A Conflict Resolution Method for an Open Avionics Architecture

Abstract

The present application discloses a conflict resolution method for an open avionics architecture, which includes: constructing a standardized ontology model in the avionics field, including entity types, semantic relationships, and constraint rules; the entity types include standard classes, resource classes, interface classes, and security classes; the semantic relationships include inclusion, inheritance, compatibility, dependency, conflict, constraint, and version evolution; the constraint rules include version consistency, functional interoperability, and performance index constraints; building a multi-standard library based on a knowledge graph; and through a conflict detection and resolution mechanism, identifying and eliminating possible standard conflicts in the design process of the open avionics architecture. The present application realizes the unified modeling of multi-source heterogeneous avionics standards, the construction of a multi-standard library, and semantic-level conflict detection and resolution.

Description

Technical Field

[0001] This application relates to the technical fields of avionics system architecture design and knowledge engineering, and particularly to a conflict resolution method for an open avionics architecture. Background Art

[0002] With the rapid development of avionics technology and the continuous improvement of system complexity, numerous frameworks and standards have emerged and evolved in the field of avionics systems, greatly promoting the diversified development of system modules and components. However, these modules and components often adhere to different technical standards, leading to numerous challenges in system interaction. To enhance system flexibility and scalability and achieve efficient functional collaboration, an open avionics architecture multi-standard library has emerged. This architecture optimizes system interoperability and development efficiency by accommodating multiple avionics technical standards, providing a better solution for the construction and maintenance of complex avionics systems.

[0003] Constructing an open avionics architecture multi-standard library mainly faces two key problems: one is the problem of standard fragmentation. Avionics systems involve dozens of standards such as DO-178C, ARINC 653, FACE (Future Airborne Capability Environment), SOSA (Sensor Open System Architecture), etc. Traditional standard libraries mainly adopt static document storage methods, lacking semantic association and intelligent retrieval capabilities. The other is the problem of inefficient compatibility management between standards. The open avionics architecture needs to support multi-standard collaboration, but manually comparing elements such as interfaces, protocols, and security level differences between standards is time-consuming and error-prone. In addition, existing open avionics architecture multi-standard libraries mainly have two defects: one is that existing knowledge base technologies (such as relational databases) are difficult to effectively express the complex hierarchical relationships and constraint conditions between avionics system standards; the other is the lack of a conflict resolution mechanism specifically for the avionics technology field, which is prone to problems such as resource allocation strategy conflicts (ARINC 653 partition strategy and FACE resource scheduling mechanism) and interface protocol mismatches. Summary of the Invention

[0004] In view of this, this application provides a conflict resolution method for an open avionics architecture, which realizes unified modeling of multi-source heterogeneous avionics standards, construction of multi-standard libraries, and semantic-level conflict detection and resolution.

[0005] This application discloses a conflict resolution method for an open avionics architecture, which includes:

[0006] Step 1: Construct a standardized ontology model for the avionics field, including entity types, semantic relationships, and constraint rules; entity types include standard classes, resource classes, interface classes, and security classes; semantic relationships include inclusion, inheritance, compatibility, dependence, conflict, constraint, and version evolution; constraint rules include version consistency, functional interoperability, and performance index constraints;

[0007] Step 2: Construct a multi-standard library based on the knowledge graph;

[0008] Step 3: Identify and eliminate possible standard conflicts in the open avionics architecture design process through the conflict detection and resolution mechanism.

[0009] Furthermore, the said Step 1 includes:

[0010] Constructing a standardized ontology model in the avionics field includes multi-standard semantic extraction and the construction and update of a multi-standard library; among them, multi-standard semantic extraction includes defining the ontology structure of the avionics field; the ontology structure of the avionics field includes standard entities, the relationship types between entities, and constraint rules; standard entities include APIs, communication protocols, and security levels; the relationship types between entities include compatibility, dependency, and conflict; constraint rules include timing restrictions and resource allocation restrictions;

[0011] Extract triple information from relevant avionics standard documents based on NLP technology to achieve the extraction of explicit and implicit semantic associations; the triple information includes the subject, predicate, and object;

[0012] Update the multi-standard library by combining expert knowledge and predefined rules, and incrementally add the key information of the new standard document to the multi-standard library.

[0013] Furthermore, the construction and update of the said multi-standard library include:

[0014] Extract semantic information from multi-source heterogeneous avionics standard documents to generate an extensible knowledge graph and construct a multi-standard library: the multi-standard library extracts key terms, concepts, and their semantic relationships from avionics standard documents of multiple heterogeneous sources; avionics standard documents of multiple heterogeneous sources include international standards, industry standards, domain standards, and manufacturer technical manuals;

[0015] Realize the mining of implicit knowledge in standard documents through natural language processing technology. At the same time, the multi-standard library supports the unified processing of structured data and unstructured data, and through data cleaning and format conversion, converts them into a standardized knowledge representation form; structured data includes interface definition tables and protocol descriptions, and unstructured data includes PDF and WORD documents;

[0016] Based on the standardized knowledge representation form, generate a knowledge graph and construct a multi-standard library to dynamically reflect the latest development of avionics standards and support the seamless integration of new standards;

[0017] When a new avionics standard is released or the original standard is iteratively updated, combine expert knowledge and predefined rules to incrementally add the key information in the new standard document to the existing multi-standard library.

[0018] Furthermore, the said Step 2 includes:

[0019] The multi-standard library adopts a hierarchical design, which includes an application service layer, a knowledge construction layer, a data acquisition layer, and a data storage layer;

[0020] The application service layer provides intelligent service functions for users, which includes a conflict detection service component, a compatibility recommendation engine component, and a user interface component; among them, the conflict detection service component can automatically identify the conflict points between avionics standards through Datalog rule inference technology and output a detailed conflict report; the compatibility recommendation engine component calculates the similarity between standards based on graph embedding technology and generates an optimal adaptation plan through the analysis of the semantic relationship between standards; the user interface component provides an interactive query function and supports natural language input;

[0021] The knowledge construction layer extracts knowledge from the original data and constructs a multi-standard library based on a knowledge graph, which includes three functional components: an entity recognition module, a relationship extraction module, and a graph update module; among them, the entity recognition module uses a BiLSTM model to identify key terms in the standard document; the relationship extraction module, based on the BERT model, mines the semantic relationship between standards to ensure the semantic integrity of the knowledge graph; the graph update module combines expert knowledge and predefined rules to optimize and update the content of the multi-standard library;

[0022] The data acquisition layer obtains data from multiple sources, cleans and integrates it, and includes three modules: an avionics standard document library, a manufacturer's technical manual, and a data cleaning component; among them, the avionics standard document library stores the original files of multiple standards; the manufacturer's technical manual provides structured data, and the structured data includes an interface definition table and a protocol description; the data cleaning component uniformly converts heterogeneous data into a standardized JSON format, and the heterogeneous data includes text, tables, and charts;

[0023] The data storage layer uses a graph database to store and manage data, and the data organization form is centered on entity types, entity relationships, and entity attributes; among them, entity types: include "standard class" entities, "resource class" entities, "interface class" entities, and "security level class" entities; entity relationships cover multiple semantic relationships; multiple semantic relationships include "contains", "inherits", "compatible", "depends on", "conflict", "constraint", "version evolution", and support the relationship modeling between the "is-a" concept and instances; entity attributes: set corresponding attribute descriptions for different entity types.

[0024] Furthermore, the data storage layer uses graph database technology to store the structured and semi-structured data of avionics standards, supporting high-concurrency access, complex queries, and data version control;

[0025] The data acquisition layer is responsible for the acquisition and integration of multi-source heterogeneous data. It collects the original files of various relevant avionics standards through web crawlers and API interfaces, and conducts data cleaning and preprocessing to build a high-quality data pool;

[0026] The knowledge construction layer is responsible for extracting knowledge from avionics standard documents, building a multi-standard library, and analyzing the compatibility, dependency, and conflict of standards to provide semantic support for the request processing of the application service layer;

[0027] The application service layer receives user requests and provides intelligent services, including parsing avionics architecture design models, identifying standard conflicts and generating repair suggestions, and intelligently recommending adapted standard combination solutions.

[0028] Furthermore, the step 3 includes:

[0029] The conflict detection and resolution mechanism consists of a rule inference engine and a dynamic priority arbitration. The rule inference engine identifies conflicts based on the Datalog rule set and supports various types of conflict detection. Various types of conflicts include compatibility conflicts, resource allocation conflicts, and interface mismatch conflicts. The dynamic priority arbitration formulates a resolution strategy according to the standard weights and application scenarios.

[0030] Furthermore, the conflict detection and resolution mechanism introduces a dynamic priority strategy, and the priority weight calculation formula is: Priority = k × airworthiness level coefficient + p × task criticality coefficient + q × manufacturer recommended level; the priority strategy is related to the priority weight; the value ranges of k, p, and q are all from 0 to 1, p is greater than q and less than k.

[0031] Furthermore, the dynamic priority arbitration mainly makes decisions based on two dimensions: standard weights and application scenarios:

[0032] Standard weights: Different avionics standards have different priorities and application scopes. The dynamic priority arbitration comprehensively evaluates the authority, applicability, and mandatory nature of each standard to determine its weight in the current design scenario;

[0033] Application scenarios: The conflict resolution strategy is adjusted in combination with specific usage scenarios. The dynamic priority arbitration dynamically adjusts the priority rules for conflict resolution according to relevant factors. The relevant factors include task requirements, performance indicators, and system configuration;

[0034] The dynamic priority arbitration generates a specific conflict resolution strategy based on the conflict detection results output by the rule inference.

[0035] Furthermore, the generation of a specific conflict resolution strategy includes:

[0036] Standard Input and Requirement Analysis: The user inputs the open avionics architecture standards to be detected and their related parameters through a graphical interface or an API. The system performs semantic analysis and standardization processing on the input requirements;

[0037] Intelligent Retrieval of Knowledge Graph: Utilize the semantic association ability of the knowledge graph to screen the standard candidate set that meets the requirements according to the user's input conditions;

[0038] Intelligent Conflict Detection and Priority Ranking: The system calls the built-in conflict detection service, combines expert experience rules and historical case data, and conducts a comprehensive conflict detection on the standard candidate set; The content of conflict detection includes version conflict, function overlap, and technical incompatibility; After the conflict detection is completed, the system performs intelligent ranking on the conflict items according to the preset priority rules, and finally generates the optimal applicable standard set for the open avionics architecture; The priority rules include standard version level, applicable scope, and industry recognition;

[0039] Verification of Standard Set and Update of Multi-Standard Library: Conduct multi-dimensional verification on the generated optimal applicable standard set for the open avionics architecture. After passing the verification, the system feeds back the new standard information and conflict resolution results to the multi-standard library to complete the dynamic update of the multi-standard library; Multi-dimensional verification includes logical consistency check, function integrity evaluation, and actual application scenario simulation.

[0040] Furthermore, the judgment process of the technical incompatibility includes:

[0041] Knowledge Representation Learning: Train all entities and relationships in the knowledge graph to generate corresponding low-dimensional vector representations;

[0042] Entity Relationship Modeling: In the vector space, the TransE algorithm expresses the relationship between entities through vector operations;

[0043] Similarity Measurement: Quantitatively evaluate the semantic similarity between different entities by calculating the metrics between vectors; The metrics include distance or dot product;

[0044] Compatibility Judgment: If the semantic similarity output by the similarity measurement is lower than the preset value, it is considered that the standards are incompatible.

[0045] Due to the adoption of the above technical solutions, the present application has the following advantages:

[0046] 1. Knowledge-driven standard management: This application breaks through the traditional management mode based on the avionics system standard document library and innovatively uses knowledge graph technology for semantic association and intelligent reasoning of avionics standards. By constructing a structured knowledge graph, the system can automatically identify and understand the deep semantic relationships between different standards, realizing intelligent management and correlation analysis of standards. This management mode not only improves the retrievability and usability of avionics standard information but also provides a solid foundation for subsequent conflict detection and resolution.

[0047] 2. Dynamic conflict resolution: This application combines a rule engine and machine learning technology to propose a set of dynamic conflict resolution mechanisms. The rule engine can quickly identify and locate conflict points between standards according to predefined business rules and industry norms, while the machine learning model continuously optimizes the accuracy and response speed of conflict detection by analyzing historical conflict cases and avionics standard change data. This dynamic resolution mechanism can discover and solve compatibility problems in real time during the standard collaborative design process, significantly improving the reliability and design efficiency of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] To more clearly illustrate the technical solutions in the embodiments of this application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this application. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0049] Figure 1 is the architecture diagram of the multi-standard library in the embodiments of this application;

[0050] Figure 2 is the schematic diagram of the avionics domain ontology in the embodiments of this application;

[0051] Figure 3 is the flowchart of the conflict resolution method for an open avionics architecture in the embodiments of this application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0052] The present application will be further described in conjunction with the drawings and embodiments. The described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art shall fall within the scope of protection of the embodiments of the present application.

[0053] See Figure 3 , this application provides an embodiment of a conflict resolution method for an open avionics architecture, which includes:

[0054] Step 1: Construct a standardized ontology model for the avionics field, including entity types, semantic relationships, and constraint rules; the entity types include standard classes, resource classes, interface classes, and security classes; the semantic relationships include inclusion, inheritance, compatibility, dependence, conflict, constraint, and version evolution; the constraint rules include version consistency, functional interoperability, and performance metric constraints;

[0055] Step 2: Construct a multi-standard library based on the knowledge graph;

[0056] Step 3: Through the conflict detection and resolution mechanism, identify and eliminate possible standard conflicts in the open avionics architecture design process.

[0057] The conflict detection and resolution method proposed in this application not only realizes the intelligent management of open avionics architecture standards, but also significantly improves the adaptability and scalability of the system. This method is applicable to complex and changeable avionics system design scenarios, can effectively reduce the uncertainty risk of standard selection, and provides strong technical support for the flexibility and scalability of the system.

[0058] Optionally, the Step 1 includes:

[0059] Constructing a standardized ontology model for the avionics field includes multi-standard semantic extraction and the construction and update of a multi-standard library; among them, multi-standard semantic extraction includes defining the ontology structure of the avionics field; the ontology structure of the avionics field includes standard entities (such as APIs, communication protocols, security levels), relationship types between entities (such as compatibility, dependence, conflict), and constraint rules (such as timing restrictions, resource allocation restrictions); standard entities include APIs, communication protocols, security levels; relationship types between entities include compatibility, dependence, conflict; constraint rules include timing restrictions, resource allocation restrictions;

[0060] Extract triple information, that is, "subject-predicate-object", from relevant avionics standard documents based on natural language processing (NLP) technology. For example, extract "ARINC 653 - defines the partitioning mechanism - time / space isolation" from the DO-178C standard, and extract "FACE 3.1 - requires compatibility - POSIX API" from the FACE standard, etc. This semantic extraction process can not only extract explicit text information, but also discover implicit semantic associations through context reasoning;

[0061] When a new avionics standard is released or the original standard is iteratively updated, key information in the new standard document can be incrementally added to the existing multi-standard library by combining expert knowledge with predefined rules. This process not only ensures the dynamic update ability of the multi-standard library but also guarantees the accuracy and consistency of the newly added content. Through this modeling method, the avionics knowledge graph can effectively solve problems existing in traditional standard libraries, such as insufficient semantic associations and lack of intelligent retrieval capabilities, providing strong knowledge support for multi-standard collaboration in avionics systems.

[0062] Optionally, the construction and update of the multi-standard library include:

[0063] Extract semantic information from multi-source heterogeneous avionics standard documents to generate an extensible knowledge graph and construct a multi-standard library: The multi-standard library extracts key terms, concepts, and their semantic relationships from avionics standard documents of multiple heterogeneous sources; the avionics standard documents of multiple heterogeneous sources include international standards, industry standards, domain-specific standards, and manufacturer technical manuals.

[0064] Implement the mining of implicit knowledge in standard documents through natural language processing technology. At the same time, the multi-standard library supports the unified processing of structured data and unstructured data, and through data cleaning and format conversion, converts them into a standardized knowledge representation form; structured data includes interface definition tables and protocol descriptions, and unstructured data includes PDF and WORD documents.

[0065] Based on the standardized knowledge representation form, generate a knowledge graph and construct a multi-standard library to dynamically reflect the latest developments of avionics standards and support seamless integration of new standards.

[0066] When a new avionics standard is released or the original standard is iteratively updated, key information in the new standard document is incrementally added to the existing multi-standard library by combining expert knowledge with predefined rules.

[0067] Optionally, step 2 includes:

[0068] See Figure 1 , the multi-standard library adopts a hierarchical design, which includes an application service layer, a knowledge construction layer, a data collection layer, and a data storage layer; each layer undertakes specific functions and works together to achieve efficient management and support of multi-source heterogeneous avionics standards.

[0069] The application service layer provides intelligent service functions for users. It includes a conflict detection service component, a compatibility recommendation engine component, and a user interface component. Among them, the conflict detection service component can automatically identify the conflict points between avionics standards through Datalog rule inference technology and output a detailed conflict report. The compatibility recommendation engine component calculates the similarity between standards based on graph embedding technology and generates an optimal adaptation plan through the analysis of the semantic relationships between standards. The user interface component provides an interactive query function and supports natural language input. As the top layer of the multi-standard library system, the application service layer is mainly responsible for receiving user requests and providing intelligent services. It can parse the avionics architecture design model (such as SysML) in real time, quickly identify the standard conflict problems existing in the system design through integrating an advanced knowledge graph inference engine, and automatically generate a detailed conflict report and repair suggestions. In addition, this layer also has a high degree of context awareness ability, can intelligently recommend the optimal combination plan of adaptation standards according to the architecture design characteristics and actual needs of the current avionics system, and help users select appropriate avionics standards in the complex standard system.

[0070] The knowledge construction layer extracts knowledge from the original data and constructs a multi-standard library based on the knowledge graph. It includes three functional components: an entity recognition module, a relationship extraction module, and a graph update module. Among them, the entity recognition module uses the BiLSTM model to identify the key terms in the standard documents. The relationship extraction module, based on the BERT model, mines the semantic relationships between standards to ensure the semantic integrity of the knowledge graph. The graph update module combines expert knowledge and predefined rules to optimize and update the content of the multi-standard library. The knowledge construction layer is the core intelligent center of the multi-standard library system, mainly responsible for extracting and constructing a structured knowledge system from a large number of avionics standard documents. Its core functions are implemented using the above avionics knowledge graph modeling method. By combining technologies such as natural language processing (NLP) and machine learning, this layer can automatically identify and extract the key terms, concepts, and their semantic relationships in the standard documents. At the same time, the knowledge construction layer can also analyze the compatibility, dependency, and conflictuality between different standards, construct a complete avionics knowledge graph, and provide strong semantic support for the upper-layer applications.

[0071] The data acquisition layer obtains data from multiple sources and performs cleaning and integration. It consists of three major modules: the avionics standard document library, the manufacturer's technical manual, and the data cleaning component. Among them, the avionics standard document library stores the original files of multiple standards; the manufacturer's technical manual provides structured data, which includes interface definition tables and protocol descriptions; the data cleaning component uniformly converts heterogeneous data into a standardized JSON format, and the heterogeneous data includes text, tables, and charts; the data acquisition layer undertakes the task of obtaining basic data for the multi-standard library system. It widely collects and integrates various avionics standard documents (including various formats such as PDF and XML), technical manuals of aircraft equipment manufacturers, airworthiness certification cases, etc., through intelligent web crawler technology and API interfaces. To ensure the quality and usability of the data, this layer can also be equipped with a complete data cleaning and preprocessing mechanism, which can automatically identify and process problems such as duplicate data and format errors, and finally build a high-quality multi-source heterogeneous avionics standard data pool, laying a solid foundation for subsequent knowledge graph construction and application services.

[0072] The data storage layer uses a graph database to store and manage data, and the data organization form is centered on entity types, entity relationships, and entity attributes. Among them, entity types include "standard class" entities, "resource class" entities, "interface class" entities, and "security level class" entities; entity relationships cover a variety of semantic relationships; the variety of semantic relationships include "contains", "inherits", "compatible", "depends on", "conflicts", "constraints", "version evolution", and support relationship modeling between the "is-a" concept and instances; entity attributes: set corresponding attribute descriptions for different entity types. As the underlying support of the multi-standard library system, the data storage layer adopts advanced graph database technology and efficiently stores and manages structured and semi-structured data related to various avionics standards according to the "entity-relationship-attribute" data model. This layer not only supports high-concurrency access to large-scale data but also has powerful data indexing and query optimization capabilities, and can quickly respond to complex query requirements from upper-layer applications. At the same time, the data storage layer also implements data version control and historical traceability functions to ensure the integrity and consistency of data during the dynamic update and evolution of the multi-standard library.

[0073] Through this hierarchical system architecture design, the multi-standard library can not only effectively solve the problems of standard fragmentation and low-efficiency compatibility management in the avionics system but also provide intelligent support for the research and development and operation and maintenance of the avionics system, significantly improving the system development efficiency and quality management level.

[0074] Optionally, the data storage layer adopts graph database technology to store structured and semi-structured data of avionics standards, supporting high-concurrency access, complex queries, and data version control;

[0075] The data acquisition layer is responsible for the acquisition and integration of multi-source heterogeneous data. It collects the original files of various relevant avionics standards through web crawlers and API interfaces, performs data cleaning and preprocessing, and constructs a high-quality data pool.

[0076] The knowledge construction layer is responsible for extracting knowledge from avionics standard documents, constructing multi-standard libraries, and analyzing the compatibility, dependency, and conflict between standards, providing semantic support for the request processing of the application service layer.

[0077] The application service layer receives user requests and provides intelligent services, including parsing avionics architecture design models, identifying standard conflicts and generating repair suggestions, and intelligent recommending suitable standard combination solutions.

[0078] Optionally, step 3 includes:

[0079] The conflict detection and resolution mechanism is also a core functional module in the multi-standard library architecture. It mainly consists of a rule inference engine and dynamic priority arbitration, aiming to effectively identify and solve potential standard conflict problems in the avionics system design process, thereby ensuring the compliance, reliability, and security of the system design.

[0080] Rule inference engine: The rule inference engine is the key unit for conflict detection. It identifies potential conflicts in system design through a predefined set of Datalog rules. Datalog is a declarative programming language suitable for reasoning about complex relationships and pattern matching. In this system, the rule inference engine defines various conflict patterns based on Datalog rules, such as compatibility conflicts, resource allocation conflicts, interface mismatch conflicts, etc. Taking the Prolog language as an example, a typical conflict detection rule can be expressed as:

[0081] conflict(X, Y) :-

[0082] requires(X, Protocol_A),

[0083] requires(Y, Protocol_B),

[0084] incompatible(Protocol_A, Protocol_B).

[0085] This rule expresses that if two components X and Y depend on incompatible protocols Protocol_A and Protocol_B respectively, then there is a conflict between them. In this way, the rule inference engine can systematically scan the avionics architecture design model (such as the SysML model), identify all potential conflict points, and generate a detailed conflict report, including conflict types, impact ranges, and possible problems.

[0086] Dynamic Priority Arbitration: Dynamic priority arbitration is a key link in conflict resolution. Its goal is to formulate a reasonable conflict resolution strategy based on the overall requirements and constraints of the system after detecting a conflict. Dynamic priority arbitration mainly makes decisions based on the following two dimensions:

[0087] a) Standard Weight: Different avionics standards have different priorities and scopes of application. For example, the airworthiness standard DO-178C usually has a higher priority due to its strict certification requirements, while industry standards such as FACE may be given priority in specific scenarios. Dynamic priority arbitration will comprehensively evaluate the authority, applicability, and enforceability of each standard to determine its weight in the current design scenario.

[0088] b) Application Scenario: The conflict resolution strategy needs to be adjusted in combination with the specific usage scenario. For example, in a combat mission scenario, the system may pay more attention to real-time performance and safety; while in a training mission scenario, it may focus more on flexibility and cost-effectiveness. Dynamic priority arbitration will dynamically adjust the priority rules for conflict resolution based on factors such as mission requirements, performance metrics, and system configuration.

[0089] The priority weight calculation formula is: Priority = 0.6 × Airworthiness Level Coefficient + 0.3 × Mission Criticality Coefficient + 0.1 × Manufacturer Recommended Level.

[0090] Based on the above evaluation results, dynamic priority arbitration will generate specific resolution strategies, such as resolving conflict issues through standard downgrading, protocol conversion, resource reallocation, etc. For example, if it is detected that there is a conflict between the partition strategy of ARINC 653 and the resource scheduling mechanism of FACE, the arbitration mechanism may recommend adopting a more compatible resource management strategy or recommend using middleware to achieve protocol adaptation.

[0091] Through the collaborative work of the rule inference engine and dynamic priority arbitration, the conflict detection and resolution mechanism can comprehensively cover various standard conflict issues in avionics system design, providing intelligent conflict detection, diagnosis, and resolution services. This not only improves the efficiency and quality of system design but also provides strong technical support for the multi-standard collaboration of avionics systems.

[0092] Optionally, the generating of specific conflict resolution strategies includes:

[0093] 1) Standard Input and Requirement Parsing: First, the user inputs the open avionics architecture standards to be detected and their related parameters through a friendly and intuitive graphical interface or API interface. The system will perform semantic parsing and standardization processing on the input requirements to ensure the accuracy and consistency of subsequent processing.

[0094] 2) Intelligent retrieval of knowledge graph: The system relies on the pre-built knowledge graph to perform accurate semantic retrieval based on the user's input conditions. Through efficient graph query algorithms and fuzzy matching technology, it quickly screens out open standard candidate sets that meet the requirements from the massive standard library. This process fully utilizes the semantic association capabilities of the knowledge graph to ensure the comprehensiveness and relevance of the candidate set.

[0095] 3) Intelligent conflict detection and priority sorting: The system calls the built-in conflict detection service, combines expert experience rules and historical case data, and conducts a comprehensive conflict analysis of the candidate standard set. The detection content includes but is not limited to potential problems such as version conflicts, functional overlaps, and technical incompatibility. After the detection is completed, the system will intelligently sort the conflicting items according to the preset priority rules (such as standard version level, scope of application, industry recognition, etc.), and finally generate an optimal open avionics architecture applicable standard set.

[0096] 4) Standard set verification and knowledge graph update: In order to ensure the reliability of the output results, the system will execute the verification unit to perform multi-dimensional verification on the generated avionics standard set, including logical consistency check, functional integrity assessment, and actual application scenario simulation. After the verification is passed, the system will feed back the new standard information and conflict resolution results to the multi-standard library, complete the dynamic update of the multi-standard library, and continuously optimize the system's intelligence level and service capabilities.

[0097] Through the above steps, the conflict detection and resolution method proposed in this application not only realizes the intelligent management of open avionics architecture standards, but also significantly improves the adaptability and scalability of the system. This method is suitable for complex and changeable avionics system design scenarios, can effectively reduce the uncertainty risk of standard selection, and provides strong technical support for the flexibility and scalability of the system.

[0098] Optionally, the process of determining technical incompatibility includes:

[0099] 1) Knowledge representation learning: First, all entities and relations in the knowledge graph are trained to generate corresponding low-dimensional vector representations.

[0100] 2) Entity relationship modeling: In the vector space, the TransE algorithm expresses the relationship between entities through vector operations. Specifically, given a triple (h, r, t), the algorithm hopes to satisfy the vector relationship h+r≈t. This concise and effective modeling method can well preserve the relationship information between entities.

[0101] 3) Similarity measurement: By calculating operations such as distance or dot product between vectors, the semantic similarity between different entities can be quantitatively evaluated.

[0102] Through such a graph embedding method, not only can the relationship between entities be quantified, but also strong support can be provided for complex system design and interoperability analysis.

[0103] See also Figure 1 , this application also provides a more specific embodiment:

[0104] 1) Application service layer:

[0105] This layer mainly provides intelligent service functions for users, including three core components: conflict detection service, compatibility recommendation engine and user interface. Among them, the conflict detection service component can automatically identify the conflict points between avionics standards through Datalog rule reasoning technology, and output detailed conflict reports, such as conflicts such as "protocol mismatch". The compatibility recommendation engine component calculates the similarity between standards based on graph embedding technology (TransE algorithm), and generates the optimal adaptation plan through in-depth analysis of the semantic relationship between standards, such as the combination recommendation of communication protocol and security certification process. The user interface component provides an intuitive interactive query function and supports natural language input, such as "communication protocol that complies with FACE 3.1", which greatly improves the convenience of user experience.

[0106] 2) Knowledge construction layer:

[0107] This layer is responsible for extracting knowledge from raw data and building a multi-standard library. It includes three functional components: entity recognition module, relationship extraction module, and graph update module. Among them, the entity recognition module uses the BiLSTM model, which can accurately identify key terms in standard documents, such as "partitioning mechanism", etc., laying the foundation for the construction of the knowledge graph. The relationship extraction module is based on the BERT model and deeply explores the semantic relationships between standards, such as "dependency" and "conflict", to ensure the semantic integrity of the knowledge graph. The graph update module combines expert knowledge and predefined rules to continuously optimize and update the content of the multi-standard library to ensure its dynamicity and accuracy.

[0108] 3) Data collection layer:

[0109] This layer is responsible for acquiring data from multiple sources, cleaning and integrating it. It mainly includes three modules: avionics standard document library, manufacturer technical manuals, and data cleaning components. Among them, the avionics standard document library stores the original files of mainstream standards such as DO-178C, FACE, ARINC653, etc. (supports multiple formats such as PDF, XML, WORD, etc.). The manufacturer's technical manual provides structured data such as interface definition tables and protocol descriptions (stored in the form of database tables). The data cleaning component uses intelligent extraction and conversion technology to uniformly convert heterogeneous data such as text, tables, and charts into a standardized JSON format to ensure data consistency and availability.

[0110] 4) Data storage layer:

[0111] This layer uses a graph database (Neo4J / JanusGraph) to efficiently store and manage data. The data organization form is centered around entities, relationships, and attributes. Among them, entity types include "standard class" entities (such as DO-178C, ARINC 653, FACE, SOSA, etc.), "resource class" entities (such as communication, navigation, radar, optoelectronics, health management, etc.), "interface class" entities (such as API, protocol, data format, etc.), and "security level class" entities (such as DAL A~D, classification level, etc.). Entity relationships cover various semantic relationships such as "contains", "inherits", "compatible", "depends on", "conflicts with", "constrains", "version evolution", etc., and support relationship modeling between the "is-a" concept and instances. Entity attributes are set with corresponding attribute descriptions for different entity types. For example, "standard class" entities include attributes such as "release date", "scope of application", "airworthiness level", etc., "interface class" entities include attributes such as "data type", "protocol type", "API type", etc., "resource class" entities include attributes such as "computing power", "bandwidth", "storage capacity", etc., and "security level class" entities include classification information such as "A / B / C / D".

[0112] See Figure 2 , the main steps of the graph embedding method proposed in this application are as follows:

[0113] In the constructed knowledge graph, there are the following three triples: (SOSA 3.0, inherits, SOSA2.0), (SOSA2.0, contains, FACE 2.0), and (SOSA 3.0, contains, ARINC653). The goal is to calculate the compatibility between SOSA 3.0 and SOSA2.0.

[0114] Step 1: Vector initialization and training: Use the TransE algorithm to generate low-dimensional vectors for each entity and relationship. The initial vectors are randomly generated as follows:

[0115] Entity vectors:

[0116] eSOSA3.0 = [0.8, −0.2, 1.0]

[0117] eSOSA2.0 = [0.5, 0.3, 0.7]

[0118] eFACE2.0 = [0.1, −0.4, 0.6]

[0119] eARINC653 = [0.9, 0.0, 0.5]

[0120] Relationship vector:

[0121] r_inheritance = [-0.1, 0.5, -0.3]

[0122] r_contains = [0.2, -0.2, 0.4]

[0123] The optimization goal of TransE is to make the positive triple satisfy h + r ≈ t. For example, for the triple (SOSA3.0, inheritance, SOSA2.0): eSOSA3.0 + r_inheritance = [0.8, -0.2, 1.0] + [-0.1, 0.5, -0.3] = [0.7, 0.3, 0.7]. Ideally, the result should be close to eSOSA2.0 = [0.5, 0.3, 0.7]. By calculating the Euclidean distance and optimizing the model parameters, the vectors are gradually adjusted to minimize the error.

[0124] Step 2: Calculate similarity

[0125] After training, the entity vectors are updated as follows:

[0126] eSOSA3.0 = [0.7, 0.1, 0.9]

[0127] eSOSA2.0 = [0.6, 0.4, 0.8]

[0128] Use cosine similarity to calculate the similarity between the two: sim(SOSA3.0, SOSA2.0) ≈ 0.95

[0129] Step 3: Result analysis

[0130] The high cosine similarity (0.95) indicates that SOSA3.0 and SOSA2.0 have almost the same direction in the vector space, which may mean they have high compatibility. This similarity reflects the close semantic association between the two, and may imply that in practical applications, SOSA3.0 can smoothly replace or work in cooperation with SOSA2.0, reducing the risks brought by compatibility issues.

[0131] See Figure 3 , the conflict detection and resolution mechanism proposed in this application is shown in the following steps:

[0132] Taking the design of the airborne mission system as an example:

[0133] 1) Input the open avionics architecture standard to be detected through the user interface. The user input requirements are as follows: "Design a navigation module that complies with FACE 3.1 and DO-178C DAL C";

[0134] 2) The knowledge graph queries the candidate set of open standards that meet the conditions according to the user input conditions;

[0135] 3) After systematically retrieving the knowledge graph, recommend compatible communication protocols (such as UDP), partition configurations (ARINC 653 time partition ≥ 50 ms), and verification toolchains (LDRA TBvision);

[0136] 4) Perform rule matching and priority sorting through the conflict monitoring service;

[0137] 5) Detect potential conflicts between "FACE requires the use of POSIX threads" and "DO-178C restricts dynamic memory allocation", and recommend replacing them with a static memory management strategy;

[0138] 6) Execute verification, update the results to the knowledge graph, and end the process after passing the verification.

[0139] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them. Although the present application has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: it is still possible to modify the specific implementation manners of the present application or make equivalent replacements, and any modification or equivalent replacement that does not depart from the spirit and scope of the present application shall be covered by the protection scope of the claims of the present application.

Claims

1. A conflict resolution method for an open avionics architecture, characterized in that Including: Step 1: Construct a standardized ontology model in the avionics field, including entity types, semantic relationships, and constraint rules; entity types include standard classes, resource classes, interface classes, and security classes; semantic relationships include inclusion, inheritance, compatibility, dependence, conflict, constraint, and version evolution; constraint rules include version consistency, functional interoperability, and performance metric constraints; Step 2: Construct a multi-standard library based on a knowledge graph; Step 3: Through a conflict detection and resolution mechanism, identify and eliminate possible standard conflicts in the process of open avionics architecture design; The said Step 1 includes: Constructing a standardized ontology model in the avionics field includes multi-standard semantic extraction and the construction and update of a multi-standard library; among them, multi-standard semantic extraction includes defining the ontology structure of the avionics field; the ontology structure of the avionics field includes standard entities, relationship types between entities, and constraint rules; standard entities include APIs, communication protocols, and security levels; relationship types between entities include compatibility, dependence, and conflict; constraint rules include timing restrictions and resource allocation restrictions; Extract triple information from relevant avionics standard documents based on NLP technology to achieve the extraction of explicit and implicit semantic associations; triple information includes subjects, predicates, and objects; Update the multi-standard library by combining expert knowledge and predefined rules, and incrementally add the key information of the new standard document to the multi-standard library; The said Step 2 includes: The multi-standard library adopts a hierarchical design, which includes an application service layer, a knowledge construction layer, a data collection layer, and a data storage layer; The application service layer provides intelligent service functions for users, which includes a conflict detection service component, a compatibility recommendation engine component, and a user interface component; among them, the conflict detection service component can automatically identify the conflict points between avionics standards through Datalog rule inference technology and output a detailed conflict report; the compatibility recommendation engine component calculates the similarity between standards based on graph embedding technology and generates an optimal adaptation plan through the analysis of the semantic relationships between standards; the user interface component provides an interactive query function and supports natural language input; The knowledge construction layer extracts knowledge from the original data and constructs a multi-standard library based on a knowledge graph, which includes three functional components: an entity recognition module, a relationship extraction module, and a graph update module; among them, the entity recognition module uses a BiLSTM model to identify the key terms in the standard document; the relationship extraction module is based on the BERT model to mine the semantic relationships between standards to ensure the semantic integrity of the knowledge graph; the graph update module combines expert knowledge and predefined rules to optimize and update the content of the multi-standard library; The data collection layer obtains data from multiple sources, cleans and integrates it, and includes three modules: an avionics standard document library, a manufacturer's technical manual, and a data cleaning component; among them, the avionics standard document library stores the original files of multiple standards; the manufacturer's technical manual provides structured data, and the structured data includes interface definition tables and protocol descriptions; the data cleaning component uniformly converts heterogeneous data into a standardized JSON format, and the heterogeneous data includes text, tables, and charts; The data storage layer uses a graph database to store and manage data, with the data organization form centered on entity types, entity relationships, and entity attributes. Among them, entity types include "standard class" entities, "resource class" entities, "interface class" entities, and "security level class" entities; entity relationships cover various semantic relationships; the various semantic relationships include "include", "inherit", "compatible", "depend on", "conflict", "constraint", "version evolution", and support the relationship modeling between the "is-a" concept and instances; entity attributes: set corresponding attribute descriptions for different entity types. The said step 3 includes: The conflict detection and resolution mechanism consists of a rule inference engine and a dynamic priority arbitration. The rule inference engine identifies conflicts based on the Datalog rule set and supports various types of conflict detection; the various types of conflicts include compatibility conflicts, resource allocation conflicts, and interface mismatch conflicts; the dynamic priority arbitration formulates a resolution strategy according to the standard weights and application scenarios.

2. The conflict resolution method for an open avionics architecture according to claim 1, characterized in that The construction and update of the said multi-standard library include: Extract semantic information from multi-source heterogeneous avionics standard documents to generate an extensible knowledge graph and construct a multi-standard library: the multi-standard library extracts key terms, concepts, and their semantic relationships from avionics standard documents of various heterogeneous sources; the avionics standard documents of various heterogeneous sources include international standards, industry standards, domain standards, and manufacturer technical manuals. Implement the mining of implicit knowledge in standard documents through natural language processing technology. At the same time, the multi-standard library supports the unified processing of structured and unstructured data, and through data cleaning and format conversion, converts them into a standardized knowledge representation form; structured data includes interface definition tables and protocol descriptions, and unstructured data includes PDF and WORD documents. Based on the standardized knowledge representation form, generate a knowledge graph and construct a multi-standard library to dynamically reflect the latest developments of avionics standards and support the seamless integration of new standards. When a new avionics standard is released or the original standard is iteratively updated, the key information in the new standard document is incrementally added to the existing multi-standard library by combining expert knowledge with predefined rules.

3. The conflict resolution method for an open avionics architecture according to claim 1, wherein The data storage layer uses graph database technology to store the structured and semi-structured data of avionics standards, supporting high-concurrency access, complex queries, and data version control. The data acquisition layer is responsible for the acquisition and integration of multi-source heterogeneous data, collects the original files of various relevant avionics standards through web crawlers and API interfaces, and performs data cleaning and preprocessing to build a high-quality data pool. The knowledge construction layer is responsible for extracting knowledge from avionics standard documents, constructing a multi-standard library, and analyzing the compatibility, dependency, and conflictiveness between standards to provide semantic support for the request processing of the application service layer. The application service layer receives user requests and provides intelligent services, including parsing the avionics architecture design model, identifying standard conflicts and generating repair suggestions, and intelligent recommending an adapted standard combination plan.

4. The conflict resolution method for an open avionics architecture according to claim 1, characterized in that The conflict detection and resolution mechanism introduces a dynamic priority strategy, and the priority weight calculation formula is: Priority = k × airworthiness level coefficient + p × task criticality coefficient + q × manufacturer recommended level; the priority strategy is related to the priority weight; the value ranges of k, p, and q are all from 0 to 1, p is greater than q and less than k.

5. A conflict resolution method for an open avionics architecture according to claim 1, characterized in that The dynamic priority arbitration makes decisions based on two dimensions: standard weight and application scenario: Standard weight: Different avionics standards have different priorities and application scopes; the dynamic priority arbitration comprehensively evaluates the authority, applicability, and enforceability of each standard to determine its weight in the current design scenario; Application scenario: The conflict resolution strategy is adjusted in combination with the specific usage scenario; The dynamic priority arbitration dynamically adjusts the priority rules for conflict resolution according to relevant factors; the relevant factors include task requirements, performance indicators, and system configuration; The dynamic priority arbitration generates specific conflict resolution strategies based on the conflict detection results output by rule reasoning.

6. The conflict resolution method for an open avionics architecture according to claim 5, characterized in that The generation of specific conflict resolution strategies includes: Standard input and requirement parsing: The user inputs the open avionics architecture standards to be detected and their related parameters through a graphical interface or API interface, and the system performs semantic parsing and standardization processing on the input requirements; Knowledge graph intelligent retrieval: Utilize the semantic association ability of the knowledge graph to screen the standard candidate set that meets the requirements according to the user's input conditions; Intelligent conflict detection and priority sorting: The system calls the built-in conflict detection service, combines expert experience rules and historical case data to conduct comprehensive conflict detection on the standard candidate set; the content of conflict detection includes version conflict, function overlap, and technical incompatibility; after the conflict detection is completed, the system performs intelligent sorting on the conflict items according to the preset priority rules, and finally generates the optimal applicable standard set for the open avionics architecture; the priority rules include standard version level, application scope, and industry recognition; Standard set verification and multi-standard library update: Conduct multi-dimensional verification on the generated optimal applicable standard set for the open avionics architecture. After the verification is passed, the system feeds back the new standard information and conflict resolution results to the multi-standard library to complete the dynamic update of the multi-standard library; the multi-dimensional verification includes logical consistency check, function integrity evaluation, and actual application scenario simulation.

7. A conflict resolution method for an open avionics architecture according to claim 6, characterized in that, The judgment process of the technical incompatibility includes: Knowledge representation learning: Train all entities and relationships in the knowledge graph to generate corresponding low-dimensional vector representations; Entity relationship modeling: In the vector space, the TransE algorithm expresses the relationship between entities through vector operations; Similarity measurement: Quantitatively evaluate the semantic similarity between different entities by calculating the metrics between vectors; the metrics include distance or dot product; Compatibility judgment: If the semantic similarity output by the similarity measurement is lower than the preset value, it is considered that the standards are incompatible.

Images