Multi-node situation unification method and system based on information diffusion mechanism

Abstract

The application discloses a multi-node situation unification method and system based on an information diffusion mechanism, and the method comprises the following steps: node information fusion: each node performs multi-node data association and data fusion on the information detected by the node and the information of other nodes received by the node through a network, and completes target unification batch and target position estimation; information diffusion transmission: each node diffuses and transmits the node information fusion result in the local field of the node; network fusion weight calculation: network fusion weights are calculated according to node related information; target position fusion estimation: the fusion estimation of the target position is completed according to the network fusion weights and node measurement weights; consistency decision: whether the iteration termination condition is met is decided according to the iteration number and the multi-node target position deviation. The application can overcome the long information transmission chain of the traditional situation unification method, the complex conflict resolution process, has good robustness, and has faster situation unification speed and better situation unification effect.

Description

Technical Field

[0001] This invention relates to the field of distributed situational awareness technology and is applied to multi-node distributed situational awareness scenarios. It unifies the number of targets, target batch numbers, target locations, etc., between each node. Specifically, it is a multi-node situational awareness unification method and system based on an information diffusion mechanism. Background Technology

[0002] The problem of unified situation awareness across multiple nodes has significant applications in various fields, including civilian and military applications. This problem requires the separate association and fusion calculation of multiple nodes to determine their target quantity, target batch number, target location, and other parameters. Traditional unified situation awareness methods include centralized methods and decentralized distributed methods.

[0003] Centralized situational awareness unification requires a central node to collect raw measurement information from all other nodes. After the central node completes the unified situational awareness, it distributes the unified situational awareness to the other nodes. Centralized situational awareness unification suffers from a long information transmission chain, slow unified situational awareness generation speed, and poor robustness.

[0004] The decentralized distributed situational awareness unification method achieves situational awareness unification by broadcasting to each other in pairs, collecting raw measurement information from all other nodes, and then processing the unified situational awareness and conflict resolution. However, decentralized distributed situational awareness unification requires a complex conflict resolution process, and the effectiveness of the unified situational awareness is difficult to guarantee. Furthermore, it is typically only suitable for scenarios involving a small number of peer nodes with ample transmission bandwidth resources. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a multi-node situational awareness unification method and system based on an information diffusion mechanism. Nodes transmit their fused post-processing results through mutual diffusion. Through information fusion and interaction between nodes, a unified situational awareness is generated iteratively and consistently. This overcomes the limitations of traditional situational awareness unification methods, which involve long information transmission chains and complex conflict resolution processes, and offers superior robustness. Compared to traditional methods, this invention achieves faster and more effective situational awareness unification.

[0006] The technical solution adopted in this invention is as follows:

[0007] On the one hand, this invention proposes a multi-node situational unification method based on an information diffusion mechanism, comprising the following steps:

[0008] S1. Node Information Fusion: Each node performs multi-node data association and data fusion based on the information detected by the node and the information received by the node from other nodes through the network, to complete the unified batching of targets and the estimation of target locations;

[0009] S2. Information diffusion and transmission: Each node diffuses and transmits the node information fusion result within its own local domain. The node information fusion result includes the total number of targets, the target batch number, and the target location.

[0010] S3. Network fusion weight calculation: Calculate the network fusion weight based on node-related information, including node network topology, node connectivity status, and node communication quality;

[0011] S4. Target Location Fusion Estimation: Based on the network fusion weights and node measurement weights, complete the fusion estimation of the target location;

[0012] S5. Consistency Decision: Based on the number of iterations and the target position deviation of multiple nodes, determine whether the iteration termination condition is met; if it is met, information diffusion and transmission will no longer be carried out; if it is not met, information diffusion and transmission will still be carried out until the iteration termination condition is met.

[0013] Furthermore, in step S1, each node uses a location association method to perform multi-node data association on the information detected by the node and the information of other nodes received by the node through the network, and completes the unified batching of targets; for each target, the covariance cross method is used to perform multi-node data fusion to complete the target location estimation.

[0014] Furthermore, in step S3, the method for calculating the network fusion weights based on node-related information includes:

[0015]

[0016] in, Represents a node and nodes The network fusion weight, Represents a node The number of adjacent nodes, Represents a node The number of adjacent nodes, Represents a node and nodes connect.

[0017] Furthermore, in step S4, the target location is estimated by using the generalized covariance cross method based on the network fusion weights and node measurement weights.

[0018] Furthermore, in step S5, the method for calculating the target position deviation includes:

[0019]

[0020] in, Gaussian distribution , Wasserstein distance, The state of the goal, Let the mean of the states at iteration time 1 be the target. Let the mean of the states at iteration time 2 be the target. Let Variance be the state variance of the objective at iteration time 1. Let Variance be the state variance of the objective at iteration time 2. Let the mean norm of the target state at iteration times 1 and 2 be . Let be the state variance norm of the target at iteration times 1 and 2.

[0021] On the other hand, this invention proposes a multi-node situational awareness system based on an information diffusion mechanism, comprising:

[0022] The node information fusion module is configured to perform multi-node data association and data fusion on the information detected by this node and the information of other nodes received by this node through the network, so as to complete the unified batching of targets and the estimation of target positions.

[0023] The information diffusion and transmission module is configured to diffuse and transmit node information fusion results within the local domain of the node itself. The node information fusion results include the total number of targets, target batch number, and target location.

[0024] The network fusion weight calculation module is configured to calculate the network fusion weight based on node-related information, including node network topology, node connectivity status, and node communication quality.

[0025] The target location fusion estimation module is configured to perform target location fusion estimation based on network fusion weights and node measurement weights.

[0026] The consistency decision module is configured to determine whether the iteration termination condition is met based on the number of iterations and the target position deviation of multiple nodes. If the condition is met, information diffusion and transmission will no longer be performed. If the condition is not met, information diffusion and transmission will still be performed until the iteration termination condition is met.

[0027] Furthermore, in the node information fusion module, each node uses a location association method to perform multi-node data association on the information detected by the node and the information of other nodes received by the node through the network, so as to complete the unified batching of targets; for each target, the covariance cross method is used to perform multi-node data fusion to complete the target location estimation.

[0028] Furthermore, in the network fusion weight calculation module, the method for calculating the network fusion weight based on node-related information includes:

[0029]

[0030] in, Represents a node and nodes The network fusion weight, Represents a node The number of adjacent nodes, Represents a node The number of adjacent nodes, Represents a node and nodes connect.

[0031] Furthermore, in the target location fusion estimation module, the generalized covariance cross method is used to complete the fusion estimation of the target location based on the network fusion weights and node measurement weights.

[0032] Furthermore, in the consistency decision module, the calculation method for the target position deviation includes:

[0033]

[0034] in, Gaussian distribution , Wasserstein distance, The state of the goal, Let the mean of the states at iteration time 1 be the target. Let the mean of the states at iteration time 2 be the target. Let Variance be the state variance of the objective at iteration time 1. Let Variance be the state variance of the objective at iteration time 2. Let the mean norm of the target state at iteration times 1 and 2 be . Let be the state variance norm of the target at iteration times 1 and 2.

[0035] The beneficial effects of this invention are as follows:

[0036] 1. This invention transmits the fusion results of the nodes through mutual diffusion between nodes. Through the information fusion and interaction between nodes, a unified situation is generated in a consistent iterative manner. This overcomes the long information transmission chain and complex conflict resolution process required by traditional situation unification methods, and has good robustness.

[0037] 2. Compared with traditional methods for unifying situational awareness, this invention avoids the problem of long information transmission chains in traditional methods, and generates unified situational awareness faster.

[0038] 3. Compared with traditional situational unification methods, the present invention can avoid complex post-processing of conflict resolution and achieve better situational unification performance. Attached Figure Description

[0039] Figure 1 This is a flowchart of the multi-node situational unification method based on the information diffusion mechanism of the present invention. Detailed Implementation

[0040] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments are now described. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention; that is, the described embodiments are only a part of the embodiments of the invention, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0041] Example 1

[0042] like Figure 1 As shown, this embodiment provides a multi-node situational awareness unification method based on an information diffusion mechanism. Nodes transmit their fusion results to each other through mutual diffusion, and a unified situational awareness is generated iteratively through information fusion and interaction between nodes. Specifically, this multi-node situational awareness unification method includes the following steps:

[0043] S1. Node Information Fusion: Each node performs multi-node data association and data fusion based on the information detected by the node and the information received by the node from other nodes through the network, to complete the unified batching of targets and the estimation of target locations;

[0044] S2. Information diffusion and transmission: Each node diffuses and transmits the node information fusion result within its own local domain. The node information fusion result includes the total number of targets, the target batch number, and the target location.

[0045] S3. Network fusion weight calculation: Calculate the network fusion weight based on node-related information, including node network topology, node connectivity status, and node communication quality;

[0046] S4. Target Location Fusion Estimation: Based on the network fusion weights and node measurement weights, complete the fusion estimation of the target location;

[0047] S5. Consistency Decision: Based on the number of iterations and the target position deviation of multiple nodes, determine whether the iteration termination condition is met; if it is met, information diffusion and transmission will no longer be carried out; if it is not met, information diffusion and transmission will still be carried out until the iteration termination condition is met.

[0048] This embodiment will explain in detail the steps of the method according to the specific calculation process and the formulas involved. See also: Figure 1 The steps shown are schematic diagrams.

[0049] Preferably, in step S1, each node uses a location association method to perform multi-node data association on the information detected by the node and the information of other nodes received by the node through the network, and completes the unified batching of targets; for each target, the covariance cross method is used to perform multi-node data fusion to complete the target location estimation.

[0050] Specifically, taking a unified situational awareness system with two nodes as an example, the covariance cross-validation method completes the correlation measurement of two different nodes with unknown correlation. , Data fusion to obtain estimates and the estimation matrix of its covariance matrix. The formula for calculating covariance using the cross-validation method is as follows:

[0051]

[0052]

[0053] Among them, parameters For the correlation measurement of two nodes and The weight, The performance of fusion estimation is determined by parameters The value of is determined by .

[0054] In step S2, each node diffuses and transmits the fusion post-processing result information of the node within its own local domain, including the total number of targets, target batch number, and target location. Traditional methods collect and share all raw measurement information of all nodes, with nodes outside their local domain transmitting the information via relays; however, the method of this invention does not require relays, but only diffuses and transmits the fusion post-processing result information of the node within its own local domain. This fusion post-processing result information includes the raw measurement characteristics of all nodes within the local domain.

[0055] Preferably, in step S3, the method for calculating the network fusion weights based on node-related information includes:

[0056]

[0057] in, Represents a node and nodes The network fusion weight, Represents a node The number of adjacent nodes, Represents a node The number of adjacent nodes, Represents a node and nodes connect.

[0058] Preferably, in step S4, the target location is estimated using the generalized covariance cross method based on the network fusion weights and node measurement weights. Specifically, taking a two-node unified situation system as an example, the generalized covariance cross method completes the correlation estimation of two different nodes with unknown correlation. , Data fusion to obtain estimates and the estimation matrix of its covariance matrix. The generalized covariance cross-validation method adds a weighted layer compared to the covariance cross-validation method. The formula for calculating the generalized covariance cross-validation method is as follows:

[0059]

[0060]

[0061] Among them, parameters For network fusion weights, ;parameter For the correlation measurement of two nodes and The weight, The performance of fusion estimation is determined by parameters The value of is determined by .

[0062] In step S5, based on the number of iterations and the target position deviation of multiple nodes, it is determined whether the iteration termination condition is met. The thresholds for the number of iterations and the target position deviation of multiple nodes can be set manually. Preferably, in step S5, the calculation method for the target position deviation includes:

[0063]

[0064] in, Gaussian distribution , The Wasserstein distance can effectively describe the target position deviation; The state of the goal, Let the mean of the states at iteration time 1 be the target. Let the mean of the states at iteration time 2 be the target. Let Variance be the state variance of the objective at iteration time 1. Let Variance be the state variance of the objective at iteration time 2. Let the mean norm of the target state at iteration times 1 and 2 be . Let be the state variance norm of the target at iteration times 1 and 2.

[0065] If the iteration termination condition is met, information diffusion and transmission will cease; if not, information diffusion and transmission will continue until the iteration termination condition is met.

[0066] Example 2

[0067] This embodiment provides a multi-node situational awareness system based on an information diffusion mechanism, including:

[0068] The node information fusion module is configured to perform multi-node data association and data fusion on the information detected by this node and the information of other nodes received by this node through the network, so as to complete the unified batching of targets and the estimation of target positions.

[0069] The information diffusion and transmission module is configured to diffuse and transmit node information fusion results within the local domain of the node itself. The node information fusion results include the total number of targets, target batch number, and target location.

[0070] The network fusion weight calculation module is configured to calculate the network fusion weight based on node-related information, including node network topology, node connectivity status, and node communication quality.

[0071] The target location fusion estimation module is configured to perform target location fusion estimation based on network fusion weights and node measurement weights.

[0072] The consistency decision module is configured to determine whether the iteration termination condition is met based on the number of iterations and the target position deviation of multiple nodes. If the condition is met, information diffusion and transmission will no longer be performed. If the condition is not met, information diffusion and transmission will still be performed until the iteration termination condition is met.

[0073] Preferably, in the node information fusion module, each node uses a location association method to perform multi-node data association on the information detected by the node and the information of other nodes received by the node through the network, so as to complete the unified batching of targets; for each target, the covariance cross method is used to perform multi-node data fusion to complete the target location estimation.

[0074] Specifically, taking a unified situational awareness system with two nodes as an example, the covariance cross-validation method completes the correlation measurement of two different nodes with unknown correlation. , Data fusion to obtain estimates and the estimation matrix of its covariance matrix. The formula for calculating covariance using the cross-validation method is as follows:

[0075]

[0076]

[0077] Among them, parameters For the correlation measurement of two nodes and The weight, The performance of fusion estimation is determined by parameters The value of is determined by .

[0078] In the information diffusion and transmission module, each node diffuses and transmits the fusion post-processing result information of the nodes within its own local domain, including the total number of targets, target batch number, and target location. Traditional methods collect and share all raw measurement information from all nodes, with nodes outside their local domain transmitting information via relays. However, the method of this invention eliminates the need for relays, diffusing and transmitting the fusion post-processing result information of the nodes only within its own local domain. This fusion post-processing result information includes the raw measurement characteristics of all nodes within the local domain.

[0079] Preferably, in the network fusion weight calculation module, the method for calculating the network fusion weight based on node-related information includes:

[0080]

[0081] in, Represents a node and nodes The network fusion weight, Represents a node The number of adjacent nodes, Represents a node The number of adjacent nodes, Represents a node and nodes connect.

[0082] Preferably, in the target location fusion estimation module, the generalized covariance cross method is used to complete the fusion estimation of the target location based on the network fusion weights and node measurement weights. Specifically, taking a two-node unified situation system as an example, the generalized covariance cross method completes the correlation estimation of two different nodes with unknown correlation. , Data fusion to obtain estimates and the estimation matrix of its covariance matrix. The generalized covariance cross-validation method adds a weighted layer compared to the covariance cross-validation method. The formula for calculating the generalized covariance cross-validation method is as follows:

[0083]

[0084]

[0085] Among them, parameters For network fusion weights, ;parameter For the correlation measurement of two nodes and The weight, The performance of fusion estimation is determined by parameters The value of is determined by .

[0086] In the consistency decision module, the iteration termination condition is determined based on factors such as the number of iterations and the target position deviation of multiple nodes. The thresholds for the number of iterations and the target position deviation of multiple nodes can be set manually. Preferably, the calculation method for the target position deviation includes:

[0087]

[0088] in, Gaussian distribution , The Wasserstein distance can effectively describe the target position deviation; The state of the goal, Let the mean of the states at iteration time 1 be the target. Let the mean of the states at iteration time 2 be the target. Let Variance be the state variance of the objective at iteration time 1. Let Variance be the state variance of the objective at iteration time 2. Let the mean norm of the target state at iteration times 1 and 2 be . Let be the state variance norm of the target at iteration times 1 and 2.

[0089] If the iteration termination condition is met, information diffusion and transmission will cease; if not, information diffusion and transmission will continue until the iteration termination condition is met.

[0090] It should be noted that, for the sake of simplicity, the foregoing method embodiments are described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

Claims

1. A multi-node situational awareness unification method based on an information diffusion mechanism, characterized in that, Includes the following steps: S1. Node Information Fusion: Each node performs multi-node data association and data fusion based on the information detected by the node and the information received by the node from other nodes through the network, to complete the unified batching of targets and the estimation of target locations; S2. Information diffusion and transmission: Each node diffuses and transmits the node information fusion result within its own local domain. The node information fusion result includes the total number of targets, the target batch number, and the target location. S3. Network fusion weight calculation: Calculate the network fusion weight based on node-related information, including node network topology, node connectivity status, and node communication quality; S4. Target Location Fusion Estimation: Based on the network fusion weights and node measurement weights, complete the fusion estimation of the target location; S5. Consistency Decision: Based on the number of iterations and the target position deviation of multiple nodes, determine whether the iteration termination condition is met; if it is met, information diffusion and transmission will no longer be performed. If the conditions are not met, information diffusion and transmission must continue until the iteration termination condition is met. In step S5, the method for calculating the target position deviation includes: in, Gaussian distribution , Wasserstein distance, The state of the goal, Let the mean of the states at iteration time 1 be the target. Let the mean of the states at iteration time 2 be the target. Let Variance be the state variance of the objective at iteration time 1. Let Variance be the state variance of the objective at iteration time 2. Let the mean norm of the target state at iteration times 1 and 2 be . Let be the state variance norm of the target at iteration times 1 and 2.

2. The multi-node situational awareness unification method based on information diffusion mechanism according to claim 1, characterized in that, In step S1, each node uses a location association method to associate multi-node data with the information detected by the node and the information of other nodes received by the node through the network, and completes the unified batching of targets; for each target, the covariance cross method is used to perform multi-node data fusion to complete the target location estimation.

3. The multi-node situational awareness unification method based on information diffusion mechanism according to claim 1, characterized in that, In step S4, the target location is estimated by using the generalized covariance cross method based on the network fusion weights and node measurement weights.

4. A multi-node situational awareness system based on an information diffusion mechanism, characterized in that, include: The node information fusion module is configured to perform multi-node data association and data fusion on the information detected by this node and the information of other nodes received by this node through the network, so as to complete the unified batching of targets and the estimation of target positions. The information diffusion and transmission module is configured to diffuse and transmit node information fusion results within the local domain of the node itself. The node information fusion results include the total number of targets, target batch number, and target location. The network fusion weight calculation module is configured to calculate the network fusion weight based on node-related information, including node network topology, node connectivity status, and node communication quality. The target location fusion estimation module is configured to perform target location fusion estimation based on network fusion weights and node measurement weights. The consistency decision module is configured to determine whether the iteration termination condition is met based on the number of iterations and the target position deviation of multiple nodes. If the conditions are met, information dissemination and transmission will cease. If the conditions are not met, information diffusion and transmission must continue until the iteration termination condition is met. In the consistency decision module, the calculation methods for the target position deviation include: in, Gaussian distribution , Wasserstein distance, The state of the goal, Let the mean of the states at iteration time 1 be the target. Let the mean of the states at iteration time 2 be the target. Let Variance be the state variance of the objective at iteration time 1. Let Variance be the state variance of the objective at iteration time 2. Let the mean norm of the target state at iteration times 1 and 2 be . Let be the state variance norm of the target at iteration times 1 and 2.

5. The multi-node situational awareness system based on an information diffusion mechanism according to claim 4, characterized in that, In the node information fusion module, each node uses a location association method to perform multi-node data association on the information detected by the node and the information of other nodes received by the node through the network, so as to complete the unified batching of targets; for each target, the covariance cross method is used to perform multi-node data fusion to complete the target location estimation.

6. The multi-node situational awareness system based on an information diffusion mechanism according to claim 4, characterized in that, In the target location fusion estimation module, the target location is estimated by using the generalized covariance cross method based on the network fusion weights and node measurement weights.

Images