Unmanned aerial vehicle ground station field fault data knowledge graph construction method

Abstract

The invention discloses an unmanned aerial vehicle ground station field fault data knowledge graph construction method, and the method comprises the steps: sorting and analyzing unmanned aerial vehicle ground station fault ontology classification according to expert knowledge, and carrying out the step-by-step subdivision of faults in each specialty until the granularity reaches a fault determinable level, thereby forming an unmanned aerial vehicle ground station field fault ontology model; adding a fault level label and a fault type label to the fault body; the method comprises the following steps: establishing an unmanned aerial vehicle ground station fault knowledge graph through fault data import, extraction and knowledge fusion, entity-relationship-entity gt; and lt; entity-attribute value-attribute gt; extracting in two modes; and according to the unmanned aerial vehicle ground station fault knowledge graph, entities, relationships and attributes in the graph are stored and updated. According to the method, the problems of low fault data utilization rate, difficulty in troubleshooting and positioning and the like in the field are solved, and the fault discovery, early warning and troubleshooting capabilities are improved.

Description

Technical Field

[0001] The present invention relates to the field of UAV ground stations, and in particular to the field of UAV ground station fault handling, and is a method for constructing a fault knowledge graph for the UAV ground station field. Background Art

[0002] Knowledge graph technology is one of the important branches of artificial intelligence. It is divided into general knowledge graph and industry knowledge graph. It has been widely used in search, medicine, health, finance and other fields, providing intelligent support for intelligent question and answer, risk prediction, etc.

[0003] With the development of UAV systems, the scale of the system is getting larger and larger. Most of the data generated by each subsystem will be collected in the UAV ground station, among which fault data is extremely important. Failures in UAV systems may cause mission failure, UAV crashes, and even casualties of unrelated personnel on the ground. Effective management and utilization of fault data are of great significance for fault analysis, troubleshooting, positioning, and task review. In the current UAV ground station system, system failures are mainly displayed as prompts, and there is a lack of necessary analysis of factors such as the cause and location of the fault. Fault troubleshooting and positioning are difficult, and the troubleshooting cycle is long, which affects the progress of the project. In fault analysis, expert knowledge is a very important type of empirical knowledge, which has a very important impact on the accurate positioning of faults. Traditional expert systems mainly use inference engines, knowledge bases, and comprehensive databases to perform data analysis, and the semantic relationship between faults is less utilized.

[0004] In different UAV ground station systems, the failures of related subsystems have certain correlations, but in existing models of equipment, the use of fault data cannot be interoperable, and related expert knowledge cannot be transferred and used. There is a lack of more efficient use of the accumulated fault troubleshooting methods, means, positioning results, solutions, etc. In order to improve the efficiency of fault location troubleshooting and data analysis, the semantic information between fault data in the field of UAV ground stations can be fully utilized. The present invention performs ontology modeling and entity relationship extraction on the fault data of each subsystem, builds a knowledge graph of faults in the field of UAV ground stations, and improves the intelligent level of fault analysis and positioning in the field of UAV ground stations. Summary of the invention

[0005] The purpose of the invention is to provide a method for constructing a knowledge graph of fault data in the field of UAV ground stations, so as to solve the problems of low utilization rate of fault data in this field, difficulty in fault troubleshooting and locating, and improve the fault discovery, warning and troubleshooting capabilities.

[0006] The object of the present invention is achieved through the following technical solutions:

[0007] A method for constructing a knowledge graph of fault data in the field of UAV ground stations, comprising the following steps:

[0008] Step 1: Based on expert knowledge, sort out and analyze the classification of UAV ground station fault ontology, and subdivide the faults in each profession step by step until the granularity reaches the level at which the faults can be determined, thereby forming a fault ontology model in the field of UAV ground stations, and adding fault level labels and fault type labels to the fault ontology;

[0009] Step 2: Based on the UAV ground station domain ontology model, a UAV ground station fault knowledge graph is established, including the import of fault data, fault data extraction and knowledge fusion of fault data; wherein, data extraction adopts two modes of <entity-relationship-entity> and <entity-attribute value-attribute>;

[0010] Step 3: According to the UAV ground station fault knowledge graph, the entities, relationships, and attributes in the graph are stored, and the stored UAV ground station fault knowledge graph is continuously updated based on the import, extraction, and fusion of fault data.

[0011] Preferably, in step 2, the imported fault data includes fault information in the equipment manual, fault data and solutions in the field of drone ground stations on the Internet, and records of faults during the use of the drone ground station.

[0012] Preferably, in step 2, the imported fault data is processed by natural language to extract knowledge entities, entity attributes and entity relationships to form <entity-relationship-entity> and <entity-attribute value-attribute> relationships, and entity alignment and relationship alignment are performed on the extracted data.

[0013] Preferably, the UAV ground station fault knowledge graph includes:

[0014] The query module is used to query the cause of the fault according to the fault phenomenon and fault level, analyze the fault principle, and find a solution; or obtain fault prediction information and avoidance methods based on the value change information.

[0015] The beneficial effects of the present invention are:

[0016] 1. The present invention establishes an ontology model in the field of UAV ground stations, classifies fault types and levels, realizes data unification, provides a unified ontology model for the establishment of fault knowledge graphs, and effectively breaks the data barriers between different UAV ground stations;

[0017] 2. The present invention makes full use of expert knowledge in the field of UAV ground stations and the semantic connections between faults, changes the original fault analysis and positioning methods, and provides a more efficient solution for fault prediction and troubleshooting. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1Flowchart for constructing the knowledge graph of UAV ground station faults in the present invention.

[0019] Figure 2 The basic data acquisition process of the UAV ground station fault knowledge graph in the present invention. DETAILED DESCRIPTION

[0020] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments.

[0021] See also Figure 1 As shown, the method for constructing a knowledge graph of fault data in the field of drone ground stations shown in this embodiment, by modeling the fault data in the field of drone ground stations, using various types of fault data, monitoring and analyzing the data in the drone ground station, helps to assist ground station users to timely discover faults, troubleshoot fault causes, analyze equipment health status, etc., and improve the intelligence level of drone ground stations. It includes the following steps:

[0022] Step 1: Based on expert knowledge, sort out and analyze the classification of UAV ground station fault ontology, and subdivide the faults in each profession step by step until the granularity reaches the level at which the fault can be determined, thereby forming a UAV ground station field fault ontology model. As an example, this embodiment divides the UAV ground station field fault ontology into UAV airborne system faults, link system faults, and ground control system faults, divides the UAV airborne system fault ontology into electromechanical faults, avionics faults, power faults, and fuselage faults, divides the link system fault ontology into line-of-sight link faults and satellite link faults, and divides the ground control system fault ontology into flight control faults, mission control faults, link control faults, and ground support faults.

[0023] According to the UAV ground station domain ontology model, a fault level label and a fault type label are added to the fault ontology, and the fault level label and the fault type label are stored separately.

[0024] Step 2: Based on the UAV ground station domain ontology model, a UAV ground station fault knowledge graph is established, including the import of fault data, fault data extraction and knowledge fusion of fault data.

[0025] The imported fault data covers Figure 2 As shown:

[0026] Including fault information in the equipment manual. This type of information is generally semi-structured data, including the fault phenomenon and fault solution;

[0027] The Internet contains information about the failures and solutions in the field of drone ground stations. This type of information is generally unstructured or semi-structured data.

[0028] Records of faults during the use of UAV ground stations, such as fault phenomena, fault location, fault analysis, etc., are generally unstructured data.

[0029] The imported fault data is processed through natural language processing to extract knowledge entities, entity attributes and entity relationships, forming <entity-relationship-entity> and <entity-attribute value-attribute> relationships, and the extracted data is aligned in entities and relationships.

[0030] The above fault data from different data sources are integrated and disambiguated under the same specification, such as OS (Operation System) and operating system are the same entity;

[0031] Step 3: Based on the UAV ground station fault knowledge graph, a graph database is used to permanently store the entities, relationships, and attributes in the graph. Based on the fault data import, extraction, and fusion, the stored UAV ground station fault knowledge graph is continuously updated.

[0032] Optionally, the UAV ground station fault knowledge graph includes:

[0033] The query module is used to query the cause of the fault according to the fault phenomenon and fault level, analyze the fault principle, and find a solution; or obtain fault prediction information and avoidance methods based on the value change information.

[0034] It should be noted that the technical solutions not described in detail in the present invention adopt well-known technologies.

[0035] It is understandable that those skilled in the art can make equivalent substitutions or changes based on the technical solution and inventive concept of the present invention, and all these changes or substitutions should fall within the protection scope of the claims attached to the present invention.

Claims

1. A method for constructing a knowledge graph of fault data in the field of UAV ground stations, characterized in that The following steps are involved: Step 1: Based on expert knowledge, sort out and analyze the classification of UAV ground station fault ontology, and subdivide the faults in each profession step by step until the granularity reaches the level at which the faults can be determined, thereby forming a fault ontology model in the field of UAV ground stations, and adding fault level labels and fault type labels to the fault ontology; Step 2: Based on the UAV ground station domain ontology model, a UAV ground station fault knowledge graph is established, including the import of fault data, fault data extraction and knowledge fusion of fault data; wherein, data extraction adopts two modes of <entity-relationship-entity> and <entity-attribute value-attribute>; Step 3: According to the UAV ground station fault knowledge graph, the entities, relationships, and attributes in the graph are stored, and the stored UAV ground station fault knowledge graph is continuously updated based on the import, extraction, and fusion of fault data.

2. A method for constructing a knowledge graph of fault data in the field of UAV ground station according to claim 1, characterized in that In step 2, the imported fault data includes fault information in the equipment manual, fault data and solutions in the field of drone ground stations on the Internet, and records of faults during the use of drone ground stations.

3. The method for constructing a knowledge graph of fault data in the field of UAV ground station according to claim 1 is characterized in that In step 2, the imported fault data is processed through natural language processing to extract knowledge entities, entity attributes and entity relationships, forming <entity-relationship-entity> and <entity-attribute value-attribute> relationships, and the extracted data is aligned in entities and relationships.

4. A method for constructing a knowledge graph of fault data in the field of UAV ground station according to claim 1, characterized in that The UAV ground station fault knowledge graph includes: The query module is used to query the cause of the fault according to the fault phenomenon and fault level, analyze the fault principle, and find a solution; or obtain fault prediction information and avoidance methods based on the value change information.

Images