Automatic translation of heterogeneous electronic system messages

Abstract

The application relates to the technical field of data processing, and discloses a heterogeneous electronic system message automatic translation method, which comprises the following steps: reconstructing each kind of heterogeneous message according to a graph structure for a plurality of heterogeneous messages which cannot directly interact; comparing all node attributes of different heterogeneous messages after reconstruction; judging the mappability between corresponding message fields; matching corresponding link rules and establishing edges according to the node corresponding to the message field with the mappability according to the heterogeneous type; searching for the shortest translation path from the source node corresponding to the source heterogeneous message to the target node corresponding to the target heterogeneous message by using a graph traversal algorithm; and realizing the automatic translation of the source heterogeneous message to the target heterogeneous message based on the shortest translation path and the link rules corresponding to the edges between the nodes on the path. The application can realize the quick analysis and translation of the heterogeneous messages output by the heterogeneous electronic system without upgrading hardware devices or changing existing system software.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a method for automatic message translation in heterogeneous electronic systems. Background Technology

[0002] Based on existing aviation system capabilities, an open system architecture is needed to distribute capabilities such as sensors, positioning, navigation, timing, and communication data links across numerous manned and unmanned platforms capable of real-time communication to perform tasks such as area surveillance and target identification. Therefore, the ability to exchange information between these platforms is crucial. However, currently, the heterogeneous electronic systems within these platforms are developed independently, such as sensors, communication systems, trackers, and display systems. Furthermore, these heterogeneous electronic systems lack a common standard for message definitions, and their information content is incompatible, preventing them from interacting with each other. The simplest solution to this interaction problem is to adopt a common standard specification for all electronic systems, but this approach also has drawbacks. For example, due to the wide range of electronic systems involved, the time, cost, and impact of equipping, deploying, upgrading, and modifying these systems would be significant, making it not currently the optimal solution.

[0003] Therefore, the following methods have been implemented to address the problem of heterogeneous message interaction: ① Local message standards, that is, establishing a translation path between every two heterogeneous messages, but for... N For heterogeneous messages, this method needs to construct... ① One translation channel is too costly technically; ② A common message standard, i.e., building a common message model, so that the interaction between each heterogeneous message can be translated through the common message model. This method is more effective when there are few messages to be translated. When the number of message types increases, the common message model to be built will be very large, and its maintenance cost will also be high. Moreover, it is impossible for all heterogeneous electronic systems to publicly disclose their message formats, making it difficult to achieve global uniformity. Summary of the Invention

[0004] To address the issue of message exchange between subsystems of the aforementioned platform electronic information system, this application provides an automatic message translation method for heterogeneous electronic systems. This method enables rapid parsing and translation of heterogeneous messages output by heterogeneous electronic systems without requiring hardware upgrades or modifications to existing system software, and also supports system integration of different types of electronic systems.

[0005] This application discloses an automatic message translation method for heterogeneous electronic systems, which includes: Step 1: For various heterogeneous messages that cannot be directly interacted with, each heterogeneous message is reconstructed according to a graph structure. Each message field of the heterogeneous message is used as a node in the graph, and the mathematical description of the message field is used as the attribute of the corresponding node. The attribute of the node includes at least the data element name, data type, byte length, and data element description. Step 2: Perform pairwise comparisons of all node attributes for the reconstructed heterogeneous messages. Based on the comparison results, determine the mappability between corresponding message fields. For nodes corresponding to message fields with mappability, match the corresponding link rules and establish edges based on the heterogeneity type between the two. The link rules are used to define the mapping relationship and data conversion method between the two message fields. A graph database with heterogeneous nodes and edges is composed of nodes, node attributes, and edges with link rules from various heterogeneous messages. Step 3: After determining the source heterogeneous message and the target heterogeneous message that need to be translated, use a graph traversal algorithm in the graph database to search for the shortest translation path from the source node corresponding to the source heterogeneous message to the target node corresponding to the target heterogeneous message. Based on the shortest translation path and the link rules corresponding to the edges between nodes on the path, the automatic translation from the source heterogeneous message to the target heterogeneous message is realized.

[0006] Furthermore, after step 3, the method further includes: If a new heterogeneous message needs to be translated, the new heterogeneous message is reconstructed according to the graph structure reconstruction rules in step 1. The attributes of its nodes are compared with those of the nodes of existing heterogeneous messages in the graph database, and the mappability is determined. Any node with mappability is selected to match the corresponding link rules to establish an edge. The graph structure of the new heterogeneous message is added to the constructed graph database, thereby realizing the automatic translation of the new heterogeneous message with all other heterogeneous messages in the graph database.

[0007] Furthermore, the graph database is a heterogeneous graph. ,in, This represents multiple heterogeneous nodes in a graph. This represents an edge between heterogeneous nodes with linked rules. , Represents a set of node types. Let the set of edge types satisfy the following condition: For a graph database composed of N heterogeneous messages, ,in Representing the Heterogeneous message types; links and The edges of the corresponding nodes are defined by a preset mapping function, where, and The value range is 1 to .

[0008] Furthermore, the heterogeneity types include syntactic heterogeneity, structural heterogeneity, and semantic heterogeneity; syntactic heterogeneity refers to differences in data type and data format of message fields; structural heterogeneity refers to differences in data structure, interface, and schema of message fields; and semantic heterogeneity refers to differences in the meaning of message-specific terms.

[0009] Furthermore, the linking rules are categorized based on the heterogeneity between the two message fields, including equal linking rules, difference linking rules, and discard linking rules.

[0010] Furthermore, the equi-linking rule states that when the data element names, data types, byte lengths, and data element descriptions in the node attributes of two message fields are completely identical and there is no heterogeneity between them, the nodes corresponding to the two message fields are directly equi-linked without any data transformation.

[0011] Furthermore, the difference linking rule is as follows: when two message fields only have syntactic heterogeneity or structural heterogeneity, the corresponding data items with syntactic heterogeneity are matched for precision or data type, and the precision bits or data types involved are processed according to the format specification of the target heterogeneous message; for fields with structural heterogeneity, data structure, interface and pattern conversion are performed to ensure that the conversion result is consistent with the format requirements of the target heterogeneous message, thus completing the difference linking of the corresponding nodes of the two message fields.

[0012] Furthermore, the discard link rule states that when two message fields have semantic heterogeneity, and the field of the source heterogeneous message cannot be represented by any data element in the target heterogeneous message, the two are determined to be unmappable, and no edge is established between the corresponding nodes of the two message fields.

[0013] Furthermore, the graph traversal algorithm is Dijkstra's algorithm, which calculates the shortest translation path from the source node to the target node in the graph database.

[0014] Furthermore, the step of calculating the shortest translation path from the source node to the target node in the graph database using the Dijkstra algorithm includes: Assume each node has a pair of labels. ,in, From the source node To the node The shortest path length is given by the fact that the shortest path from a node to itself is the zero path, which has a length of 0. From the source node To the node Nodes on the shortest path The previous node; solve from the source node To the node The shortest paths include: Initialization: The source node is set to , Empty From the source node To the source node The length of the shortest path, For the source node The source node on the shortest path to itself The previous point; all other points are set to , , From the source node To the node The length of the shortest path, Indicates unknown values; marks source nodes. ,remember All other nodes are unmarked; Distance update: Check all marked nodes To its directly connected unmarked nodes The distance, and set , It is a node To the node The direct connection distance; Select the next node: From all unlabeled nodes, select... smallest node ;node It is selected as a node in the shortest path and marked as marked; Update predecessor: Find node The previous point, find the node directly connected to the marked node. nodes As the previous node, set ; Execute in a loop: If all nodes have been marked, the algorithm completes. Then, based on the predecessor node... By backtracking, we can obtain the source node. To the node The shortest path, otherwise Then proceed to distance update until all nodes have been marked.

[0015] Due to the adoption of the above technical solution, this application has the following advantages: 1. No hardware upgrades or disruption of existing system software are required. This invention does not require modification to existing electronic systems, nor does it mandate the adoption of universal interface standards. Instead, it rapidly creates the necessary connections based on existing capabilities, avoiding the need to upgrade hardware or existing system software for interaction. Based on this technology, each subsystem only needs to submit an interface, describe the messages it will use, collect all messages it wishes to interact with, remodel them using a graph structure, and link them to field nodes of any heterogeneous messages in the established graph database. This enables automatic translation with other heterogeneous messages.

[0016] 2. Fast compilation speed. This invention supports the rapid implementation of integrated system architecture capabilities. For N heterogeneous messages, only N translation channels need to be built to achieve automatic translation of any two messages. Furthermore, when integrating a new subsystem, simply linking the messages of the new subsystem with any heterogeneous message in the already built graph database enables interaction with any other subsystem. Therefore, compiling a new integrated system takes only a few minutes, reducing time costs. Each specific implementation is customized and optimized for the requirements of the integrated system architecture. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments recorded in the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a schematic diagram of a graph database built based on heterogeneous messages according to an embodiment of this application.

[0019] Figure 2 This is a schematic diagram of the translation path of heterogeneous messages in an embodiment of this application.

[0020] Figure 3 This is a schematic diagram of an example message graph structure and its linking relationships in an embodiment of this application.

[0021] Figure 4 This is a schematic diagram of an example heterogeneous message translation path in an embodiment of this application. Detailed Implementation

[0022] The present application will be further described in conjunction with the accompanying drawings and embodiments. The described embodiments are only some, not all, of the embodiments of the present application. All other embodiments obtained by those skilled in the art should fall within the protection scope of the embodiments of the present application.

[0023] See Figure 1This application provides an embodiment of an automatic message translation method for heterogeneous electronic systems, which includes: Step 1: For various heterogeneous messages that cannot be directly interacted with, each heterogeneous message is reconstructed according to a graph structure. Each message field of the heterogeneous message is used as a node in the graph, and the mathematical description of the message field is used as the attribute of the corresponding node. The attribute of the node includes at least the data element name, data type, byte length, and data element description. Step 2: Perform pairwise comparisons of all node attributes for the reconstructed heterogeneous messages. Based on the comparison results, determine the mappability between corresponding message fields. For nodes corresponding to message fields with mappability, match the corresponding link rules and establish edges based on the heterogeneity type between the two. The link rules are used to define the mapping relationship and data conversion method between the two message fields. A graph database with heterogeneous nodes and edges is composed of nodes, node attributes, and edges with link rules from various heterogeneous messages. Step 3: After determining the source heterogeneous message and the target heterogeneous message that need to be translated, use a graph traversal algorithm in the graph database to search for the shortest translation path from the source node corresponding to the source heterogeneous message to the target node corresponding to the target heterogeneous message. Based on the shortest translation path and the link rules corresponding to the edges between nodes on the path, the automatic translation from the source heterogeneous message to the target heterogeneous message is realized.

[0024] Figure 1 In this process, after obtaining the messages that need to be translated, each message is created as a graphical database. Figure 2 Specifically, this refers to the translation path of the subfield "ID" in heterogeneous messages 1 through 6 in the graph database. Based on this, automatic translation of subfields in various heterogeneous messages can be achieved.

[0025] Optionally, after step 3, the method further includes: If a new heterogeneous message needs to be translated, the new heterogeneous message is reconstructed according to the graph structure reconstruction rules in step 1. The attributes of its nodes are compared with those of the nodes of existing heterogeneous messages in the graph database, and the mappability is determined. Any node with mappability is selected to match the corresponding link rules to establish an edge. The graph structure of the new heterogeneous message is added to the constructed graph database, thereby realizing the automatic translation of the new heterogeneous message with all other heterogeneous messages in the graph database.

[0026] Optionally, the graph database is a heterogeneous graph. ,in, This represents multiple heterogeneous nodes in a graph. This represents an edge between heterogeneous nodes with linked rules. , Represents a set of node types. Let the set of edge types satisfy the following condition: For a graph database composed of N heterogeneous messages, ,in Representing the Heterogeneous message types; links and The edges of the corresponding nodes are defined by a preset mapping function, where, and The value range is 1 to .

[0027] Optionally, the heterogeneity types include syntactic heterogeneity, structural heterogeneity, and semantic heterogeneity; the syntactic heterogeneity refers to the differences in data type and data format of message fields; the structural heterogeneity refers to the differences in data structure, interface, and schema of message fields; and the semantic heterogeneity refers to the differences in the meaning of message-specific terms.

[0028] Optionally, the linking rules are divided according to the heterogeneity type between the two message fields, including equal linking rules, difference linking rules, and discard linking rules.

[0029] Optionally, the equi-linking rule is as follows: when the data element names, data types, byte lengths, and data element descriptions in the node attributes of two message fields are completely identical and there is no heterogeneity between them, the nodes corresponding to the two message fields are directly equi-linked without any data transformation.

[0030] Optionally, the difference linking rule is as follows: when two message fields only have syntactic heterogeneity or structural heterogeneity, the corresponding data items with syntactic heterogeneity are matched for precision or data type, and the precision bits or data types involved are processed according to the format specification of the target heterogeneous message; the fields with structural heterogeneity are converted in terms of data structure, interface and pattern, so that the conversion result is consistent with the format requirements of the target heterogeneous message, and the difference linking of the corresponding nodes of the two message fields is completed.

[0031] Optionally, the discard link rule is as follows: when two message fields have semantic heterogeneity, and the field of the source heterogeneous message cannot be represented by any data element in the target heterogeneous message, the two are determined to be unmappable, and no edge is established between the corresponding nodes of the two message fields.

[0032] Optionally, the graph traversal algorithm is Dijkstra's algorithm, which calculates the shortest translation path from the source node to the target node in the graph database.

[0033] Optionally, calculating the shortest translation path from the source node to the target node in the graph database using the Dijkstra algorithm includes: Assume each node has a pair of labels. ,in, From the source node To the node The shortest path length is given by the fact that the shortest path from a node to itself is the zero path, which has a length of 0. From the source node To the node Nodes on the shortest path The previous node; solve from the source node To the node The shortest paths include: Initialization: The source node is set to , Empty From the source node To the source node The length of the shortest path, For the source node The source node on the shortest path to itself The previous point; all other points are set to , , From the source node To the node The length of the shortest path, Indicates unknown values; marks source nodes. ,remember All other nodes are unmarked; Distance update: Check all marked nodes To its directly connected unmarked nodes The distance, and set , It is a node To the node The direct connection distance; Select the next node: From all unlabeled nodes, select... smallest node ;node It is selected as a node in the shortest path and marked as marked; Update predecessor: Find node The previous point, find the node directly connected to the marked node. nodes As the previous node, set ; Execute in a loop: If all nodes have been marked, the algorithm completes. Then, based on the predecessor node... By backtracking, we can obtain the source node. To the node The shortest path, otherwise Then proceed to distance update until all nodes have been marked.

[0034] For ease of understanding, this application provides a more specific embodiment: It is known that a UAV flight control system A sends an example message to a ground-based integrated display and control system C. The message sent is as follows: ①Target number: 3 ② Target longitude:

[0035] ③ Target latitude:

[0036] ④ Target altitude: 6000 meters ⑤ Recommended flight speed: 586.5 m / s ⑥ Drone ID: 31 ⑦ Working mode: 1 The message content sent by system A needs to be translated so that system C can successfully interpret it.

[0037] Based on the message formats of systems A, B, and C given in Tables 1, 2, and 3, reconstruct the graph structure model of the three heterogeneous message types A, B, and C as follows: Figure 3 And pairwise associations are established between systems A and B, and between systems B and C, linking identical subfields. Furthermore, the optimal translation path for subfields from system A to system C in a graph-structured database search is as follows: Figure 4 .

[0038] Table 1. Message format of System A to be translated

[0039] Table 2 Message Format of System B

[0040] Table 3. Message format of the target system C after translation

[0041] The linking rules between the message formats of System A and System B are shown in Table 4. Therefore, if an example message from System A is converted to the message format of System B, the message will be converted to: ①Target number: 3 ② Target longitude:

[0042] ③ Target latitude:

[0043] ④ Target altitude: 6000.00 meters ⑤ Recommended flight speed: 586.5 m / s ⑥ Working mode: 1 Table 4 System A Linking rules for System B

[0044] The linking rules between the message formats of System B and System C are shown in Table 5. Then, messages in System B format are translated into System C format, thus achieving message translation between Systems A, B, and C as shown in Table 6. The final message format is then converted to: ①Target number: 3 ② Target X coordinate (Geocentric coordinate system): rice ③ Target Y coordinate (Geocentric coordinate system): rice ④ Target Z coordinate (Geocentric coordinate system): rice ⑤ Recommended flight speed: 586.5 m / s ⑥ Working mode: 1 Therefore, the message can be received and parsed by system C to obtain specific data.

[0045] Table 5 System B Linking rules for system C

[0046] Table 6 Translation of System A to System C

[0047] System B is translated into System C through coordinate transformation, converting the geographic coordinate system (latitude, longitude, and altitude) into the geocentric-fixed coordinate system (ECEF). The given latitude, longitude, and altitude coordinates are... To convert to ECEF coordinates The conversion formula is as follows:

[0048] in, The geodetic height is the distance from a point on the reference ellipsoid along the normal direction to the reference ellipsoid. Latitude, unit: rad, with North latitude being positive. Longitude is expressed in rad, with east longitude being positive. e is the first eccentricity, and N is the radius of curvature of the reference ellipsoid. The calculation formula is as follows: The WGS84 ellipsoid parameters are as follows:

[0049] Where a is the major semi-axis and f is the flattening.

[0050] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application and not to limit them. Although this application has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of this application. Any modifications or equivalent substitutions that do not depart from the spirit and scope of this application should be covered within the protection scope of the claims of this application.

Claims

1. A method for automatic message translation in heterogeneous electronic systems, characterized in that, include: Step 1: For various heterogeneous messages that cannot be directly interacted with, each heterogeneous message is reconstructed according to a graph structure. Each message field of the heterogeneous message is used as a node in the graph, and the mathematical description of the message field is used as the attribute of the corresponding node. The attribute of the node includes at least the data element name, data type, byte length, and data element description. Step 2: Perform pairwise comparisons of all node attributes for the reconstructed heterogeneous messages. Based on the comparison results, determine the mappability between corresponding message fields. For nodes corresponding to message fields with mappability, match the corresponding linking rules according to the heterogeneity type between them. The linking rules are used to define the mapping relationship and data conversion method between the two message fields. The linking rules are divided according to the heterogeneity type between the two message fields, including equi-linking rules, difference-linking rules, and discard-linking rules. The equi-linking rule: When the data element names, data types, byte lengths, and data element descriptions in the node attributes of the two message fields are completely identical, and there is no heterogeneity between them, directly link the nodes corresponding to the two message fields equi-linked without any data transformation. The difference-linking rule: When the two message fields only have syntactic heterogeneity or structural heterogeneity, perform precision or data type matching on the corresponding data items with syntactic heterogeneity, and process the involved precision bits or data types according to the format specification of the target heterogeneous message. For fields exhibiting structural heterogeneity, data structure, interface, and schema transformations are performed to ensure that the transformation results are consistent with the format requirements of the target heterogeneous message, thus completing the difference link between the corresponding nodes of the two message fields. The discard link rule states that when two message fields exhibit semantic heterogeneity, and the field of the source heterogeneous message cannot be represented by any data element in the target heterogeneous message, they are determined to be unmappable, and no edge is established between the corresponding nodes of the two message fields. A graph database with heterogeneous nodes and edges is composed of nodes, node attributes, and edges with link rules from various heterogeneous messages. Step 3: After determining the source heterogeneous message and the target heterogeneous message that need to be translated, use a graph traversal algorithm in the graph database to search for the shortest translation path from the source node corresponding to the source heterogeneous message to the target node corresponding to the target heterogeneous message. Based on the shortest translation path and the link rules corresponding to the edges between nodes on the path, the automatic translation from the source heterogeneous message to the target heterogeneous message is realized.

2. The method for automatic message translation in heterogeneous electronic systems according to claim 1, characterized in that, After step 3, the following is also included: If a new heterogeneous message needs to be translated, the new heterogeneous message is reconstructed according to the graph structure reconstruction rules in step 1. The attributes of its nodes are compared with those of the nodes of existing heterogeneous messages in the graph database, and the mappability is determined. Any node with mappability is selected to match the corresponding link rules to establish an edge. The graph structure of the new heterogeneous message is added to the constructed graph database, thereby realizing the automatic translation of the new heterogeneous message with all other heterogeneous messages in the graph database.

3. The method for automatic message translation in heterogeneous electronic systems according to claim 1, characterized in that, The graph database is a heterogeneous graph. ,in, This represents multiple heterogeneous nodes in a graph. This represents an edge between heterogeneous nodes with linked rules. , Represents a set of node types. Let the set of edge types satisfy the following condition: For a graph database composed of N heterogeneous messages, ,in Representing the Heterogeneous message types; links and The edges of the corresponding nodes are defined by a preset mapping function, where, and The value range is 1 to .

4. The method for automatic message translation in heterogeneous electronic systems according to claim 1, characterized in that, The graph traversal algorithm is Dijkstra's algorithm, which calculates the shortest translation path from the source node to the target node in the graph database.

5. The method for automatic message translation in heterogeneous electronic systems according to claim 4, characterized in that, The calculation of the shortest translation path from the source node to the target node in the graph database using the Dijkstra algorithm includes: Assume each node has a pair of labels. ,in, From the source node To the node The shortest path length is given by the fact that the shortest path from a node to itself is the zero path, which has a length of 0. From the source node To the node Nodes on the shortest path The previous node; solve from the source node To the node The shortest paths include: Initialization: The source node is set to , Empty From the source node To the source node The length of the shortest path, For the source node The source node on the shortest path to itself The previous point; all other points are set to , , From the source node To the node The length of the shortest path, Indicates unknown values; marks source nodes. ,remember All other nodes are unmarked; Distance update: Check all marked nodes To its directly connected unmarked nodes The distance, and set , It is a node To the node The direct connection distance; Select the next node: From all unlabeled nodes, select... smallest node ;node It is selected as a node in the shortest path and marked as marked; Update predecessor: Find node The previous point, find the node directly connected to the marked node. nodes As the previous node, set ; Execute in a loop: If all nodes have been marked, the algorithm completes. Then, based on the predecessor node... By backtracking, we can obtain the source node. To the node The shortest path, otherwise Then proceed to distance update until all nodes have been marked.

Images