A method and system for constructing a collaborative network model of a multi-domain control system

Abstract

The present invention discloses a method and system for constructing a collaborative network model of a multi-domain control system, which relates to the technical field of data processing, and includes the following steps: S1: Cluster existing control elements into multiple domains to obtain a domain set #imgabs0#, and construct an initial network model with the control elements as nodes and the relationships between the nodes as edges; The method and system for constructing the collaborative network model of the multi-domain control system can improve the efficiency of establishing the collaborative network model of the control system, and the established collaborative network model of the control system has a strong correlation, thereby improving the efficiency of control decision-making, fully considering the contribution degree of the selected nodes in HL, further improving the efficiency of collaborative control, and through reasonable configuration of HL, it will not cause waste of HL, saves costs, sends the sent information through multiple carriers, and further improves the confidentiality.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and specifically to a method and system for constructing a multi-domain control system collaborative network model. Background Art

[0002] The component systems of the composition system of the multi-domain control system collaborative network model are large in number, and the ways of mutual connection and interaction are complex. There are situations of mutual cooperation and competition under limited resources. When problems occur in the cooperation between component systems, under the pressure of competition, the ways of interaction between component systems may change, thereby causing changes in the system structure and further causing changes in the system behavior. On the other hand, the future environment is uncertain. When the environment changes, it is necessary for the system to adapt to the new changes, so it is required that the system make corresponding structural and behavioral changes. Therefore, a method and system for constructing a multi-domain control system collaborative network model are proposed. Summary of the Invention

[0003] Aiming at the deficiencies of the prior art, the present invention provides a method and system for constructing a multi-domain control system collaborative network model, which solves the problems raised in the above background art.

[0004] To achieve the above objectives, the present invention is realized through the following technical solutions: A method for constructing a multi-domain control system collaborative network model includes the following steps:

[0005] S1: Cluster the existing control elements and divide them into multiple domains to obtain a domain set , and construct an initial network model with the control elements as nodes and the relationships between the nodes as edges;

[0006] S2: Calculate the repulsion of the newly discovered control element with each domain in the existing domain set . If it is repulsive to each domain in the domain set , execute ;

[0007] If there are non-repulsive domains, the set of non-repulsive domains is denoted as , then calculate the similarity of the newly discovered control element v i with each non-repulsive domain in the set of non-repulsive domains ;

[0008] S3: If is less than the threshold, execute ; if is greater than or equal to the threshold, then place the newly discovered control element in In the largest domain, calculate the newly discovered control elements and the topological correlation degree with the existing control elements in the domain, and connect them with edges;

[0009] S4: According to the newly discovered control elements establish a new domain and add the domain to the domain set ;

[0010] S5: Repeat S2, S3 and S4 until all newly discovered control elements are clustered to form a collaborative network model of the control system.

[0011] Optionally, the control elements include control monitoring nodes, decision nodes, HL selected nodes and target nodes;

[0012] The control monitoring nodes obtain the information of the target nodes in real time and transmit it to the decision nodes through the encrypted communication network;

[0013] The decision nodes include an information fusion module and an intelligent decision-making module;

[0014] The information fusion module summarizes the information provided by multiple control monitoring nodes and generates control objectives based on the summarized information;

[0015] The intelligent decision-making module includes a data selection module, a graph attention mechanism model layer and a BP neural network;

[0016] The data selection module selects entities from the collaborative network model of the control system according to the control objective at time, one of the entities includes a target node and HL selected nodes, the HL selected nodes form an HL node set ; The feature vectors of the target of the HL selected nodes at

[0017] time are respectively input into each graph attention mechanism model in the graph attention mechanism model layer; The graph attention mechanism model layer includes graph attention mechanism models. The th graph attention mechanism model outputs a completion vector for the target at time, so that the output vector of the graph attention mechanism model layer is ; The BP neural network generates based on the input vector At a moment control index items

[0018] Optionally, according to the above, it can be obtained that:

[0019] (1)

[0020] In formula (1), is the activation function,

[0021]

[0022] is the nth HL selected node in the HL selected node set and is the contribution degree of the feature vector of the selected target of the nth HL selected node to the feature vector of the selected target of the nth HL selected node ; is the activation function; is the parameter from the input layer to the hidden layer of the graph attention mechanism model; represents the parameter matrix; represents concatenating and and together.

[0023] Optionally, in step S4: establishing a new domain according to the newly discovered control element and adding the domain to the domain set The selection of the HL selected node set in and the selection of the m HL selected nodes in it include the following steps: S41:

[0024] ; ;

[0025] S42: Calculate the achievement ability value of the HL selected node set for the target according to formula (2): (2)

[0026] In the formula, is the achievement ability of the HL selected node in the HL selected node set for the target ;

[0027] S43: Judge that if , it is considered that the HL node set is not sufficient to achieve the target , , and then return to S42; if , it is considered that the target is selected of the HL node set is sufficient to achieve the target , and then S44 is executed, where is the target completion value of;

[0028] S44: Output the HL selected node set and the target corresponding to the selected set .

[0029] Optionally, the control monitoring node, the decision node, and the HL selected node all include a communication device, and the communication device includes a wireless transmitter and a wireless receiver;

[0030] The wireless transmitter includes: a first measurement unit, a transmission controller, and a transmission unit;

[0031] The first measurement unit measures the electromagnetic wave environment between the wireless transmitter of the first node and the wireless receiver of the second node that communicate with each other according to the wireless resources;

[0032] The transmission controller determines the wireless resources and parameters used in the transmission of the bit information sequence based on the measurement results of the first measurement unit;

[0033] The transmission unit selects a frequency combination according to the wireless resources and generates a bit group symbol according to the frequency combination;

[0034] The wireless receiver includes: a second measurement unit, a reception controller, a reception unit, and a decoding unit;

[0035] The second measurement unit measures the electromagnetic wave environment between the wireless transmitter of the first node and the wireless receiver of the second node that communicate with each other according to the wireless resources;

[0036] The reception controller estimates the frequency combination based on the measurement results of the second measurement unit;

[0037] The reception unit receives the bit group symbol;

[0038] The decoding unit decodes the received bit group symbol based on the frequency combination.

[0039] Optionally, the transmission unit includes a source input module, a bit grouping module, a mapper, and a modulator;

[0040] The source input module is used to receive the bit information sequence to be modulated after source coding;

[0041] The bit grouping module divides the bit information sequence into groups of K bits and provides it to the symbol mapper, The bit sequence has combinations;

[0042] The mapper maps one combination into a bit group symbol according to the agreed confidentiality rules;

[0043] The modulator selects carrier frequencies from the wireless resources measured by the first measurement unit, and modulates 4 adjacent symbols to four phases of one of the carrier frequencies of

[0044] Optionally, the decoding unit includes a plurality of receiving modules, a multiplex demodulation module and a calculation module;

[0045] Each of the receiving modules converts the signal received by the receiving antenna into electrical data;

[0046] The multiplex demodulation module includes a multiplication module, an integration module and a modulo module;

[0047] The multiplication module is used to multiply each electrical data by the complex conjugate value of the bit group symbol and send it to the integration module;

[0048] The integration module integrates the received signal within one symbol length in the time domain, and then sends it to the modulo module to take its absolute value to obtain a received value, and the calculation module obtains the bit confidence probability according to the received value.

[0049] Optionally, the calculation module includes a probabilistic neural network, the probabilistic neural network includes an input layer, a hidden layer and an output layer, the input layer inputs the received values of the multiplex demodulation module, and the output layer outputs the bit confidence probability.

[0050] A multi-domain control system collaborative network model construction system includes a storage medium and several processors, the storage medium stores computer program codes, and the processors call the computer program codes in the storage medium to implement the method described above.

[0051] The present invention provides a multi-domain control system collaborative network model construction method and system, having the following beneficial effects:

[0052] The multi-domain control system collaborative network model construction method and system can improve the efficiency of establishing a control system collaborative network model, and the established control system collaborative network model has a strong correlation, thereby improving the efficiency of control decision-making, fully considering the contribution degree of the nodes selected by HL, further improving the efficiency of collaborative control, and through reasonable configuration of HL, it will not cause waste of HL, saving costs, sending the information through multiple carriers, and thus improving the confidentiality. Description of the Drawings

[0053] Figure 1 It is a flowchart of the method for constructing the collaborative network model of the control system of the present invention;

[0054] Figure 2 It is a block diagram of the composition of the intelligent decision-making module of the decision-making node of the present invention;

[0055] Figure 3 It is a flowchart of the HL configuration method of the present invention;

[0056] Figure 4 It is a block diagram of the composition of the wireless transmitter of each node of the present invention;

[0057] Figure 5 It is a block diagram of the composition of the wireless receiver of each node of the present invention. Specific implementation mode

[0058] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0059] Embodiment 1

[0060] Please refer to Figures 1 to 5 , the present invention provides a technical solution: a method for constructing a collaborative network model of a multi-domain control system, including the following steps:

[0061] S1: Cluster the existing control elements and divide them into multiple domains to obtain a domain set , and construct an initial network model with the control elements as nodes and the relationships between the nodes as edges;

[0062] S2: Calculate the repulsion between the newly discovered control element and each domain in the existing domain set . If it is repulsive to each domain in the domain set , execute S4; if there are non-repulsive domains, the set of non-repulsive domains is denoted as , then calculate the similarity i between the newly discovered control element v and each non-repulsive domain in the set of non-repulsive domains;

[0063] S3: If is less than the threshold, execute ; if is greater than or equal to the threshold, then place the newly discovered control element in the largest domain, calculate the topological correlation degree between the newly discovered control element and the existing control elements in the domain, and connect them with edges;

[0064] S4: According to the newly discovered control elements Establish a new domain , and add the domain to the domain set ;

[0065] S5: Repeat S2, S3, and S4 until all the newly discovered control elements are clustered to complete the formation of the collaborative network model of the control system.

[0066] Through the technical solution proposed by the present invention, the efficiency of establishing the collaborative network model of the control system is improved, and the established collaborative network model of the control system has strong correlation, thus improving the efficiency of control decision-making.

[0067] Among them, the control elements include control monitoring nodes, decision nodes, HL selected nodes, and target nodes;

[0068] The control monitoring nodes obtain the information of the target nodes in real time and transmit it to the decision nodes through the encrypted communication network;

[0069] The decision nodes include an information fusion module and an intelligent decision-making module;

[0070] The information fusion module summarizes the information provided by multiple control monitoring nodes and generates control objectives based on the summarized information;

[0071] As Figure 2 shown, the intelligent decision-making module includes a data selection module, a graph attention mechanism model layer, and a BP neural network;

[0072] The data selection module selects entities from the collaborative network model of the control system according to the control objective at time, one of the entities is a target node and the HL selected nodes form the HL node set ; The feature vectors of the HL selected nodes at

[0073] time are respectively input into each graph attention mechanism model in the graph attention mechanism model layer; The graph attention mechanism model layer includes graph attention mechanism models, and the th graph attention mechanism model outputs the completion vector for the target at time, so that the output vector of the graph attention mechanism model layer is Input vector at a moment Generate At the moment of control index items;

[0074] From the above, we can get:

[0075] (1)

[0076] In the formula, is the activation function,

[0077]

[0078] is the in the set of HL selected nodes th HL selected node, and the feature vector of the selected target For the th HL selected node, the contribution degree of the feature vector of the selected target ; is the activation function; is the parameter from the input layer to the hidden layer of the graph attention mechanism model; represents the parameter matrix; means to and concatenate them.

[0079] The above proposed scheme fully considers the contribution degree among the HL selected nodes, thereby improving the control efficiency of collaborative control;

[0080] As Figure 3 shown, S4: Establish a new domain according to the newly discovered control element , and add the domain to the domain set . The selection of the in the set of HL selected nodes HL selected nodes includes the following steps:

[0081] S41: ;

[0082] S42: Calculate the achievement ability value of the set of HL selected nodes for the target according to formula (2): (2)

[0083] In the formula, is the achievement ability of the HL selected node in the set of HL selected nodes for the target ;

[0084] S43: Determine that if , then it is considered that the HL node set is insufficient to achieve the goal , , and then return to S42; if , then it is considered that the selected target 's HL node set is sufficient to achieve the goal , and then execute S44, where is the completion value of the target .

[0085] S44: Output the HL selected node set and the corresponding selected target of this set .

[0086] Through the proposed solution content, the HL is reasonably configured, thus not causing waste of the HL, and thereby saving costs.

[0087] Among them, the control monitoring node, the decision-making node, and the HL selected node all include a communication device, and the communication device all includes a wireless transmitter and a wireless receiver;

[0088] As Figure 4 shown, the wireless transmitter includes: a first measurement unit, a transmission controller, and a transmission unit;

[0089] The first measurement unit measures the electromagnetic wave environment between the wireless transmitter of the first node and the wireless receiver of the second node that communicate with each other according to wireless resources;

[0090] The transmission controller decides the wireless resources and parameters used in the transmission of the bit information sequence based on the measurement results of the first measurement unit;

[0091] The transmission unit selects a frequency combination according to the wireless resources and generates a bit group symbol according to the frequency combination.

[0092] The transmission unit includes a source input module, a bit grouping module, a mapper, and a modulator;

[0093] The source input module is used to receive the bit information sequence to be modulated after source coding;

[0094] The bit grouping module divides the bit information sequence into groups of K bits and provides it to the symbol mapper, The bit sequence has combinations;

[0095] The mapper maps one combination to one bit group symbol according to the agreed confidentiality rule;

[0096] A modulator selects from the radio resources measured by a first measurement unit carrier frequencies, and modulates four adjacent symbols onto four phases of one of the carrier frequencies of the

[0097] Figure 5 As shown in Figure 5 a radio receiver includes: a second measurement unit, a reception controller, a reception unit, and a decoding unit;

[0098] The second measurement unit measures the electromagnetic wave environment between the radio transmitter of a first node and the radio receiver of a second node that communicate with each other according to radio resources;

[0099] The reception controller estimates a frequency combination based on the measurement result of the second measurement unit,

[0100] The reception unit receives a bit group signal according to radio resources;

[0101] The decoding unit decodes the received bit group symbols based on the frequency combination.

[0102] Among them, the decoding unit includes: a plurality of reception modules, a multiplex demodulation module, and a calculation module,

[0103] Each reception module converts the signal received by the reception antenna into electrical data, and each demodulation module includes a multiplication module, an integration module, and a modulo module;

[0104] The multiplication module is used to multiply each electrical data by the complex conjugate value of the bit group symbol and send it to the integration module;

[0105] The integration module integrates the received signal within one symbol length in the time domain, and then sends it to the modulo module to take its absolute value to obtain a reception value; the calculation module obtains a bit confidence probability according to the reception value.

[0106] The calculation module includes a probabilistic neural network, and the probabilistic neural network includes an input layer, a hidden layer, and an output layer. The input layer inputs the reception values of the multiplex demodulation module; the output layer outputs the bit confidence probability.

[0107] This solution can send the transmitted information through multiple carriers, thereby improving the confidentiality.

[0108] Embodiment 2

[0109] The present invention also provides a multi-domain control system collaborative network model construction system, including a storage medium and several processors. The storage medium stores computer program codes, and the processors call the computer program codes in the storage medium to implement the above-mentioned method.

[0110] The above are only the preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, making equivalent substitutions or changes should be covered within the protection scope of the present invention.

Claims

1. A method for constructing a collaborative network model of a multi-domain control system, characterized in that: It includes the following steps: S1: Cluster the existing control elements into multiple domains to obtain the domain set A r = [A r1 , …, A ri , …, A rI , and construct an initial network model with the control elements as nodes and the relationships between the nodes as edges; S2: Calculate the newly discovered control element v new with the existing domain set A r for each domain A ri in it. If it is mutually exclusive with each domain A r in the domain set A ri execute S4; If there are non - exclusive domains, the set of non - exclusive domains is denoted as A r =[A r1 ,…,A rk ,…,A rK , then calculate the similarity Sim(v i with each non - exclusive domain in the set of non - exclusive domains, Sim(v new ,A rk ); S3: If Sim(v new , A rk ) is less than the threshold, execute S4; if Sim(v new , A rk ) is greater than or equal to the threshold, then place the newly discovered control element S new in the domain with the largest Sim(v new , A rk ), calculate the topological correlation degree between the newly discovered control element v new and the existing control elements in the domain, and connect them with edges; S4: Establish a new domain A according to the newly discovered control factor v new and add domain A rnew to the domain set A rnew ; r ​ S5: Repeat S2, S3, and S4 until all newly discovered control elements v new The clustering is completed to form a collaborative network model of the control system; The control elements include a control monitoring node, a decision-making node, an HL selection node, and a target node; The control monitoring node obtains the information of the target node in real time and transmits it to the decision-making node through an encrypted communication network; The decision-making node includes an information fusion module and an intelligent decision-making module; The information fusion module summarizes the information provided by multiple control monitoring nodes, and generates a control target based on the summarized information; The intelligent decision-making module includes a data selection module, a graph attention mechanism model layer, and a BP neural network; At time t, the data selection module selects N + 1 entities from the control system collaborative network model according to the control objective. Among the N + 1 entities, there is a target node and N HL selected nodes, and the N HL selected nodes form the HL node set V N ; Input the feature vectors of the targets of the N HL selected nodes at time t into each graph attention mechanism model in the graph attention mechanism model layer respectively; The graph attention mechanism model layer includes N graph attention mechanism models, and the nth graph attention mechanism model outputs a completion vector for the target at time t Thus, the output vector of the graph attention mechanism model layer is The BP neural network generates Q control index terms at time t based on the input vector at time t ​ According to the above, we can get: In formula (1), σ is the activation function, is the j-th HL selected node in the HL selected node set V N and the feature vector h of the selected target j is the contribution degree of the feature vector h of the selected target to the n-th HL selected node; LeakyReLU is the activation function; ρ is the parameter from the input layer to the hidden layer of the graph attention mechanism model; W represents the parameter matrix; || means concatenating Wh n and Wh n together. j ​ 2. The method for constructing a collaborative network model of a multi-domain control system according to claim 1, wherein: S4: According to the newly discovered control element v new Establish a new domain A rnew , and add domain A rnew to the domain set A r The selection of N HL selected nodes in the HL selected node set V in N includes the following steps: S41: N ← 1; S42: Calculate the HL selected node set V according to formula (2) N For the target Achievement ability value of: (2) where A n is the HL selected node set V N in the HL selected nodes the ability to achieve the target ; S43: Judgment, if then it is considered that the HL node set V N is insufficient to achieve the goal make N ← N + 1, and then return to S42; if then it is considered that the target is selected of the HL node set V N is sufficient to achieve the goal then execute S44, where L d is the completion value of the target ; S44: Output the HL selected node set V N and the targets corresponding to the selected set 3. A method for constructing a collaborative network model of a multi-domain control system according to claim 1, characterized in that: The control monitoring node, the decision-making node, and the HL selection node all include communication devices, and the communication devices include wireless transmitters and wireless receivers; The wireless transmitter includes: a first measurement unit, a transmission controller, and a transmission unit; The first measurement unit measures the electromagnetic wave environment between the wireless transmitter of the first node and the wireless receiver of the second node that communicate with each other according to wireless resources; The transmission controller determines the wireless resources and parameters used in the transmission of the bit information sequence based on the measurement results of the first measurement unit; The transmission unit selects a frequency combination according to the wireless resources, and generates a bit group symbol according to the frequency combination; The wireless receiver includes: a second measurement unit, a reception controller, a reception unit, and a decoding unit; The second measurement unit measures the electromagnetic wave environment between the wireless transmitter of the first node and the wireless receiver of the second node that communicate with each other according to wireless resources; The reception controller estimates the frequency combination based on the measurement results of the second measurement unit; The reception unit receives the bit group symbol; The decoding unit decodes the received bit group symbol based on the frequency combination.

4. A method for constructing a collaborative network model of a multi-domain control system according to claim 3, characterized in that: The transmission unit includes a source input module, a bit grouping module, a mapper, and a modulator; The source input module is used to receive the bit information sequence to be modulated after source coding; The bit grouping module divides the bit information sequence into groups of K bits and provides them to the symbol mapper. The K-bit sequence has 2 K combinations; The mapper maps a combination into a bit group symbol according to the agreed confidentiality rules; The modulator selects M carrier frequencies from the wireless resources measured by the first measurement unit, and modulates 4 adjacent symbols onto four phases of one of the M carrier frequencies.

5. A method for constructing a collaborative network model of a multi-domain control system according to claim 3, characterized in that: The decoding unit includes multiple reception modules, a multiplex demodulation module, and a calculation module; Each of the reception modules converts the signal received by the reception antenna into electrical data; The multiplex demodulation module includes a multiplication module, an integration module, and a modulo module; The multiplication module is used to multiply each electrical data by the complex conjugate value of the bit group symbol and send it to the integration module; The integration module integrates the received signal within one symbol length in the time domain, and then sends it to the modulo module to take its absolute value to obtain the received value, and the calculation module obtains the bit confidence probability according to the received value.

6. A method for constructing a multi-domain control system collaborative network model according to claim 5, characterized in that: The calculation module includes a probability neural network, the probability neural network includes an input layer, a hidden layer, and an output layer, the input layer inputs the received value of the multiplex demodulation module, and the output layer outputs the bit confidence probability.

7. A system for constructing a collaborative network model of a multi-domain control system, characterized in that: It includes a storage medium and several processors. The storage medium stores computer program codes, and the processors call the computer program codes in the storage medium to implement the method according to any one of claims 1-2.

Images