Fault diagnosis method of satellite power supply system based on knowledge graph

Abstract

The invention provides a fault diagnosis method of a satellite power supply system based on a knowledge graph, and relates to the technical field of aerospace. The method comprises the following steps: based on a computer language rule, extracting entities in historical data of a satellite power supply system, and determining an association relationship between the entities to form a triple so as to construct a knowledge graph; wherein the entity sequentially comprises equipment, a subsystem, a single machine, a fault type, a fault criterion and a data parameter according to a hierarchical relationship; segmenting the fault criterion into a plurality of decision expressions based on the logical relational symbols; and performing fault analysis and diagnosis on the abnormal data of the power supply subsystem in the obtained fault time period based on a judgment formula obtained by segmentation, and realizing fault positioning in combination with an entity association relationship in the knowledge graph. According to the fault diagnosis method, the sensor data is matched with the fault mode in the knowledge graph, and the fault reason can be quickly positioned in a structured and intelligent mode, so that the diagnosis time is shortened.

Description

Technical Field

[0001] The present invention relates to the field of aerospace technology, and in particular to a fault diagnosis method and system for a satellite power supply system based on a knowledge graph. Background Art

[0002] Satellite power systems are crucial for maintaining normal on-orbit satellite operations. Failures in satellite power systems account for the highest proportion of satellite failure types. Therefore, maintaining satellite power systems is crucial. Fault diagnosis is a crucial tool for timely identifying faults, implementing effective repair measures, and ensuring the normal operation of satellites.

[0003] Currently, fault diagnosis for satellite power systems often relies on threshold detection, which uses a preset threshold range to examine current telemetry data to determine whether the system is healthy. Because these methods fail to fully exploit the relationships between data, they suffer from several issues, including fragmented data storage, difficulty in data acquisition and processing, limited visualization of the fault location logic, insufficient fault data mining capabilities, and low fault location accuracy. Summary of the Invention

[0004] In response to the above technical problems, the present invention proposes a fault diagnosis method and system for satellite power systems based on knowledge graphs.

[0005] A first aspect of the present invention discloses a fault diagnosis method for a satellite power system based on a knowledge graph, the method comprising:

[0006] S1, based on computer language rules, identifies entities in the historical data of the satellite power system and determines the association relationship between entities to form triples to build a knowledge graph. Entities include equipment, subsystems, single machines, fault types, fault criteria, and data parameters.

[0007] S2, split the fault criterion into multiple decision formulas;

[0008] S3, based on the judgment formula obtained by segmentation, performs fault analysis and diagnosis on the abnormal data of the power subsystem within the fault time period, and combines the entity association relationship in the knowledge graph to realize fault location.

[0009] Step S1 specifically includes:

[0010] S11, based on computer language rules, normalizes the format of historical data of the satellite power system to convert the historical data into data in a specified format containing logical relationships; the historical data of the satellite power system includes historical telemetry information, design parameters of the satellite power system, and fault information;

[0011] S12, identifying entities in the normalized data, and determining associations between the entities to form triples;

[0012] S13 adopts a graph structure, uses nodes and edges to represent triples, and forms a knowledge graph.

[0013] Historical telemetry information includes battery voltage and discharge current; the design parameters of the satellite power system include the rated capacity of the battery pack cells, the number of series and parallel connections, the open-circuit voltage and the short-circuit current; the fault information includes the normal data detection results, the fault characteristic table data and the satellite power system fault plan.

[0014] Step S2 specifically includes:

[0015] Based on the logical relationship symbols, the fault judgment criteria are divided into multiple judgment formulas.

[0016] Step S3 specifically includes:

[0017] S31, preprocessing abnormal data;

[0018] S32, for the pre-processed abnormal data, traverse each decision formula in each unit time to perform threshold judgment to determine the fault diagnosis result;

[0019] S33, based on the fault diagnosis results in all time periods, determines whether a specified type of fault occurs and its corresponding time period of occurrence, and combines the entity information of high-level nodes in the knowledge graph to achieve fault location.

[0020] In step S31 , preprocessing of abnormal data includes: unifying the time type of each source data, removing data duplication values, unifying the time granularity of multi-source data, and filling missing values.

[0021] Fill missing values ​​using forward filling.

[0022] A second aspect of the present invention discloses a fault diagnosis system for a satellite power system based on a knowledge graph, the system comprising:

[0023] The first processing module is configured to identify entities in the historical data of the satellite power system based on computer language rules, and determine the association relationship between the entities to form triples to construct a knowledge graph; wherein the entities include equipment, subsystems, single machines, fault types, fault criteria, and data parameters;

[0024] The second processing module is configured to divide the fault judgment criterion into multiple judgment formulas;

[0025] The third processing module is configured to perform fault analysis and diagnosis on the abnormal data of the power subsystem within the fault time period based on the judgment formula obtained by segmentation, and to locate the fault by combining the entity association relationship in the knowledge graph.

[0026] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps of the knowledge graph-based fault diagnosis method for a satellite power system described in the first aspect of the present invention.

[0027] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program implements the steps of the knowledge graph-based fault diagnosis method for a satellite power system described in the first aspect of the present invention.

[0028] In summary, the proposed solution has the following technical advantages: Compared with existing methods, this invention primarily optimizes fault diagnosis methods for satellite power systems by constructing a knowledge graph. The knowledge graph represents knowledge in a structured manner, enabling complex fault diagnosis-related knowledge to be presented and understood clearly and intuitively. Furthermore, structured knowledge is easier for computers to process and reason about, thereby supporting automated decision-making and intelligent services. The knowledge graph can also integrate knowledge from different data sources into a unified knowledge base, improving both knowledge quality and data interoperability. Furthermore, the knowledge graph supports semantic-based search, allowing users to enter natural language queries and obtain more accurate and relevant results. The reasoning mechanism based on the knowledge graph can infer implicit knowledge, such as indirect relationships between entities, thereby providing more intelligent decision support. Based on this, the fault diagnosis method can support automated fault diagnosis. By matching sensor data with fault patterns in the knowledge graph and using a structured and intelligent approach, the cause of the fault can be quickly located, shortening diagnosis time. This improves diagnostic efficiency and reduces the burden of manual diagnosis, particularly in situations where equipment failures are frequent and complex. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0030] Figure 1 Flowchart of a fault diagnosis method for a satellite power system based on a knowledge graph according to an embodiment of the present invention

[0031] Figure 2 This is a structural diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0033] The present invention provides a fault diagnosis method for satellite power systems based on a knowledge graph. This method, by establishing a knowledge graph model in this field, assists in fault data analysis, thereby achieving more accurate fault diagnosis and location. The knowledge graph provides a structured knowledge representation. Through its construction, fault diagnosis-related knowledge, including equipment, subsystems, individual machines, fault types, fault characteristics, fault criteria, and related signals, can be stored in a structured form, and features can be correlated. The knowledge graph can also integrate fault diagnosis knowledge from a typical fault library, including fault characteristics, fault criteria, and fault descriptions. By integrating the knowledge graph with the typical fault library, a fault diagnosis knowledge base with complex hierarchical relationships can be formed, improving the accuracy and reliability of diagnosis. The knowledge graph can also be dynamically updated and maintained, allowing it to continuously expand and improve as new fault data is collected and historical experience accumulates. This dynamic nature enables the fault diagnosis system to promptly reflect the latest fault knowledge and repair methods, maintaining the accuracy and timeliness of diagnosis. In this paper, the data characteristics of satellite power systems are first aligned to the time granularity of the data, duplication is removed, and missing values ​​are supplemented. Then, based on the fault data characteristics, fault data manifestations, and fault judgment criteria, computer language fault threshold rules are generated. Furthermore, a domain knowledge graph is constructed by integrating satellite product design knowledge and operational scenario knowledge. Leveraging the knowledge-guided retrieval and reasoning analysis capabilities of the knowledge graph, it can assist in fault data analysis and enhance the visualization of the fault diagnosis process, providing support for fault diagnosis and decision-making.

[0034] The technical solution of the present invention will be further described below in conjunction with the accompanying drawings of the embodiments.

[0035] The first aspect of the present invention discloses a fault diagnosis method for a satellite power system based on a knowledge graph. Figure 1 , the method comprising:

[0036] S1, based on computer language rules, identifies entities in the historical data of the satellite power system and determines the association relationship between entities to form triples to build a knowledge graph. Entities include equipment, subsystems, single machines, fault types, fault criteria, and data parameters.

[0037] Historical data is pre-imported and stored in a designated location. The designated location includes:

[0038] A time series data storage module, which is used to store historical telemetry information of the satellite power system, wherein the historical telemetry information includes battery voltage and discharge current;

[0039] a relational data storage module for storing design parameters of the satellite power system, wherein the design parameters include the rated capacity of a battery pack, the number of series and parallel connections, the open circuit voltage, and the short circuit current; and

[0040] A distributed file storage module is used to store fault information, wherein the fault information includes normal data detection results, fault feature table data, and satellite power system fault plans.

[0041] Step S1 specifically includes:

[0042] S11, based on computer language rules, normalizes the format of historical data of the satellite power system to convert the historical data into data in a specified format containing logical relationships; the historical data of the satellite power system includes historical telemetry information, design parameters of the satellite power system, and fault information;

[0043] Taking fault data characteristics, that is, fault information, as an example, first, based on computer language rules, the fault data characteristics are converted into a complex expression in a specified format containing logical relationships, where each decision formula contains a data parameter code, a relationship symbol and a threshold value. The decision formulas are connected by logical symbols such as AND or NOT, and the decision formula may contain one or more sub-decision formulas.

[0044] For example, the fault data feature described in natural language, "the output array current of the windsurfing panel is too small (70A threshold), and the angle between the windsurfing panel normal and the sun is less than 5° or greater than 355°", will first be converted into a pseudocode expression "(M001+M002<70)and(((M057>=0and M057<=5)or(M057>=355and M057<=360))and((M040>=0and M040<=5)or(M040>=355and M040<=360)))", and stored in the knowledge graph in the form of knowledge.

[0045] S12, identifying entities in the normalized data, and determining the associations between the entities to form triples; the triples can be summarized as (high-level entity, subordinate relationship edge, low-level entity).

[0046] S13 adopts a graph structure, uses nodes and edges to represent triples, and forms a knowledge graph.

[0047] Historical telemetry information includes battery voltage and discharge current; satellite power system design parameters include battery pack cell rated capacity, number of series and parallel connections, open-circuit voltage, and short-circuit current; and fault information includes normal data detection results, fault characteristic table data, and satellite power system fault contingency plans. Sources of fault information include historical fault data, empirical data, and fault data constructed through data simulation. For example, multiple satellite power system experts provide fault names, data characteristics, fault data manifestations, fault judgment criteria, and related data parameters, analysis of the fault-affected individual machines and impact domains, and treatment measures. These fault data characteristics, fault data manifestations, and fault judgment criteria can also be extracted from historical fault information or constructed through fault data simulation. Specifically, satellite power system fault data simulation simulates possible fault conditions that may occur in actual operation of the satellite power subsystem. Based on the characteristics of normal data and fault type, fault data is correspondingly constructed to generate fault data through simulation.

[0048] S2, split the fault criterion into multiple decision formulas;

[0049] Optionally, based on logical relationship symbols, the fault judgment criteria are segmented into multiple decision formulas. The expressions in the knowledge graph are segmented by logical relationship symbols to extract all the minimum set decision formulas in the expressions. Taking the above expression as an example, it is segmented into multiple decision formulas: M001+M002<70, M057>=0, M057<=5, M057>=355, M057<=360, M040>=0, M040<=5, M040>=355, M040<=360.

[0050] S3, based on the judgment formula obtained by segmentation, performs fault analysis and diagnosis on the abnormal data of the power subsystem within the fault time period, and combines the entity association relationship in the knowledge graph to realize fault location.

[0051] Use sensors to obtain abnormal data. Optionally, based on the start and end time of the fault diagnosis, obtain abnormal data of the power subsystem. The abnormal data may include, for example, common power system fault indicators such as battery overcharge protection status, single cell voltage telemetry values, and regulation circuit fault codes.

[0052] Step S3 specifically includes:

[0053] S31, preprocessing abnormal data;

[0054] In step S31, preprocessing the abnormal data includes: unifying the time type of each source data, removing data duplication, unifying the time granularity of multiple source data, and filling missing values. Optionally, the missing values ​​are filled by forward filling method.

[0055] S32, for the pre-processed abnormal data, traverse each decision formula in each unit time to perform threshold judgment to determine the fault diagnosis result;

[0056] S33, based on the fault diagnosis results in all time periods, determines whether a specified type of fault occurs and its corresponding time period of occurrence, and combines the entity information of high-level nodes in the knowledge graph to achieve fault location.

[0057] The knowledge graph can be dynamically updated and maintained. As new fault data is collected and expert experience accumulates, the knowledge graph can be continuously expanded and improved. Based on this, one embodiment of the present invention further includes step S4: updating the knowledge graph. After diagnosis is complete, for example, when new equipment is put into use or a new fault type is discovered, the knowledge graph is dynamically updated to reflect the latest fault knowledge.

[0058] In one embodiment of the present invention, the knowledge graph is constructed based on the satellite power system's design parameters, historical telemetry data, fault information, and other data. Specifically, the power subsystem's design parameters, historical telemetry data, fault information, and other related data must first be collected and classified for import. Telemetry information from the power subsystem, such as battery voltage and discharge current, is stored in a time-series data storage module, such as a time-series database. Relevant equipment parameters of the power subsystem, such as the rated capacity of battery cells, number of series and parallel connections, open-circuit voltage, and short-circuit current, are stored in a relational data storage module, such as a relational database. Furthermore, auxiliary files, such as constant data detection results, fault characteristic table data, and Beidou satellite power system fault response plans, are stored in a distributed file system.

[0059] Since these imported data are multi-source data, there may be situations such as misaligned timestamps between the data, different time granularities between and within the data, repeated sampling at the same time, and missing values ​​in individual time periods. Based on this, in one embodiment of the present invention, before storing the data, the data is further preprocessed, including: unifying the time type of each source data, removing duplicate data values, unifying the time granularity of multi-source data, and supplementing missing values. In one embodiment of the present invention, based on the characteristics of Beidou satellite data, the forward filling method is adopted to supplement missing values.

[0060] As mentioned previously, for fault judgment criteria, multiple satellite power system experts can provide fault names, data characteristics, fault data manifestations, fault judgment criteria and related influencing variables, analysis of the fault's impact on individual machines and the impact domain, and remediation measures. Based on the fault data characteristics, fault data manifestations, and fault judgment criteria, combined with computer language rules, computer language fault threshold rules are generated. Furthermore, the Beidou satellite power system fault information format is standardized to facilitate the subsequent automated identification of fault entities and relationships.

[0061] By combining knowledge graphs, fault information can be effectively organized and managed. Therefore, after standardizing the fault information format, entities in the fault information can be identified and extracted, and the entities can be aligned to ensure consistency in the knowledge graph.

[0062] Then, the relationships between entities are extracted from the fault information to form triples, such as (equipment, including, subsystem), (subsystem, including, single machine), (fault judgment criteria, attributes, fault characteristics), etc. These relationships reflect the associations between equipment, subsystems, faults, and signals.

[0063] Finally, using the graph structure of the knowledge graph, these triplets are represented as nodes and edges to form a knowledge graph, where nodes represent entities and edges represent relationships. For example, a device node is connected to other subsystem, fault, and signal nodes through edges, forming a complex fault relationship network.

[0064] A second aspect of the present invention discloses a fault diagnosis system for a satellite power system based on a knowledge graph, the system comprising:

[0065] The first processing module is configured to identify entities in the historical data of the satellite power system based on computer language rules, and determine the association relationship between the entities to form triples to construct a knowledge graph; wherein the entities include equipment, subsystems, single machines, fault types, fault criteria, and data parameters;

[0066] The second processing module is configured to divide the fault judgment criterion into multiple judgment formulas;

[0067] The third processing module is configured to perform fault analysis and diagnosis on the abnormal data of the power subsystem within the fault time period based on the judgment formula obtained by segmentation, and to locate the fault by combining the entity association relationship in the knowledge graph.

[0068] The described fault diagnosis method and system can be applied to the power systems of Beidou satellites and other satellites, enabling accurate diagnosis and location of satellite power system faults, thereby optimizing the satellite power system's operational state. For example, for a battery pack overcharge fault in a satellite power system, fault diagnosis requires obtaining the battery pack's overcharge status, cell voltage values, or the performance characteristics of the power controller's main and secondary function regulation circuits based on the fault knowledge provided by the knowledge graph.

[0069] A third aspect of the present invention discloses an electronic device. The electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps of the knowledge graph-based fault diagnosis method for a satellite power system described in the first aspect of the present invention.

[0070] Figure 2 FIG. 1 is a structural diagram of an electronic device according to an embodiment of the present invention. Figure 2 As shown, the electronic device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the electronic device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, near field communication (NFC) or other technologies. The display screen of the electronic device can be a liquid crystal display or an electronic ink display screen, and the input device of the electronic device can be a touch layer covering the display screen, or a button, trackball or touchpad provided on the electronic device housing, or an external keyboard, touchpad or mouse.

[0071] Those skilled in the art will understand that Figure 2 The structure shown in the figure is only a structural diagram of the part related to the technical solution of the present disclosure, and does not constitute a limitation on the electronic device to which the solution of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0072] A fourth aspect of the present invention discloses a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the computer program implements the steps of the knowledge graph-based fault diagnosis method for a satellite power system described in the first aspect of the present invention.

[0073] The above embodiments are intended only to illustrate the technical solutions of the present invention and are not intended to limit the same. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the technical solutions described in the above embodiments may be modified or some or all of the technical features thereof may be replaced with equivalents, and such modifications or replacements do not deviate from the essence of the corresponding technical solutions within the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fault diagnosis method for satellite power supply system based on knowledge graph, characterized in that: The method comprises: S1, based on computer language rules, identifies entities in the historical data of the satellite power system and determines the association relationship between entities to form triples to build a knowledge graph. Entities include equipment, subsystems, single machines, fault types, fault criteria, and data parameters. S2, split the fault criterion into multiple decision formulas; S3, based on the judgment formula obtained by segmentation, performs fault analysis and diagnosis on the abnormal data of the power subsystem within the fault time period, and combines the entity association relationship in the knowledge graph to realize fault location.

2. The method according to claim 1, characterized in that Step S1 specifically includes: S11, based on computer language rules, normalizes the format of historical data of the satellite power system to convert the historical data into data in a specified format containing logical relationships; the historical data of the satellite power system includes historical telemetry information, design parameters of the satellite power system, and fault information; S12, identifying entities in the normalized data, and determining associations between the entities to form triples; S13 adopts a graph structure, uses nodes and edges to represent triples, and forms a knowledge graph.

3. The method according to claim 2, characterized in that Historical telemetry information includes battery voltage and discharge current; the design parameters of the satellite power system include the rated capacity of the battery pack cells, the number of series and parallel connections, the open-circuit voltage and the short-circuit current; the fault information includes the normal data detection results, the fault characteristic table data and the satellite power system fault plan.

4. The method according to claim 2, characterized in that Step S2 specifically includes: Based on the logical relationship symbols, the fault judgment criteria are divided into multiple judgment formulas.

5. The method according to claim 1, wherein Step S3 specifically includes: S31, preprocessing abnormal data; S32, for the pre-processed abnormal data, traverse each decision formula in each unit time to perform threshold judgment to determine the fault diagnosis result; S33, based on the fault diagnosis results in all time periods, determines whether a specified type of fault occurs and its corresponding time period of occurrence, and combines the entity information of high-level nodes in the knowledge graph to achieve fault location.

6. The method according to claim 5, characterized in that In step S31 , preprocessing of abnormal data includes: unifying the time type of each source data, removing data duplication values, unifying the time granularity of multi-source data, and filling missing values.

7. The method according to claim 6, characterized in that Fill missing values ​​using forward filling.

8. A fault diagnosis system for satellite power supply system based on knowledge graph, characterized in that: The system comprises: The first processing module is configured to identify entities in the historical data of the satellite power system based on computer language rules, and determine the association relationship between the entities to form triples to construct a knowledge graph; wherein the entities include equipment, subsystems, single machines, fault types, fault criteria, and data parameters; The second processing module is configured to divide the fault judgment criterion into multiple judgment formulas; The third processing module is configured to perform fault analysis and diagnosis on the abnormal data of the power subsystem within the fault time period based on the judgment formula obtained by segmentation, and to locate the fault by combining the entity association relationship in the knowledge graph.

9. An electronic device, characterized in that: The electronic device includes a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, it implements the steps in the fault diagnosis method of a satellite power system based on a knowledge graph as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps of the fault diagnosis method of a satellite power system based on a knowledge graph according to any one of claims 1 to 7 are implemented.

Images