CN-121979491-A - Space software demand dynamic management system based on atlas and big data technology

CN121979491ACN 121979491 ACN121979491 ACN 121979491ACN-121979491-A

Abstract

The invention relates to a space software demand dynamic management system based on a graph and big data technology, which integrates a knowledge graph and a big data technology solution, generates normalized and structured demand items by automatically analyzing original information such as task demands, system design scheme reports, subsystem design reports, various analysis reports and the like, builds a demand association network by using the knowledge graph technology, realizes traceability and dynamic change of demand item sources, and guarantees space information safety by adopting an offline processing mode. Through preliminary verification, the scheme can shorten the processing period of the required information by more than 60%, the structuring rate of the key information is improved to 95%, and technical support is provided for the dynamic management of the requirements of the spaceflight software.

Inventors

PENG LIZHANG
FAN RONG
LUO BIN
LI JIANGDAO
WANG HUA
LIU ZHIJUAN
JIANG CHENGFU
SHI LI

Assignees

上海宇航系统工程研究所

Dates

Publication Date: 20260505
Application Date: 20251223

Claims (12)

1. A space software demand dynamic management system based on a map and big data technology is characterized by comprising a data layer, a technical layer, a functional layer and an application layer; The data layer is responsible for storing multi-source demand information and map data, wherein the multi-source demand information comprises an original document library and a structured demand library; The technical layer comprises a big data analysis module and a knowledge graph management module, wherein the big data analysis module realizes the automatic analysis of multi-source unstructured demand information; the functional layer provides a demand information acquisition and processing function, a demand map construction and query function, a demand dynamic change management function and a demand statistics and analysis function for a user according to the big data analysis module and the knowledge map management module used for the demand calling technical layer; And an application layer for providing application services for developing different roles.
2. The space software requirement dynamic management system based on the map and big data technology of claim 1, wherein the original document library stores the original documents of task requirements, system design scheme reports, subsystem design reports and various analysis reports in an encrypted format, and sets access rights; the structured demand library stores normalized demand entries processed by the big data analysis module; The knowledge graph base stores required knowledge graph data based on a graph database, wherein the knowledge graph data comprises an entity table, a relation table and an attribute table, and supports graph query and association analysis.
3. The space software demand dynamic management system based on the map and big data technology of claim 1, wherein the big data analysis module comprises a data acquisition and preprocessing sub-module, an unstructured information structured conversion sub-module, a demand quality multi-dimensional verification detection sub-module and a structured item output and butt joint sub-module; The data acquisition and preprocessing sub-module is used for accessing a multi-source demand document from the data layer, wherein the multi-source demand document comprises a task demand, a system design report, a subsystem design report and an analysis report; The unstructured information structured conversion sub-module is used for constructing an aerospace demand entity recognition model based on a natural language processing technology, automatically extracting key entities in a demand document processed by the data acquisition and preprocessing sub-module, recognizing the association relationship among the entities and forming a structured demand item, wherein the key entities comprise a demand item name, an index requirement, constraint conditions, a responsibility department and delivery time; The demand quality multi-dimensional verification detection sub-module is used for verifying and detecting the integrity, consistency and rationality of the structured demand items, automatically marking abnormal items and prompting manual auditing And the structured item output and butt joint sub-module is used for auditing and correcting the structured demand items checked and detected by the demand quality multi-dimensional checking and detecting sub-module, outputting the checked and corrected structured demand items to the subsequent modules, and simultaneously reserving a manual intervention interface.
4. The space software demand dynamic management system based on the map and big data technology of claim 3, wherein the data acquisition and preprocessing submodule performs format standardization processing on the received multi-source demand document, removes redundant information in the document and unifies formats and terms, and is specifically as follows: (1) Format normalization of received multi-source documents Extracting text content in a document by adopting an optical character recognition technology aiming at a PDF document, and extracting editable text by combining a PDF analysis tool; Extracting title, paragraph and table data, and converting the table data into a 'line-column' structured format; Extracting formula content by adopting a mathematical formula identification tool, and converting data points in the diagram into a numerical sequence; (2) Deleting redundant information, including header footers, watermarks, blank sections and repeated expression in the document; (3) Based on the term library in the field of aerospace requirements, the nonstandard expression is automatically replaced.
5. The space software demand dynamic management system based on the map and big data technology as claimed in claim 3, wherein the unstructured information structured transformation submodule forms a structured demand entry, and the method is specifically as follows: Constructing a space demand entity recognition model by using the BERT pre-training model, and training the space demand entity recognition model by taking historical demand data in the space field as a training set, wherein the space demand entity recognition model is input into a demand document processed by a data acquisition and preprocessing sub-module segmented according to paragraphs, and the entities in the document are output; obtaining an entity according to the input demand document by utilizing the trained aerospace demand entity identification model; extracting relation conditions by adopting a BERT-RE model according to the complex expression relation among the entities; The "entity-relationship" is combined into a structured demand entry in a space demand standard format.
6. The space software demand dynamic management system based on the map and big data technology as claimed in claim 3, wherein the demand quality multidimensional checking and detecting submodule checks and detects the integrity, consistency and rationality of the structured demand items, and the system is characterized by comprising the following specific steps: Checking whether the necessary filling field of the structured requirement item is missing, marking the item of the missing field as incomplete, and prompting a possible supplement source; Consistency detection, namely judging whether expressions of the same requirement item in different documents are consistent or not; Rationality detection, namely judging whether the logic relationship among the item indexes is rational or not based on historical data and the demand indexes of the similar aerospace models; According to the detection, a 'demand quality detection report' is output, wherein the 'demand quality detection report' comprises qualified items and problem items, and the problem items are classified and counted according to the 'severity', wherein the serious problems comprise serious problems, general problems and acceptable problems, the serious problems comprise the fact that the existing core field is missing and the key index is conflicted, the problem needs to be immediately modified, the general problems comprise the fact that the existing term is irregular, the index rationality is to be confirmed, the problem needs to be checked and confirmed, and the acceptable problems comprise the fact that the units are non-uniform and the non-core field is missing, and the problem is reserved for subsequent optimization.
7. The space software demand dynamic management system based on the map and big data technology of claim 6, wherein the structured item output and docking submodule is specifically implemented as follows: the demand engineer carries out manual correction on marked problem items through a demand quality detection report; the structured item output and butt joint sub-module synchronizes the detected qualified structured items and the corrected detected qualified structured items to a knowledge-graph library through an offline interface, and provides data support for the requirement association management; The structured item output and docking submodule automatically records the whole process log of extraction, detection, auditing and output of each item, including operators, operation time and modified content, and ensures traceability.
8. The aerospace software demand dynamic management system based on the graph and big data technology of claim 1, wherein the knowledge graph analysis module comprises a demand knowledge graph construction sub-module, a demand source tracing sub-module and a demand dynamic change management sub-module; The demand knowledge graph construction module is used for defining core entities, relations and attributes of the aerospace software demand field and constructing a field ontology model, converting the structured demand items generated by the big data analysis module into triplet data, and storing the triplet data in a graph database to form a demand knowledge graph; the demand source tracing sub-module is used for rapidly positioning source documents of any demand items, extracting chapters and related responsible persons through map query sentences; and the demand dynamic change management sub-module automatically records change nodes in the map when the demand is changed, wherein the change nodes comprise change reasons, change contents, change time and approvers, and analyzes change influence ranges through a relation path so as to realize traceability and controllability of the change.
9. The space software demand dynamic management system based on the map and big data technology as claimed in claim 1, wherein the demand knowledge map modeling submodule comprises a space software demand knowledge map design unit, a demand data access and preprocessing unit and a knowledge map construction unit; the space software demand knowledge graph design unit is used for defining entities, relations, attributes and constraint rules of the graph to form a standardized graph mode; the demand data access and preprocessing unit is used for extracting structured demand items from the big data analysis module; the knowledge graph construction unit is used for converting the extracted structured demand items into 'entity-relation-attribute' triples, and importing the 'entity-relation-attribute' triples into a graph database to form a complete demand knowledge graph; The knowledge graph construction submodule comprises: The entity construction subunit is used for creating map entities in batches according to entity types; The relation construction subunit is used for constructing the association relation between the entities in a 'rule matching + manual complement' mode based on the relation type defined by the structured demand items and the pattern mode; an automated relation construction subunit, for the relation with clear data association, creating the association relation in batches through SQL inquiry and Cypher sentences, and for the relation of logic judgment, triggering automatic matching based on a rule base; And the manual complement and audit subunit is used for providing an interactive interface for the complex relation of the automatic construction failure, and the requirement engineer and the test engineer are used for collaborative complement.
10. The aerospace software-demand dynamic management system based on the atlas and big data technology of claim 9, wherein the knowledge atlas construction unit further comprises an atlas quality checking and optimizing subunit; The map quality checking and optimizing subunit: counting the number of entities without association relation to the constructed knowledge graph through Cypher query, and if yes, marking and triggering manual processing; Checking the attribute consistency and the relation logic consistency of the same entity, if the attribute consistency and the relation logic consistency are inconsistent, marking abnormality and triggering manual processing; And deleting the duplicate entity and invalid relationship.
11. The space software demand dynamic management system based on the map and big data technology of claim 1, wherein the function layer provides the function service for the user, and the system is characterized in that: The requirement information acquisition and processing function is that the multi-source original document in the data layer is supported to be imported, and a big data analysis module is called to automatically complete structural conversion and quality detection, so that a normalized requirement item is generated; The demand pattern constructing and inquiring function is that a knowledge pattern management module is called to automatically construct a demand knowledge pattern, a demand source tracing and related inquiring function is provided, and related information can be rapidly positioned through a demand ID, a name or a source document; And the requirement dynamic change management function is used for recording requirement change history, analyzing change influence range, generating a change report and supporting change approval process management. And the demand statistics and analysis function is used for counting the number, type, state and change times of the demands, generating a visual report and providing support for developing decisions.
12. The space software demand dynamic management system based on the map and big data technology of claim 1, wherein the application layer provides application services for developing different roles, and comprises: providing services for checking the overall progress of the demand, changing statistical reports, grasping the demand management state and making a development plan for development managers; A demand engineer is oriented to complete the generation of demand items, the quality audit and the change application by using a demand information acquisition and processing function and a demand dynamic change management function; And the method is oriented to software designers and test engineers, and accurate demand information is acquired through a demand pattern construction and query function, software design and test case writing are guided, and the influence of demand change on design and test is known.

Description

Space software demand dynamic management system based on atlas and big data technology Technical Field The invention belongs to the technical field of software design, and relates to a space software demand dynamic management system based on a map and big data technology. Background The space software is used as a core control unit of space equipment, and the accuracy and the dynamics of demand management directly determine the equipment development quality and the task success rate. In the early development and in the process of the current aerospace model, the management of the demand information still depends on manual collection and arrangement, and the problems of long period, non-uniform standard, unstructured key information, low utilization rate, difficult dynamic change tracing and the like exist. Based on the research background, aiming at the requirements of the space software management, research on the space software requirement dynamic management based on the atlas and big data technology is developed, and technical support is provided for the space software requirement dynamic management through the application of the big data technology and the knowledge atlas. Disclosure of Invention Aiming at the current situation that the space software needs are multi-source, dynamic and complex, the invention overcomes the defects of the prior art and provides a space software needs dynamic management system based on a map and big data technology. The solution of the invention is as follows: A space software demand dynamic management system based on a map and big data technology comprises a data layer, a technical layer, a functional layer and an application layer; The data layer is responsible for storing multi-source demand information and map data, wherein the multi-source demand information comprises an original document library and a structured demand library; The technical layer comprises a big data analysis module and a knowledge graph management module, wherein the big data analysis module realizes the automatic analysis of multi-source unstructured demand information; the functional layer provides a demand information acquisition and processing function, a demand map construction and query function, a demand dynamic change management function and a demand statistics and analysis function for a user according to the big data analysis module and the knowledge map management module used for the demand calling technical layer; And an application layer for providing application services for developing different roles. Preferably, the original document library stores the original documents of task demands, system design scheme reports, subsystem design reports and various analysis reports in an encryption format, and sets access rights; the structured demand library stores normalized demand entries processed by the big data analysis module; The knowledge graph base stores required knowledge graph data based on a graph database, wherein the knowledge graph data comprises an entity table, a relation table and an attribute table, and supports graph query and association analysis. Preferably, the big data analysis module comprises a data acquisition and preprocessing sub-module, an unstructured information structuring conversion sub-module, a required quality multi-dimensional checking detection sub-module and a structuring item output and butt joint sub-module; The data acquisition and preprocessing sub-module is used for accessing a multi-source demand document from the data layer, wherein the multi-source demand document comprises a task demand, a system design report, a subsystem design report and an analysis report; The unstructured information structured conversion sub-module is used for constructing an aerospace demand entity recognition model based on a natural language processing technology, automatically extracting key entities in a demand document processed by the data acquisition and preprocessing sub-module, recognizing the association relationship among the entities and forming a structured demand item, wherein the key entities comprise a demand item name, an index requirement, constraint conditions, a responsibility department and delivery time; The demand quality multi-dimensional verification detection sub-module is used for verifying and detecting the integrity, consistency and rationality of the structured demand items, automatically marking abnormal items and prompting manual auditing And the structured item output and butt joint sub-module is used for auditing and correcting the structured demand items checked and detected by the demand quality multi-dimensional checking and detecting sub-module, outputting the checked and corrected structured demand items to the subsequent modules, and simultaneously reserving a manual intervention interface. Preferably, the data acquisition and preprocessing submodule performs format standardization processing on the received multisource requiremen