A data fusion architecture design method for multi-level unified situation generation of a distributed platform

Abstract

The application discloses a data fusion architecture design method for multi-stage unified situation generation of a distributed platform, and belongs to the technical field of information fusion. Firstly, each platform shares target track information detected by a heterogeneous sensor on a network; meanwhile, each platform collects the target track information shared by other platforms on the network, and forms situation information of the platform after pre-processing, association and fusion processing; finally, the situation target information is numbered by a unified numbering rule, and unified situations of the platforms in the network are generated, so as to provide situation support for cross-domain cooperative operation.

Description

Technical Field

[0001] This invention relates to the field of information fusion technology, and more specifically, to a data fusion architecture design method for generating multi-level unified situations on distributed platforms. Background Technology

[0002] Cross-domain collaborative command and control operations involve diverse platforms, leading to problems such as significant differences in sensor target characteristics and large information transmission latency in networks. Simultaneously, the complexity, adversarial nature, and uncertainty of air combat pose considerable challenges to establishing a unified situation across different platforms. Existing research on airborne information fusion technology primarily focuses on similar, small-scale platforms, lacking research on data fusion architecture design methods for generating multi-level unified situation across cross-domain platforms. This fails to meet the requirements of future distributed cross-domain collaborative command and control operations. Therefore, providing a data fusion architecture design method for generating multi-level unified situation across distributed platforms has practical military significance. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a data fusion architecture design method for multi-level unified situation generation of distributed platforms for cross-domain collaborative command and control scenarios. By designing a multi-level unified situation architecture of COP, CTP and FCP, and a target batch allocation principle based on distance priority, the invention achieves the unification of situation among distributed combat platforms and improves the unified situation perception capability of cross-domain platforms.

[0004] To achieve the purpose of this invention, the technical solution adopted is as follows: a data fusion architecture design method for multi-level unified situation generation on a distributed platform, comprising the following steps: S1. Each platform shares the target point information detected by heterogeneous sensors to the network; S2. Each platform obtains target point information shared by other platforms in the network; S3. After receiving all network-shared point information, each platform performs preprocessing, correlation and fusion processing to form COP situation information; S4. Each platform selects the original target point information from the COP situation information based on the list provided by the system, performs preprocessing, association and fusion processing to form CTP situation information; S5. Each platform selects the original target point information from the CTP situation information based on the list provided by the system, performs preprocessing, correlation and fusion processing to form FCP situation information; S6. Use FCP situation information to update CTP situation, and use CTP situation information to update COP situation. S7. Calculate the distance between the target and the origin of the geodetic coordinate system in the COP situation. The closer the target is, the higher the priority of the target batch number allocation. The farther the target is, the lower the priority of the target batch number allocation. S8. The target retrieves a number from the numbering library in priority order as a temporary number. After the number remains unchanged within a set time, it is output as a final number.

[0005] Furthermore, in step S7, the distance between the target and the origin of the geodetic coordinate system is calculated in the geodetic coordinate system as the principle for target batch number allocation.

[0006] The beneficial effects of this invention are as follows: This method provides a data fusion architecture design method for multi-level unified situation generation on distributed platforms. Based on shared network point information, it achieves cross-domain platform distributed fusion through multi-level fusion and unified numbering rules, improving the openness, correctness, and consistency of the fusion architecture. This method is applicable to complex cross-domain platform formation and collaborative combat situation unification, providing reference information to help our personnel grasp the battlefield situation and make reasonable plans. Attached Figure Description

[0007] Figure 1 Sensor fusion architecture diagram.

[0008] Figure 2 yes Figure 1 A magnified view of the receiving end.

[0009] Figure 3 yes Figure 1 A magnified view of the fusion algorithm.

[0010] Figure 4 yes Figure 1 A magnified view of the transmitting end.

[0011] Figure 5 Simulation results are generated into a graph. Detailed Implementation

[0012] The present invention will be specifically described below through embodiments. It should be noted that the following embodiments are only used to further illustrate the present invention, but are not limited thereto, unless otherwise stated.

[0013] The specific embodiments of the present invention are described in detail below with reference to the technical solutions: According to one embodiment of the present invention, a data fusion architecture design method for multi-level unified situation generation on a distributed platform is provided. First, each platform shares target point information detected by heterogeneous sensors on the network. Simultaneously, each platform collects target point information shared by other platforms on the network, performs preprocessing, correlation, and fusion processing to form its own situation information. Finally, through a unified numbering rule, the situation target information is numbered to generate a unified situation across all platforms within the network, providing situational support for cross-domain collaborative operations.

[0014] Specifically, the following steps are included: a. Each platform shares the target point trace information detected by its heterogeneous sensors to the network. At the same time, each platform obtains the target point trace information shared by other platforms in the network. b. After receiving all network-shared point information, each platform performs preprocessing (including coordinate system transformation, dimension unification, etc.), association (multi-hypothesis association), and fusion (IMM adaptive filtering) of the target information to form COP situation information (update cycle 20 seconds). Each platform selects the original target point information from the COP situation information based on the list information provided by the system, performs preprocessing, association, and fusion to form CTP situation information (update cycle 1 second). Each platform selects the original target point information from the CTP situation information based on the list provided by the system, performs preprocessing, association, and fusion to form FCP situation information (update cycle 25 milliseconds). c. Use the situation information generated by FCP to update the CTP situation in reverse, and use the situation information generated by CTP to update the COP situation in reverse. d. Calculate the distance between the target and the origin of the geodetic coordinate system in the COP situation. The closer the target is, the higher the priority of the target batch number assignment. The farther the target is, the lower the priority of the target batch number assignment. The target obtains a number from the numbering library according to the priority order as a temporary number. After 4 seconds if there is no change, it is output as the final number.

[0015] In step a above, each platform shares its own sensor data point information. Simultaneously, each platform collects data point information shared by other platforms within the network and performs data fusion processing.

[0016] In step b above, CTP situation information is generated from COP situation information, and FCP situation information is generated from CTP situation information, thus achieving the downward decomposition and generation of a unified situation at three levels. Simultaneously, to ensure the accuracy of the correlation between slower-refreshing situations, faster-refreshing situations are used to supplement the slower-refreshing situations upwards.

[0017] In step d above, the distance between the target and the origin of the geodetic coordinate system is calculated and used as the principle for target batch number allocation. Simultaneously, by setting temporary batch numbers, the problem of inconsistent batch number allocation caused by link latency is resolved.

[0018] Figure 2 This is the target data processing architecture at the receiving end (callback mode). Its core is the hierarchical fusion and standardized processing of multi-source data, with the following logical flow: First, it connects to three data sources: this platform (inertial navigation, radar, optical radar, etc.), other platforms (inertial navigation, radar), and target lists (SIAP, SD). Radar data from this platform is split into COP / CTP / FCP levels using the original target list numbers. Electronic warfare data is split into two modules based on target alarm status after fusion level splitting. Data from other platforms is pre-splitted into COP / CTP / FCP levels and then filtered using target numbers. After AOI filtering and dimension-missing filtering, all data is finally matched and cached with the target data through synchronization caching logic (based on a target-to-inertial navigation timescale difference of <100ms). Subsequently, this architecture, through hierarchical splitting, cross-domain filtering, and timescale matching, achieves collaborative fusion of multi-platform and multi-sensor data, providing a technical link for the mutual understanding and sharing of target data in multi-domain operations.

[0019] Figure 3 The entire business process of multi-link parallel collaboration is presented. Taking COP data processing as an example: First, COP data from another platform is accessed. The spatiotemporal registration node completes the alignment of time and space dimensions based on the uniform extrapolation of the target's three-dimensional velocity. Then, the spatial registration stage is entered to unify the data coordinate system. The registered data and the tracks in the COP file library are input together into the track association module based on GNN for association matching. The matching result is divided into nodes based on the association success judgment: if the association is successful, the COP track status is updated and the fusion recognition stage is entered; if the association fails, the operation of creating a new COP track is executed. Finally, the updated or newly created tracks are integrated by the fused COP output nodes, sorted from near to far according to the distance to the fixed point on Earth, and a unique ID is generated through the R / M sliding window mechanism. The track data is then input into a GNN-based track association module for matching. The matching result is then split into nodes based on the association success judgment: if the association is successful, the COP track status is updated and the fusion recognition stage is entered to complete the secondary verification and integration of track information; if the association fails, the operation of creating a new COP track is executed; finally, the updated or newly created track is summarized by the fused COP output nodes and written to the COP file library to complete the persistent storage of data.

[0020] CTP and FCP data are processed using the same method described above. At the COP / CTP / FCP level, bidirectional data flow is performed, with IDs distributed to lower levels and target state vectors distributed to higher levels, enabling the sharing and synchronization of fused track information across multi-level systems.

[0021] Figure 4 This is a hierarchical data distribution process at the sending end (queue buffer mode). Data to be transmitted is uniformly stored and scheduled in the queue buffer module. Then, based on a multi-domain collaborative hierarchical data standard, COP, CTP, and FCP data are separated from the queue buffer according to priority or type, and distributed to their respective COP, CTP, and FCP data channels. This achieves ordered output of multi-level data, adapting to the differentiated data needs of different levels of nodes in multi-domain operations. COP data returned to the receiving end, along with the results obtained from the target list through target number lookup, enters the fusion-level filtering stage, where data of the corresponding level is selected by number and integrated into the data processing at the receiving end. Through the buffer scheduling and hierarchical distribution mechanism of the queue buffer, the stability and hierarchical targeting of data transmission at the sending end are ensured.

[0022] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data fusion architecture design method for multi-level unified situation generation on a distributed platform, characterized in that, Includes the following steps: S1. Each platform shares the target point information detected by heterogeneous sensors to the network; S2. Each platform obtains target point information shared by other platforms in the network; S3. After receiving all network-shared point information, each platform performs preprocessing, correlation and fusion processing to form COP situation information; S4. Each platform selects the original target point information from the COP situation information based on the list provided by the system, performs preprocessing, association and fusion processing to form CTP situation information; S5. Each platform selects the original target point information from the CTP situation information based on the list provided by the system, performs preprocessing, correlation and fusion processing to form FCP situation information; S6. Use FCP situation information to update CTP situation, and use CTP situation information to update COP situation. S7. Calculate the distance between the target and the origin of the geodetic coordinate system in the COP situation. The closer the target is, the higher the priority of the target batch number allocation. The farther the target is, the lower the priority of the target batch number allocation. S8. The target retrieves a number from the numbering library in priority order as a temporary number. After the number remains unchanged within a set time, it is output as a final number. COP data processing flow: First, COP data from other platforms is accessed. The spatiotemporal registration node aligns the time and spatial dimensions based on the target's three-dimensional velocity through uniform extrapolation. Then, the spatial registration stage is entered to unify the data coordinate system. The registered data and the tracks in the COP file library are input into the track association module based on GNN for association matching. The matching result is split into nodes based on the association success judgment: if the association is successful, the COP track status is updated and the fusion recognition stage is entered. If the association fails, a new COP track is created. Finally, the updated or newly created tracks are integrated by the merged COP output nodes, sorted from near to far by distance to a fixed point on Earth, and a unique ID is generated through the R / M sliding window mechanism. The track data is then input into a GNN-based track association module for matching; The matching result is successfully associated and the node is divided: if the association is successful, the COP track status is updated and the fusion recognition stage is entered to complete the secondary verification and integration of track information; If the association fails, a new COP track is created. Finally, the updated or newly created tracks are aggregated by the merged COP output nodes and written to the COP file library to complete the persistent storage of data.

2. The data fusion architecture design method for multi-level unified situation generation on a distributed platform according to claim 1, characterized in that, In step S7, the distance between the target and the origin of the geodetic coordinate system is calculated in the geodetic coordinate system as the principle for target batch number allocation.

Images