changes among the indexes in the sequencing results, so that the optimal command control network architecture can be screened out more scientifically and reasonably, and the method plays a key role in the win or loss of future war.
The embodiments of the present invention have been described in detail with reference to the drawings, but the present invention is not limited to the above embodiments, and various changes can be made within the knowledge of those skilled in the art without departing from the spirit of the present invention.
Claims (7)
1. The architecture optimization method based on double weight change and TOPSIS-gray association is characterized by comprising the following steps of S1-S7, and screening of an optimal command control network architecture is completed aiming at the combat mission of the battlefield environment:
step S1: constructing each command control network architecture according to a battlefield environment, wherein each command control network architecture is divided into three command control layers, namely an information network, a command control network and a fire striking network;
step S2: according to a complex network theory, abstracting information exchange relation among command control layers, communication transmission paths, interaction relation among command mechanisms among the command control layers and membership relation among the command mechanisms of a command control network architecture, and describing by adopting an undirected communication graph to establish a command control network architecture model;
step S3: constructing a command network architecture index system, wherein the command network architecture index system comprises a plurality of indexes for evaluating the performance of the established command control network architecture model, and optimizing the command control network architecture according to the indexes;
the indexes contained in the index system of the finger-controlled network architecture in the step S3 are combat capability, average path length, aggregation coefficient and network density respectively;
the combat capability is calculated based on information entropy, belongs to benefit type indexes, and is calculated according to the following formula:
；
in the method, in the process of the invention,p i for co-operating combat nodesv i Is a function of the information source probability error of (a),H i for co-operating combat nodesv i Is an information entropy of (a);
the information entropy of the single command control network architecture is as follows:
；
in the method, in the process of the invention,qfor co-operating combat nodesv i Is used in the number of (a) and (b),H j for a single nodev j Is used for the information entropy of (a),bas a number of individual nodes,H l controlling network architecture for single conductorlIs an information entropy of (a);
the combat ability is as follows:
；
in the method, in the process of the invention,Cthe ability to perform a battle is indicated,C l controlling network architecture for single conductorlIs a combat capability of (1);
the average path length belongs to a cost indicator, as follows:
；
in the method, in the process of the invention,Lthe average path length is indicated as such,d ij representing nodes in a command control network architecturev i Andv j the shortest distance between;Nrepresenting the total number of nodes of the command control network architecture;
the aggregation factor belongs to the benefit index, and the following formula is shown:
；
in the method, in the process of the invention,representing nodesv i Is used for the aggregation coefficient of (a),w ij representing nodesv i Sum nodev j The edge weight of the middle part is calculated,w jk representing nodesv j Sum nodev k The edge weight of the middle part is calculated,w ki representing nodesv k Sum nodev i Edge weight between nodesv i Sum nodev k Is a nodev j Is a neighbor node of (a);
the network density belongs to the benefit index, and the following formula is shown:
；
in the method, in the process of the invention,NandMrespectively representing the total number of nodes of the command control network architecture and the total number of actual connecting edges,D i representing nodesv i Network density of (2);
step S4: weighting each index in the index system of the index control network architecture respectively, wherein the weights comprise subjective weights and objective weights, subjective weights are assigned to each index by using a hierarchical analysis method, objective weights are assigned to each index by using an improved entropy weight method, and the subjective weights and the objective weights of each index are combined according to subjective and objective factors to obtain a first-layer index weight;
step S5: carrying out variable weight processing on the first layer of index weights by utilizing a variable weight theory to obtain variable weights which change according to the change of the target state values, and obtaining second layer of index weights according to the variable weights;
step S6: calculating a target evaluation value of the index weight of the second layer by using a TOPSIS method and a gray correlation method respectively;
step S7: and determining a target comprehensive evaluation value by combining target evaluation values of the second-layer index weights calculated by the TOPSIS method and the gray correlation method, sequencing each constructed command control network architecture, and completing screening of the optimal command control network architecture by taking the command control network architecture with the highest sequencing as the optimal command control network architecture.
2. The architecture optimization method based on double variation weights and TOPSIS-ash correlation according to claim 1, wherein the command mechanism in the intelligence information network in step S1 comprises a sensing unit and an intelligence center unit, the command mechanism in the command control network comprises a command control unit, and the command mechanism in the fire striking network comprises a fire striking unit and a soft killing unit.
3. The architecture optimization method based on double-varying-weight and TOPSIS-gray correlation according to claim 1, characterized by that step S2 builds node sets in each finger hierarchyWherein node->An abstract representation of each command mechanism in each command level, wherein n is the total number of nodes in the command level; constructing a set of edges in each finger hierarchy +.>Wherein the edge->Representing the connected edges among the nodes, wherein m is the total number of edges in the command hierarchy; and constructing an undirected connected graph G= (V, E) according to the node set V and the edge set E.
4. The architecture optimization method based on double variation weights and TOPSIS-gray correlation according to claim 3, wherein the subjective weight determination method of each index in step S4 is as follows:
assigning values to the important relations between every two indexes respectively, and constructing a judgment matrixIn a matrixa ij Representation matrixAMiddle (f)iIndividual elements and the firstjComparing the importance of the individual elements with the result value;
determination matrixAMaximum feature root of (2)λ max Corresponding feature vectorwThe following formula is shown:
；
for characteristic vectorwAnd carrying out consistency detection, wherein the consistency detection is shown in the following formula:
；
；
in the method, in the process of the invention,CIto judge matrixAIs used for the general consistency index of (c),RIas an index of the average random consistency,CRif the consistency is detected as an indexCR<0.1, then judge the judgment matrixAIf the consistency requirement is met, otherwise, returning to adjust the judgment matrixAUp to the judgment matrixAMeets the consistency requirement, and the feature vectorwSubjective weight as index after normalizationw s ；
The objective weight determining method of each preset index in the step S4 is as follows:
is provided withx ij Is the firstiFirst command control network architecturejThe value of the individual index(s),i=1,2,…，m，j=1,2,…，n；min order to command the number of network architectures to be controlled,nis the index number; normalizing each index:
；
in the method, in the process of the invention,y ij is thatx ij Normalized values;
calculate the firstiFirst command control network architecturejInformation entropy of individual indexz ij The formula is as follows:
；
；
in the method, in the process of the invention,E j represent the firstjInformation entropy of individual index, wheny ij When=0, use 10 -4 Correctiony ij ；
According to the firstjInformation entropy of individual indexE j The objective weight is calculated as follows:
；
in the method, in the process of the invention,w ae (j) Represent the firstjObjective weights of the individual indicators;
combining subjective weightsw s Objective weightw ae Obtaining a first layer of index weightw pre The formula is as follows:
；
in the method, in the process of the invention,βrepresenting the ratio coefficient of subjective weight to first layer index weight, 1-βAnd the scale coefficient of the objective weight to the index weight of the first layer is represented.
5. The architecture optimization method based on dual variable weights and TOPSIS-gray correlation according to claim 4, wherein the method for performing variable weight processing on the first layer index weight by using the variable weight theory in step S5, and obtaining the second layer index weight is as follows:
the equalization sub-function is constructed as follows:
；
；
；
in the method, in the process of the invention,F(P j )、F(P 1 ，…，P 4 ) In order to equalize the sub-functions,P j represent the firstjThe number of the indexes is equal to the number of the indexes,j=1 indicates the ability to combat,j=2 represents the average path length,j=3 represents an aggregation coefficient,j=4 denotes network density;αis an adjustment factor;
according to the weight-changing theory, deriving the equalization subfunction to obtain a state weight-changing vectorS(P 1 ，…，P 4 ) The following formula:
；
since the final weight sum is 1,S(P 1 ) When the index weight is smaller, the other index weights are correspondingly increased, and the index weight is shown as the following formula:
；
in the method, in the process of the invention,is the firstjThe first layer of index weights of the individual indices,W(P j ) Is the firstjIndividual indexP j Is a second level of index weight.
6. The architecture optimization method based on double variation weights and TOPSIS-gray correlation according to claim 5, wherein the target evaluation value method for calculating the second layer index weight by using TOPSIS method in step S6 is as follows:
for containingmPersonal command control network architecturenDecision-making problems to be ordered of individual indexes, and establishing a decision matrix，a ij Representation matrixAMiddle (f)iLine 1jElements of a column; normalized decision matrix->Obtaining a matrixWhereinr ij The formula is as follows:
；
pair matrixRWeighting to obtain weighted normalized matrix：
；
In the method, in the process of the invention,w j is the firstjThe weight value of the individual attribute(s),v ij is a weighted normalized matrixVMiddle (f)iLine 1jElements of a column;
then normalize the matrix according to the weightingVDetermining positive and negative ideal solutionsAndthe following formula is shown:
；
；
in the method, in the process of the invention,i=1，2，…，m，V j + to get an ideal understandingV + Is the first of (2)jThe number of attributes that can be used in the method,V j - is a negative ideal solutionV - Is the first of (2)jA plurality of attributes;
calculating the distance from each scheme to positive and negative ideal solutionsAnd->The following formula is shown:
；
；
the relative closeness is calculated and aligned according to the following:
；
in the method, in the process of the invention,C i the relative closeness of the scheme is used as a target evaluation value of the index weight of the second layer calculated by the TOPSIS method;
in step S6, the method for calculating the target evaluation value of the second layer index weight by using the gray correlation method is as follows:
establishing a normalized decision matrixXThe following formula is shown:
；
wherein the method comprises the steps ofx mn Representing decision matrixXFirst, themLine 1nColumn elements, determining a reference number column characterizing a data featureComparison series of values with attributed factors +.>Wherein->The method comprises the steps of carrying out a first treatment on the surface of the Ash correlation coefficient->The following formula is shown:
；
；
in the middle ofTaking the gray correlation coefficient as the resolution coefficient>As a degree of association with a reference seriesr i And ordered as shown in the following formula:
；
degree of associationr i And a target evaluation value of the second-layer index weight calculated as a gray correlation method.
7. The architecture optimization method based on double weight change and TOPSIS-gray correlation according to claim 6, wherein the specific method of step S7 is as follows:
the correlation degree is as followsr i The calculation is improved:
；
；
in the middle ofAnd->Is positive and negative gray correlation coefficient->And->Respectively the improved positive and negative gray correlation degrees,W(P k ) Is the firstkIndividual indexP k Is a second layer index weight of (2);
comprehensively calculating the change in the index multidimensional space distance and shape by using a linear weighting method:
；
in the method, in the process of the invention,representing the decision maker's preference degree for the distance and shape change between the indexes, < >>Representing the selected firstiValues and the most of the schemesDistance of the best solution,/->Representing the selected firstiThe value of each scheme is the distance from the worst scheme;
calculating relative closenessAdThe formula is as follows:
；
in relative proximity toAdAs a target comprehensive evaluation value, according to the relative closeness of each command control network architectureAdThe command control network architecture is ordered from a big order to a small order.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311669373.3A CN117436282B (en) | 2023-12-07 | 2023-12-07 | Architecture optimization method based on dual variable weight and TOPSIS-gray correlation |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311669373.3A CN117436282B (en) | 2023-12-07 | 2023-12-07 | Architecture optimization method based on dual variable weight and TOPSIS-gray correlation |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117436282A CN117436282A (en) | 2024-01-23 |
| CN117436282B true CN117436282B (en) | 2024-04-02 |
Family
ID=89553602
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311669373.3A Active CN117436282B (en) | 2023-12-07 | 2023-12-07 | Architecture optimization method based on dual variable weight and TOPSIS-gray correlation |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117436282B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008102624A (en) * | 2006-10-17 | 2008-05-01 | Toshiba Corp | Evaluation method of command control system, and simulation system using the same |
| CN102542347A (en) * | 2011-12-28 | 2012-07-04 | 东南大学 | Method for comprehensively evaluating electric energy quality |
| CN106953754A (en) * | 2017-03-16 | 2017-07-14 | 大连大学 | Charge network survivability Measurement Method based on operation link entropy |
| CN114444203A (en) * | 2022-01-25 | 2022-05-06 | 中国人民解放军空军工程大学 | Battlefield situation variable-weight-based aerial cluster threat assessment method |
| CN114912783A (en) * | 2022-03-22 | 2022-08-16 | 中国人民解放军战略支援部队航天工程大学 | Equipment combat effectiveness dynamic evaluation method based on virtual ideal solution and grey correlation degree |
| CN114971230A (en) * | 2022-05-10 | 2022-08-30 | 华能国际电力股份有限公司上海石洞口第二电厂 | Coal blending combustion effect evaluation method based on combined weighted-improved TOPSIS method |
| CN116739428A (en) * | 2023-06-20 | 2023-09-12 | 中国人民解放军陆军炮兵防空兵学院 | Method for analyzing target value of fire striking of subjective and objective weighted TOPSIS (top-down sequence of steps of analysis) soldier |
-
2023
- 2023-12-07 CN CN202311669373.3A patent/CN117436282B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008102624A (en) * | 2006-10-17 | 2008-05-01 | Toshiba Corp | Evaluation method of command control system, and simulation system using the same |
| CN102542347A (en) * | 2011-12-28 | 2012-07-04 | 东南大学 | Method for comprehensively evaluating electric energy quality |
| CN106953754A (en) * | 2017-03-16 | 2017-07-14 | 大连大学 | Charge network survivability Measurement Method based on operation link entropy |
| CN114444203A (en) * | 2022-01-25 | 2022-05-06 | 中国人民解放军空军工程大学 | Battlefield situation variable-weight-based aerial cluster threat assessment method |
| CN114912783A (en) * | 2022-03-22 | 2022-08-16 | 中国人民解放军战略支援部队航天工程大学 | Equipment combat effectiveness dynamic evaluation method based on virtual ideal solution and grey correlation degree |
| CN114971230A (en) * | 2022-05-10 | 2022-08-30 | 华能国际电力股份有限公司上海石洞口第二电厂 | Coal blending combustion effect evaluation method based on combined weighted-improved TOPSIS method |
| CN116739428A (en) * | 2023-06-20 | 2023-09-12 | 中国人民解放军陆军炮兵防空兵学院 | Method for analyzing target value of fire striking of subjective and objective weighted TOPSIS (top-down sequence of steps of analysis) soldier |
Non-Patent Citations (7)
| Title |
|---|
| RESEARCH ON BALLISTIC MISSILE THREAT ASSESSMENT BASED ON MAXIMUM DEVIATION;Chengsheng Pan等;ICIC International;20231031;第9卷(第10期);4131-4138 * |
| 基于复杂网络的防空反导体系抗毁性评估;李腾达等;军事运筹与评估;20231028;第38卷(第5期);34-39 * |
| 基于指标双层变权和TOPSIS-灰关联的多目标威胁评估方法;王建伟等;电光与控制;20210409;第28卷(第6期);1-6 * |
| 基于改进遗传算法的装备保障效能评估指标约简;潘成胜等;计算机仿真;20210715;第38卷(第7期);223-229 * |
| 基于灰色关联-组合赋权的建筑工程项目评标模型研究;吴夏萍;张建坤;;项目管理技术;20180210(第02期);47-53 * |
| 基于级差最大化组合赋权TOPSIS灰关联投影法的区域碳效率动态评价;余鹏;马珩;周福礼;;运筹与管理;20191225(第12期);174-181 * |
| 战场指挥控制系统的网络控制特性及其评估;张练达, 卢昱;装备指挥技术学院学报;20051028(第05期);108-111 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117436282A (en) | 2024-01-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111866954B (en) | User selection and resource allocation method based on federal learning | |
| Li et al. | A systematic approach to heterogeneous multiattribute group decision making | |
| CN110490422B (en) | Target combat effectiveness situation assessment method based on game cloud model | |
| CN111313957B (en) | Hybrid satellite communication system resource allocation method based on classification multi-objective optimization | |
| CN114444203B (en) | Air cluster threat assessment method based on battlefield situation variable weight | |
| CN111163487B (en) | Communication waveform comprehensive transmission performance evaluation method and system | |
| CN112258062A (en) | System contribution degree determination method and device | |
| CN111832817A (en) | Small world echo state network time sequence prediction method based on MCP penalty function | |
| CN114444389A (en) | Air attack target dynamic threat assessment method based on combined empowerment and improved VIKOR | |
| CN113177583B (en) | Aerial target clustering method | |
| CN117436282B (en) | Architecture optimization method based on dual variable weight and TOPSIS-gray correlation | |
| CN113919507A (en) | Federal learning method based on DAG block chain | |
| CN116739428A (en) | Method for analyzing target value of fire striking of subjective and objective weighted TOPSIS (top-down sequence of steps of analysis) soldier | |
| CN116934148A (en) | Air attack target threat assessment method based on clustering combination weighting | |
| CN116846592A (en) | Intelligent decision-making system and method based on attack and defense game model | |
| Liu et al. | A new method to air target threat evaluation based on Dempster-Shafer evidence theory | |
| CN113919425B (en) | Autonomous aerial target allocation method and system | |
| CN112612300B (en) | Multi-platform intelligent decision-making multi-target game method and device | |
| CN112612207B (en) | Multi-target game solving method and system under uncertain environment | |
| CN113887888A (en) | Adaptive scoring and associated recommendation method and system for combined hit target | |
| CN111654855B (en) | Trust updating method in underwater wireless sensor network based on AHP | |
| Han et al. | Operational effectiveness evaluation of space-based information system based on improved ahp-bpnn model | |
| Longyue et al. | Optimal index shooting policy for layered missile defense system | |
| Huang et al. | Malicious Models-based Federated Learning in Fog Computing Networks | |
| CN106910148B (en) | Collaborative filtering-based adaptive pushing method for command elements |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |