Method, device and medium for constructing engine fault atlas and determining engine faults

Abstract

The present disclosure relates to the technical field of engines, and particularly to a method, device, and medium for constructing an engine fault atlas and determining engine faults. The method for constructing an engine fault atlas includes: obtaining an engine fault log; performing preliminary mining on the engine fault log to obtain a first fault information set; performing secondary mining on the first fault information set based on a fault information prompt template to obtain a second fault information set, and constructing an engine fault atlas based on the first fault information set, the relationship types between the second fault information set and fault information. The present disclosure can improve the accuracy and reliability of engine fault location.

Description

Technical Field

[0001] The present disclosure relates to the technical field of engines, and particularly to a method, device, and medium for constructing an engine fault atlas and determining engine faults. Background Art

[0002] As the core component of a power device, the normal operation of an engine plays a decisive role in the safety of the power device.

[0003] In the related art, during the maintenance of an engine, in the case of an engine failure, the fault phenomenon information provided by the user can be processed based on a language model for fault location.

[0004] However, in the fault location and fault maintenance solutions provided in the related art, the training of the language model usually focuses on the understanding of semantics and language, resulting in deviations in the output engine fault location results. Summary of the Invention

[0005] In view of the above problems, the present disclosure is proposed. The present disclosure provides a method, device, and medium for constructing an engine fault atlas and determining engine faults, which can improve the accuracy and reliability of engine fault location.

[0006] According to one aspect of the present disclosure, a method for constructing an engine fault atlas is provided, including:

[0007] Obtain an engine fault log;

[0008] Perform preliminary mining on the engine fault log to obtain a first set of fault information, where the fault information in the first set of fault information includes engine fault phenomenon information, fault cause information, and fault maintenance solution information;

[0009] Perform secondary mining on the first set of fault information based on a fault information prompt template to obtain a second set of fault information, where the fault information in the second set of fault information includes engine fault component information, fault mode information, and fault type information;

[0010] Construct an engine fault atlas based on the first set of fault information, the second set of fault information, and the relationship type between the fault information.

[0011] According to another aspect of the present disclosure, a method for determining an engine fault is provided, including:

[0012] Obtain fault query information;

[0013] Calculate the matching degree between the fault query information and each node in the engine fault map to obtain the matching value between the fault query information and each node, where the engine fault map is determined based on the above-mentioned engine fault map construction method;

[0014] Determine the node corresponding to the maximum matching value as the target node, and in the engine fault map, determine the engine fault information through the target node, where the fault information includes fault causes and / or fault repair suggestions.

[0015] According to another aspect of the present disclosure, there is provided an electronic device including a memory, a processor, and a computer program stored on the memory, and the processor executes the computer program to implement the above method.

[0016] According to still another aspect of the present disclosure, there is provided a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed by a processor, the above method is implemented.

[0017] The engine fault map construction and engine fault determination method, device, and medium provided by the present disclosure can obtain fault phenomenon information, fault cause information, fault maintenance plan information, as well as fault component information, fault mode information, and fault type information under engine fault conditions through two excavations of engine fault logs, and construct an engine fault map based on the six types of fault information and the relationship types between the fault information, so as to integrate and standardize the engine fault information, and facilitate determining a more accurate and reliable engine fault location result based on the engine fault map.

[0018] It should be understood that both the foregoing general description and the following detailed description are exemplary and intended to provide further explanation of the claimed technology. Brief Description of the Drawings

[0019] By describing the embodiments of the present disclosure in more detail in conjunction with the accompanying drawings, the above and other objects, features, and advantages of the present disclosure will become more apparent. The drawings are used to provide a further understanding of the embodiments of the present disclosure, and constitute a part of the specification, and are used to explain the present disclosure together with the embodiments of the present disclosure, and do not constitute a limitation to the present disclosure. In the drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 It is a flowchart illustrating the engine fault map construction method according to an embodiment of the present disclosure.

[0021] Figure 2 It is a schematic diagram illustrating a partial engine fault map according to an embodiment of the present disclosure.

[0022] Figure 3 It is a flowchart showing the engine fault determination method according to an embodiment of the present disclosure.

[0023] Figure 4 It is a block diagram showing the engine fault atlas construction device according to an embodiment of the present disclosure.

[0024] Figure 5 It is a block diagram showing the engine fault atlas construction device according to an embodiment of the present disclosure.

[0025] Figure 6 It is a schematic diagram showing a computer program product according to an embodiment of the present disclosure.

[0026] Figure 7 It is a hardware block diagram showing an electronic device according to an embodiment of the present disclosure. Detailed implementation manners

[0027] In order to make the objectives, technical solutions, and advantages of the present disclosure more obvious, exemplary embodiments according to the present disclosure will be described in detail below with reference to the accompanying drawings. Apparently, the described embodiments are only a part of the embodiments of the present disclosure, rather than all the embodiments of the present disclosure. It should be understood that the present disclosure is not limited by the exemplary embodiments described herein.

[0028] In the related art, during the engine fault handling process, the fault phenomenon information provided by the user is usually processed based on a language model to locate the engine fault.

[0029] However, since the training of the language model usually focuses on the understanding of semantics and language, the language model pays more attention to the language and semantic analysis of the large language model during the process of processing the fault phenomenon information, resulting in insufficient attention to the fault handling information, and causing deviations in the output engine fault location results.

[0030] To solve the above problems, an embodiment of the present disclosure provides an engine fault atlas construction method. This engine fault atlas construction method can be applied to terminal devices such as computers, laptops, or tablet computers, as Figure 1 shown, including:

[0031] Step S101, obtaining an engine fault log;

[0032] Step S102, performing preliminary mining on the engine fault log to obtain a first fault information set;

[0033] Among them, the fault information in the first fault information set includes the fault phenomenon information, fault cause information, and fault maintenance plan information of the engine;

[0034] Step S103, performing secondary mining on the first fault information set based on a fault information prompt template to obtain a second fault information set;

[0035] Among them, the fault information in the second fault information set includes the fault component information, fault mode information, and fault type information of the engine;

[0036] Step S104: Construct an engine fault map based on the first fault information set, the second fault information set, and the relationship type between the fault information.

[0037] In summary, the engine fault map construction method provided by the embodiments of the present disclosure can obtain the fault phenomenon information, fault cause information, fault maintenance plan information, as well as the fault component information, fault mode information, and fault type information under engine fault conditions through two - time mining of the engine fault log, and construct an engine fault map based on the six types of fault information and the relationship type between the fault information, so as to integrate and standardize the engine fault information, and facilitate determining a more accurate and reliable engine fault location result based on the engine fault map.

[0038] The following Figure 1 elaborates in detail the specific implementation manners of each step in the

[0039] In step S101, the terminal device acquires the engine fault log.

[0040] In the embodiments of the present disclosure, the engine fault log is usually a fault log written by maintenance personnel under engine fault conditions, and the engine fault log may include record information of different models of engines in different fault states.

[0041] In an alternative implementation manner, the engine fault log may be stored in the terminal device, and the process of the terminal device acquiring the engine fault log may include: in response to the engine fault log acquisition instruction, reading the engine fault log.

[0042] In an alternative implementation manner, when the data volume of the engine fault log is large, the engine fault log may be stored in an external database, and the process of the terminal device acquiring the engine fault log may include: in response to the engine fault log acquisition instruction, sending an engine fault log acquisition request to the database, and receiving the engine fault log data returned by the database to obtain the engine fault log.

[0043] In step S102, the terminal device performs a preliminary mining on the engine fault log to obtain a first fault information set.

[0044] In the embodiments of the present disclosure, the fault information in the first fault information set includes the fault phenomenon information, fault cause information, and fault maintenance plan information of the engine.

[0045] In an alternative embodiment, the process by which the terminal device performs preliminary mining on the engine fault log to obtain the first fault information set may include: extracting keywords from the engine fault log; wherein, when the keyword is a preset keyword, the engine fault log is processed based on a keyword matching model to obtain the first fault information set. When the engine fault log contains a preset keyword, the engine fault log can be processed based on a keyword matching method to obtain the first fault information set, so as to improve the reliability of the first fault information set obtained from the preliminary mining of the engine fault log.

[0046] In an alternative embodiment, when the keyword is not a preset keyword, the engine fault log is processed based on a semantic recognition model to obtain the first fault information set; when the engine fault log does not contain a preset keyword, the engine fault log can be processed based on a semantic understanding method to obtain the first fault information set, so as to improve the integrity of the first fault information set obtained from the preliminary mining of the engine fault log.

[0047] It should be noted that in the embodiments of the present disclosure, the preset keyword can be determined according to actual needs, and the embodiments of the present disclosure do not make any limitations in this regard.

[0048] Step S103, perform secondary mining on the first fault information set based on a fault information prompt template to obtain a second fault information set;

[0049] In the embodiments of the present disclosure, the fault information in the second fault information set includes the fault component information, fault mode information, and fault type information of the engine.

[0050] It should be noted that the process by which the terminal device performs secondary mining on the first fault information set based on a fault information prompt template to obtain a second fault information set can generally be implemented through large language models (LLMs). The large language model has strong context semantic understanding ability, which can improve the accuracy of the fault information in the obtained second fault information set. Moreover, the prompt template can supplement the background knowledge in the field of engine fault maintenance and guide the large language model to perform fault knowledge mining in the engine field, alleviating the problems of lack of knowledge and hallucinations of the large language model in the professional field of engine fault maintenance, so as to further ensure the reliability of the fault information in the mined second fault information set.

[0051] It is understandable that the prompt template is used to assist the large language model in more accurately mining the fault information in the second fault information set; the prompt template may include task description information, and the task description information includes task processing object description information and task goal description information; the fault template may also include task background information, and the fault template may also include the output format of the mined fault information.

[0052] Among them, the prompt template usually includes a prompt template for mining fault components, a prompt template for mining fault modes, and a prompt template for mining fault types; in the prompt template for mining fault modes, the task background information is multiple existing fault modes; in the fault template for mining fault types, the task background information is multiple existing fault types.

[0053] For example, in the prompt template for mining fault components, the task description information is: Please extract the components that may have failed from the following fault phenomena and fault causes. If there are multiple components, separate them with # in the middle. Only the names of the fault components are required to be output, and no explanations or information other than the fault names need to be output.

[0054] Among them, the task processing object description information is: Please based on the following fault phenomena and fault causes; the keywords in this task processing object description information are: fault phenomena and fault causes.

[0055] The task goal description information is: Extract the components that may have failed; the keywords in this task goal description information are: fault components.

[0056] The output format of the fault information is: If there are multiple components, separate them with # in the middle. Only the names of the fault components are required to be output, and no explanations or information other than the fault names need to be output.

[0057] In the prompt template for mining fault modes, the task description information is: Please combine the existing fault modes: blade fracture, blade damage, EGT indication system failure, variable stator vane system failure, variable bleed valve system failure, low-pressure rotor imbalance, fan blade boss lap, icing, foreign object ingestion, surge, high-pressure rotor imbalance, vibration indication system failure, oil leakage, lubricating oil indication system failure, bearing cavity oil return valve failure, accessory drive system failure, and based on the following fault phenomena and fault causes, select the fault mode from them. Only the names of the fault modes are required to be output, and no explanations or information other than the fault names need to be output.

[0058] Among them, the task background information is: Please combine the existing failure modes: blade fracture, blade damage, EGT indication system failure, adjustable stator vane system failure, adjustable bleed valve system failure, low-pressure rotor imbalance, fan blade boss lap joint, icing, foreign object ingestion, surge, high-pressure rotor imbalance, vibration indication system failure, oil leakage, lubricating oil indication system failure, bearing cavity oil return valve failure, accessory drive system failure.

[0059] The description information of the task processing object is: According to the following failure phenomena and failure reasons; The keywords in the description information of this task processing object are: failure phenomena and failure reasons.

[0060] The description information of the task objective is: Select its failure mode from them; The keyword in the description information of this task objective is: failure mode.

[0061] The output format of the failure information is: It is required to only output the failure mode name, without outputting any explanations and information other than the failure name.

[0062] In the hint template for mining failure types, the task description information is: Please select its failure type from the following failure phenomena and failure reasons according to the existing failure types: in-flight shutdown, ground shutdown, bleed air failure, abnormal engine speed, engine vibration, abnormal engine exhaust temperature, engine surge, engine start failure, engine-driven pump failure, engine control system failure. It is required to only output the failure type name, without outputting any explanations and information other than the failure name.

[0063] Among them, the task background information is: Please according to the existing failure types: in-flight shutdown, ground shutdown, bleed air failure, abnormal engine speed, engine vibration, abnormal engine exhaust temperature, engine surge, engine start failure, engine-driven pump failure, engine control system failure.

[0064] The description information of the task processing object is: According to the following failure phenomena and failure reasons; The keywords in the description information of this task processing object are: failure phenomena and failure reasons.

[0065] The description information of the task objective is: Select its failure type from them; The keyword in the description information of this task objective is: failure type.

[0066] The output format of the failure information is: It is required to only output the failure type name, without outputting any explanations and information other than the failure name.

[0067] In an alternative embodiment, the process by which the terminal device performs secondary mining on the first fault information set based on the fault information prompt template to obtain a second fault information set includes: identifying the task description information in the fault information prompt template, where the task description information includes task processing object description information and task target description information; then, when the keywords in the task processing object description information are fault phenomenon and fault cause, and the keyword in the task target description information is the faulty component, processing the fault phenomenon information and the fault cause information to obtain the faulty component information. The fault phenomenon information and the fault cause information can be processed according to the prompt template for determining the faulty component to obtain more accurate faulty component information.

[0068] For example, assume that the prompt template for mining the faulty component is the prompt template in the above embodiment. If the fault phenomenon information is: Before executing the "flight number" Guangzhou - Chongqing flight, due to the "aircraft number" replacing the right engine as planned, it is reported that the right engine TBV and HPTACC are leaking during the test run. And the fault cause information is: TBV fault. Then the faulty component information determined by the terminal device based on the large language model is: TBV#HPTACC.

[0069] In an alternative embodiment, the process by which the terminal device performs secondary mining on the first fault information set based on the fault information prompt template to obtain a second fault information set includes: identifying the task description information in the fault information prompt template, where the task description information includes task processing object description information, task target description information, and task background information; then, when the keywords in the task processing object description information are fault phenomenon and fault cause, and the keyword in the task target description information is the target keyword, processing the fault phenomenon information and the fault cause information, and determining the fault mode information or fault type information from the task background information, where the target keyword is the fault mode or fault type. The fault phenomenon information and the fault cause information can be processed according to the prompt template for determining the fault mode or fault type, and combined with the task background information, to determine more accurate fault mode or fault type information.

[0070] Exemplarily, assume that the hint template for mining failure modes is the hint template in the above embodiment. If the failure phenomenon information is: Before executing the "flight number" Guangzhou-Chongqing flight, due to the "aircraft number", the right engine was replaced as planned. During the test run, it was reported that the right engine's transmitbleed valve (TBV) and high-pressure turbine active clearance control (HPTACC) leaked oil. And, the failure cause information is: TBV failure. Then the failure mode information determined by the terminal device based on the large language model is: Adjustable bleed valve system failure.

[0071] Assume that the hint template for mining failure types is the hint template in the above embodiment. If the failure phenomenon information is: Before executing the "flight number" Guangzhou-Chongqing flight, due to the "aircraft number", the right engine was replaced as planned. During the test run, it was reported that the right engine's TBV and HPTACC leaked oil. And, the failure cause information is: TBV failure. Then the failure mode information determined by the terminal device based on the large language model is: Engine vibration.

[0072] In step S104, the terminal device constructs an engine fault map based on the first fault information set, the second fault information set, and the relationship type between the fault information.

[0073] In the embodiment of the present disclosure, the nodes in the engine fault map can be any fault information in the first fault information set and the second fault information set, and the node relationship in the engine fault map can be the relationship between the fault information represented by two nodes.

[0074] It should be noted that in the embodiments of the present disclosure, in order to construct an engine fault atlas, it is necessary to define nodes and node relationships; the nodes are divided into six types, including fault phenomena, faulty components, fault causes, maintenance suggestions, fault types, and fault modes. Among them, fault phenomena can record various phenomena of faults and provide detailed information for fault analysis. The language for defining fault phenomenon nodes includes the unique identifier of the fault phenomenon and information for detailed description of the fault phenomenon; faulty components are used to identify specific components or parts that have failed in an aeroengine. The language for defining faulty component nodes includes the unique code of the faulty component and the description information of the working state of the faulty component, such as normal, warning, fault, etc.; fault causes are used to record the direct and indirect causes of faults to provide a basis for in-depth fault analysis. The language for defining fault cause nodes includes the unique identifier of the fault cause and the detailed description of the fault cause; maintenance suggestions represent solutions provided for specific fault phenomena and causes. The language for defining maintenance suggestion nodes includes the unique identifier of the maintenance suggestion and the specific content of the maintenance suggestion; fault types are used to classify faults and provide support for the overall management of faults. The language for defining fault type nodes includes the unique identifier of the fault type and the name of the fault type; fault modes represent possible fault evolution modes in specific engine parts. The language for defining fault mode nodes includes the unique identifier of the fault mode and the detailed description of the fault mode.

[0075] Node relationships are divided into four categories, including location relationships, causal relationships, suggestion relationships, and classification relationships. Among them, location relationships represent the compositional relationships between the engine and the faulty parts, as well as between faulty parts and faulty parts, and are used to provide information on the association between the internal structure of the engine and specific faulty parts; causal relationships are used to describe the causal relationships between fault causes and fault phenomena, as well as the mutual association relationships between fault causes and fault causes (i.e., a certain fault cause may be the direct or indirect cause of another fault cause occurring); suggestion relationships represent specific maintenance measures that should be taken for a specific fault cause; classification relationships are used to describe the engine parts to which the fault phenomena belong, the fault types to which the fault phenomena belong, the faulty parts where the fault causes are located, or the fault modes to which the fault causes belong.

[0076] It can be understood that in the embodiments of the present disclosure, the language for defining nodes can be determined based on actual needs, and the embodiments of the present disclosure do not make any limitations in this regard. For example, in the case of using Cypher language to define nodes, the Cypher language for defining fault phenomenon nodes is as follows:

[0077] CREATE(faultSymptom:FaultSymptom{

[0078] symptomId: 'unique_identifier',

[0079] description:'symptom_description',

[0080] / / Other attributes...

[0081] [[ID=9}}

[0082] Among them, symptomId is the unique identifier of the fault symptom, and description is the information for detailed description of the fault symptom.

[0083] The Cypher language for defining the fault component node is:

[0084] CREATE (faultComponent: FaultComponent {

[0085] componentId: 'unique_identifier',

[0086] description: 'component_description',

[0087] / / Other attributes...

[0088] [[ID=30}}

[0089] Among them, componentId is the unique code of the fault component, and description is the description information of the working state of the fault component.

[0090] The Cypher language for defining the fault cause node is:

[0091] CREATE (faultCause: FaultCause {

[0092] causeId: 'unique_identifier',

[0093] description: 'cause_description',

[0094] / / Other attributes...

[0095] [[ID=51}}

[0096] Among them, causeId is the unique identifier of the fault cause, and description is the detailed description of the fault cause.

[0097] The Cypher language for defining the maintenance suggestion node is:

[0098] CREATE(repairAdvice:RepairAdvice{

[0099] adviceId:'unique_identifier',

[0100] suggestion:'repair_suggestion',

[0101] / / Other attributes...

[0102] })

[0103] Among them, adviceId is the unique identifier of the repair advice, and suggestion is the specific content of the repair advice.

[0104] The Cypher language for defining the fault type node is:

[0105] CREATE(faultType:FaultType{

[0106] typeId:'unique_identifier',

[0107] name:'fault_type_name',

[0108] / / Other attributes...

[0109] })

[0110] Among them, typeId is the unique identifier of the fault type, and name is the name of the fault type.

[0111] The Cypher language for defining the fault mode node is:

[0112] CREATE(faultMode:FaultMode{

[0113] modeId:'unique_identifier',

[0114] description:'mode_description',

[0115] / / Other attributes...

[0116] })

[0117] Among them, modeId is the unique identifier of the fault mode, and description is the detailed description of the fault mode.

[0118] Similarly, in the case of defining node relationships using the Cypher language, the Cypher language for defining the location relationship is as follows:

[0119] CREATE(engine)-[:COMPOSED_OF]->(faultLocation)

[0120] CREATE(faultLocation)-[:COMPOSED_OF]->(faultLocation)

[0121] Among them, engine represents the engine, faultLocation represents the fault location, and COMPOSED_OF represents the location association relationship.

[0122] The Cypher language for defining the causal relationship is as follows:

[0123] CREATE(faultCause)-[:LEADS_TO]->(faultSymptom)

[0124] CREATE(faultCause)-[:LEADS_TO]->(faultCause)

[0125] Among them, faultCause represents the fault cause, faultSymptom represents the fault symptom, and LEADS_TO represents the causal association relationship.

[0126] The Cypher language for defining the suggestion relationship is as follows:

[0127] CREATE(faultCause)-[:SUGGESTS]->(repairAdvice)

[0128] Among them, repairAdvice represents the repair advice, and SUGGESTS represents the suggestion association relationship.

[0129] The Cypher language for defining the classification relationship is as follows:

[0130] CREATE(faultSymptom)-[:BELONGS_TO]->(faultLocation)

[0131] CREATE(faultSymptom)-[:BELONGS_TO]->(FaultType)

[0132] CREATE(faultCause)-[:BELONGS_TO]->(faultLocation)

[0133] CREATE(faultCause)-[:BELONGS_TO]->(faultMode)

[0134] Among them, FaultType is the fault type, faultMode is the fault mode, and BELONGS_TO represents the classification association relationship.

[0135] In an alternative embodiment, the process by which the terminal device constructs an engine fault map based on the first fault information set, the second fault information set, and the relationship type between the fault information may include: in response to a knowledge map construction operation, obtaining a plurality of knowledge map triples, where the nodes in the knowledge map triples are any two fault information in the first fault information set and the second fault information set, and the node relationship in the knowledge map triples is the relationship type between the two fault information; then, generating the engine fault map based on the knowledge map triples. The fault phenomenon information, fault cause information, fault maintenance plan information, fault component information, fault mode information, and fault type information may be used as the nodes of the engine fault map, and combined with the pre-determined relationship type between pairwise fault information, the engine fault map is automatically constructed to realize the mapping of the engine fault information in the engine fault log, so as to more standardly and reasonably represent the engine fault information.

[0136] It should be noted that, in the embodiments of the present disclosure, the process by which the terminal device constructs an engine fault map based on the first fault information set, the second fault information set, and the relationship type between the fault information may be implemented based on a graph database software (Neo4j), and the visualization and efficient construction of the engine fault map may be realized.

[0137] Exemplarily, as Figure 2 shown, Figure 2 shows a schematic diagram of a partial engine fault map constructed in the embodiments of the present disclosure. Among them, the nodes include a junction box, an EGT indication system fault, an abnormal engine exhaust temperature, etc., and the node relationships include a causal relationship: lead-to, a location relationship: locate, and a classification relationship: classify, etc.

[0138] The embodiments of the present disclosure provide an engine fault determination method, which can be applied to a terminal device. As Figure 3 shown, it includes:

[0139] Step S301, obtaining fault query information;

[0140] Step S302: Calculate the matching degree between the fault query information and each node in the engine fault atlas to obtain the matching value between the fault query information and each node;

[0141] Among them, the engine fault atlas is determined based on the engine fault atlas construction method in the above embodiment;

[0142] Step S303: Determine the node corresponding to the maximum matching value as the target node, and in the engine fault atlas, determine the engine fault information through the target node;

[0143] Among them, the fault information includes the fault cause and / or the fault repair suggestion, and the fault information may also include the fault mode and the fault type.

[0144] In summary, the engine fault determination method provided by the embodiments of the present disclosure can compare the fault query statement with the node description information in the engine fault atlas to determine the target node that meets the query requirements, and determine the engine fault information through the target node. Since the engine fault atlas is constructed by mining the fault phenomenon information, fault cause information, fault maintenance plan information, as well as the fault component information, fault mode information and fault type information in the engine fault log, and the relationship types between the predefined fault information, it can integrate and standardize the engine fault information, and can obtain more accurate and reliable engine fault determination results.

[0145] In an alternative implementation manner, the process of the terminal device obtaining the fault query information may include: in response to the fault query information input operation, obtaining the fault query information; or, in response to the fault query request sent by the user terminal, parsing the fault query request to obtain the fault query information.

[0146] It should be noted that the terminal device may determine the matching value between the fault query information and each node based on any grammar matching algorithm, and the embodiments of the present disclosure do not limit this.

[0147] In an alternative embodiment, the process of calculating the matching degree between the fault query information and each node in the engine fault atlas may include: determining the semantic encoding value of the fault query information, and determining the semantic encoding value of the text information of each node; then, based on the semantic encoding value of the fault query information and the semantic encoding value of the text information of each node, determining the recall rate and precision of the fault query information and the text information of each node; further, based on the recall rate and precision of the fault query information and the text information of each node, determining the matching value between the fault query information and each node. The matching degree between the fault query information and each node in the engine fault atlas can be calculated based on the BERT matching model. Since the grammar matching process based on the BERT matching model requires combining the recall rate and precision of the fault query information and the text information of each node to determine the final matching value, the accuracy of the obtained matching value can be improved.

[0148] For example, assume that the fault query information x is <x1,…,x k >, and for the text information of each node is The terminal device can input the fault query information and the text information of the node into the BERT model for encoding to obtain X = <X1,…,X k > and two semantic encoding values respectively, where k represents the number of words after word segmentation of the fault query information and the text information of the node;

[0149] Further, the process by which the terminal device determines the recall rate of the fault query information and the text information of each node based on the semantic encoding value of the fault query information and the semantic encoding value of the text information of the node can be implemented based on the first formula, and the first formula is:

[0150]

[0151] In formula 1, R BERT represents the recall rate, x i represents the i-th word among the k word-segmented words of the fault query information, represents the j-th word among the k word-segmented words of the text information of the node.

[0152] The process by which the terminal device determines the precision of the fault query information and the text information of each node based on the semantic encoding value of the fault query information and the semantic encoding value of the text information of the node can be implemented based on the second formula, and the second formula is:

[0153]

[0154] Based on the recall rate and precision of the fault query information and the text information of each node, the process for the terminal device to determine the matching value between the fault query information and each node can be implemented based on the third formula, where the third formula is:

[0155]

[0156] In an alternative embodiment, in the engine fault map, the process of determining engine fault information through the target node may include: in the engine fault map, determining candidate nodes associated with the target node. Further, determining candidate nodes with a causal relationship with the target node as the first target candidate nodes, and extracting fault causes from the text description information of the first target candidate nodes; and / or, determining candidate nodes with a recommendation relationship with the target node as the second target candidate nodes, and extracting fault repair recommendations from the text description information of the second target candidate nodes; and / or, determining candidate nodes with a classification relationship with the target node as the third target candidate nodes, and determining the fault type and fault mode from the text description information of the third target candidate nodes.

[0157] In an alternative embodiment, in response to obtaining a fault information update instruction, determining the node corresponding to the second-largest matching value as the updated target node; further, in the engine fault map, determining updated engine fault information through the updated target node. It is possible to re-determine updated engine fault information when the user does not recognize the determined engine fault information during the process of the user querying engine fault information, thereby improving the flexibility of determining engine fault information.

[0158] It can be understood that in the embodiments of the present disclosure, the process for the terminal device to determine updated engine fault information through the updated target node in the engine fault map may refer to the process of determining engine fault information through the target node in the above embodiments, and the present disclosure will not elaborate on this.

[0159] An exemplary embodiment of the present disclosure provides an engine map construction device, and this engine map construction device may be a chip of a terminal device. Figure 4 The functional module schematic block diagram of the engine map construction device according to an exemplary embodiment of the present disclosure is shown. As Figure 4 shown, the engine map construction device 400 includes:

[0160] A first acquisition module 401, configured to acquire engine fault logs;

[0161] The first mining module 402 is configured to perform preliminary mining on the engine fault log to obtain a first set of fault information, where the fault information in the first set of fault information includes fault phenomenon information, fault cause information, and fault maintenance plan information of the engine;

[0162] The second mining module 403 is configured to perform secondary mining on the first set of fault information based on a fault information prompt template to obtain a second set of fault information, where the fault information in the second set of fault information includes fault component information, fault mode information, and fault type information of the engine;

[0163] The construction module 404 is configured to construct an engine fault map based on the first set of fault information, the second set of fault information, and the relationship type between the fault information.

[0164] Optionally, the second mining module 403 is configured to:

[0165] Identify the task description information in the fault information prompt template, where the task description information includes task processing object description information and task target description information;

[0166] When the keywords in the task processing object description information are fault phenomenon and fault cause, and the keyword in the task target description information is fault component, process the fault phenomenon information and the fault cause information to obtain the fault component information.

[0167] Optionally, the second mining module 403 is configured to:

[0168] Identify the task description information in the fault information prompt template, where the task description information includes task processing object description information, task target description information, and task background information;

[0169] When the keywords in the task processing object description information are fault phenomenon and fault cause, and the keyword in the task target description information is the target keyword, process the fault phenomenon information and the fault cause information, and determine the fault mode information or fault type information from the task background information, where the target keyword is fault mode or fault type.

[0170] Optionally, the first mining module 402 is configured to:

[0171] Extract the keywords in the engine fault log;

[0172] When the keyword is a preset keyword, process the engine fault log based on a keyword matching model to obtain the first set of fault information.

[0173] Optionally, the building block 404 is configured to:

[0174] In response to a knowledge graph construction operation, obtain a plurality of knowledge graph triples, where the nodes in the knowledge graph triples are any two pieces of fault information in the first fault information set and the second fault information set, and the node relationship in the knowledge graph triples is the relationship type between the two pieces of fault information;

[0175] Generate the engine fault graph based on the knowledge graph triples.

[0176] An exemplary embodiment of the present disclosure provides an engine fault determination device, and the engine fault determination device may be a chip of a terminal device. Figure 5 FIG. shows a schematic block diagram of functional modules of an engine fault determination device according to an exemplary embodiment of the present disclosure. As Figure 5 shown, the engine fault determination device 500 includes:

[0177] A second acquisition module 501, configured to acquire fault query information;

[0178] A matching module 502, configured to calculate a matching degree between the fault query information and each node in the engine fault graph to obtain a matching value between the fault query information and each node, where the engine fault graph is determined based on the above-mentioned engine fault graph construction method;

[0179] A determination module 503, configured to determine the node corresponding to the maximum matching value as the target node, and in the engine fault graph, determine engine fault information through the target node, where the fault information includes a fault cause and / or a fault repair suggestion.

[0180] Optionally, the device further includes an update module 504, configured to:

[0181] In response to obtaining a fault information update instruction, determine the node corresponding to the second largest matching value as the updated target node;

[0182] In the engine fault graph, determine updated engine fault information through the updated target node.

[0183] Optionally, the matching module 502 is configured to:

[0184] Determine a semantic encoding value of the fault query information, and determine a semantic encoding value of the text information of each node;

[0185] Determine the recall rate and precision of the fault query information and the text information of each node according to the semantic coding value of the fault query information and the semantic coding value of the text information of each node;

[0186] Determine the matching value between the fault query information and each node based on the recall rate and precision of the fault query information and the text information of each node.

[0187] An exemplary embodiment of the present disclosure further provides an electronic device, including: at least one processor; and a memory communicatively connected to the at least one processor. The memory stores a computer program executable by the at least one processor, and when the computer program is executed by the at least one processor, it is configured to cause the electronic device to execute the method according to the embodiment of the present disclosure.

[0188] An exemplary embodiment of the present disclosure further provides a non-transitory computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor of a computer, it is configured to cause the computer to execute the method according to the embodiment of the present disclosure.

[0189] As Figure 6 shown, an exemplary embodiment of the present disclosure further provides a computer program product 600, including a computer program 601, wherein when the computer program is executed by a processor of a computer, it is configured to cause the computer to execute the method according to the embodiment of the present disclosure.

[0190] Referring Figure 7 , the structural block diagram of an electronic device 700 that can be used as a terminal device of the present disclosure will now be described. It is an example of a hardware device applicable to various aspects of the present disclosure. The electronic device is intended to represent various forms of digital electronic computer devices, such as, for example, a laptop computer, a desktop computer, a workbench, a personal digital assistant, a server, a blade server, a mainframe computer, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, for example, a personal digital processor, a cellular phone, a smart phone, a wearable device, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0191] As Figure 7As shown, the electronic device 700 includes a computing unit 701, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 702 or a computer program loaded from a storage unit 708 into a random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the electronic device 700 can also be stored. The computing unit 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input / output (I / O) interface 705 is also connected to the bus 704.

[0192] A plurality of components in the electronic device 700 are connected to the I / O interface 705, including: an input unit 706, an output unit 707, a storage unit 708, and a communication unit 709. The input unit 706 can be any type of device capable of inputting information into the electronic device 700. The input unit 706 can receive input numerical or character information and generate key signal inputs related to user settings and / or function controls of the electronic device. The output unit 707 can be any type of device capable of presenting information and can include, but is not limited to, a display, a speaker, a video / audio output terminal, a vibrator, and / or a printer. The storage unit 708 can include, but is not limited to, magnetic disks and optical disks. The communication unit 709 allows the electronic device 700 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks and can include, but is not limited to, a modem, a network card, an infrared communication device, a wireless communication transceiver, and / or a chipset, such as a BluetoothTM device, a WiFi device, a WiMax device, a cellular communication device, and / or the like.

[0193] The computing unit 701 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 701 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 701 executes the various methods and processes described above. For example, in some embodiments, the method of the exemplary embodiments of the present disclosure can be implemented as a computer software program tangibly embodied in a machine-readable medium, such as the storage unit 708. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 700 via the ROM 702 and / or the communication unit 709. In some embodiments, the computing unit 701 can be configured to execute the method of the exemplary embodiments of the present disclosure by any other appropriate means (e.g., by means of firmware).

[0194] The program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program codes cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on the machine, partially on the machine, as a stand-alone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0195] In the context of the present disclosure, a machine-readable medium may be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0196] As used in the present disclosure, the terms “machine-readable medium” and “computer-readable medium” refer to any computer program product, apparatus, and / or device (e.g., a disk, an optical disc, a memory, a programmable logic device (PLD)) for providing machine instructions and / or data to a programmable processor, including a machine-readable medium that receives machine instructions as a machine-readable signal. The term “machine-readable signal” refers to any signal for providing machine instructions and / or data to a programmable processor.

[0197] In order to provide interaction with a user, the systems and techniques described herein may be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0198] The systems and techniques described herein can be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend, middleware, or frontend components. The components of the system can be interconnected to each other by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), and the Internet.

[0199] A computer system can include a client and a server. The client and the server are generally remote from each other and typically interact through a communication network. The client-server relationship is created by computer programs that run on the respective computers and have a client-server relationship with each other.

[0200] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer programs or instructions. When the computer program or instructions are loaded and executed on a computer, the processes or functions described in the embodiments of the present disclosure are executed in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, a terminal, a user device, or other programmable devices. The computer program or instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer program or instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired or wireless manner. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or a data center that integrates one or more available media. The available medium can be a magnetic medium, such as a floppy disk, a hard disk, or a magnetic tape; it can also be an optical medium, such as a digital video disc (DVD); or it can be a semiconductor medium, such as a solid state drive (SSD).

[0201] Although the present disclosure has been described in connection with specific features and their embodiments, it will be apparent that various modifications and combinations can be made without departing from the spirit and scope of the present disclosure. Accordingly, this specification and the drawings are merely exemplary illustrations of the present disclosure as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the present disclosure. Obviously, those skilled in the art can make various changes and modifications to the present disclosure without departing from the spirit and scope of the present disclosure. Thus, if these modifications and variations of the present disclosure fall within the scope of the claims of the present disclosure and their equivalent technologies, the present disclosure is also intended to include these changes and modifications.

Claims

1. A method for constructing an engine fault atlas, characterized in that Including: Obtain the engine fault log; Based on a keyword matching model or a semantic recognition model, preliminarily mine the engine fault log to obtain a first set of fault information, where the fault information in the first set of fault information includes engine fault phenomenon information, fault cause information, and fault maintenance plan information; Through a large language model, based on a fault information prompt template, perform secondary mining on the first set of fault information to obtain a second set of fault information, where the fault information in the second set of fault information includes engine fault component information, fault mode information, and fault type information; Based on the first set of fault information, the second set of fault information, and the relationship types between the fault information, construct an engine fault map; Among them, constructing the engine fault map based on the first set of fault information, the second set of fault information, and the relationship types between the fault information includes: In response to a knowledge graph construction operation, obtain multiple knowledge graph triples, where the nodes in the knowledge graph triples are any two pieces of fault information in the first set of fault information and the second set of fault information, and the node relationship in the knowledge graph triples is the relationship type between the two pieces of fault information; Generate the engine fault map based on the knowledge graph triples.

2. The method for constructing an engine fault atlas according to claim 1, wherein, The secondary mining of the first set of fault information through the large language model based on the fault information prompt template to obtain the second set of fault information includes: Identify the task description information in the fault information prompt template, where the task description information includes task processing object description information and task target description information; When the keyword in the task processing object description information is a fault phenomenon and a fault cause, and the keyword in the task target description information is a fault component, process the fault phenomenon information and the fault cause information to obtain the fault component information.

3. The method for constructing an engine fault atlas according to claim 1, wherein, The secondary mining of the first set of fault information through the large language model based on the fault information prompt template to obtain the second set of fault information includes: Identify the task description information in the fault information prompt template, where the task description information includes task processing object description information, task target description information, and task background information; When the keyword in the task processing object description information is a fault phenomenon and a fault cause, and the keyword in the task target description information is a target keyword, process the fault phenomenon information and the fault cause information, and determine the fault mode information or fault type information from the task background information, where the target keyword is a fault mode or a fault type.

4. The method for constructing an engine fault atlas according to claim 1, wherein The preliminary mining of the engine fault log based on a keyword matching model or a semantic recognition model to obtain the first set of fault information includes: Extract the keywords in the engine fault log; When the keyword is a preset keyword, process the engine fault log based on the keyword matching model to obtain the first set of fault information.

5. A method for determining engine faults, characterized in that Including: Obtain fault query information; Calculate the matching degree between the fault query information and each node in the engine fault map to obtain the matching value between the fault query information and each node, where the engine fault map is determined based on the engine fault map construction method described in any one of claims 1-4; Determine the node corresponding to the maximum matching value as the target node, and in the engine fault map, determine the engine fault information through the target node, where the fault information includes the fault cause and / or the fault repair suggestion.

6. The engine fault determination method according to claim 5, characterized in that, The method further includes: In response to obtaining a fault information update instruction, determine the node corresponding to the second largest matching value as the updated target node; In the engine fault map, determine the updated engine fault information through the updated target node.

7. The engine fault determination method according to claim 5, characterized in that, The calculating the matching degree between the fault query information and each node in the engine fault map to obtain the matching value between the fault query information and each node includes: Determine the semantic encoding value of the fault query information, and determine the semantic encoding value of the text information of each node; According to the semantic encoding value of the fault query information and the semantic encoding value of the text information of each node, determine the recall rate and precision of the fault query information and the text information of each node; Based on the recall rate and precision of the fault query information and the text information of each node, determine the matching value between the fault query information and each node.

8. An electronic device, comprising a memory, a processor, and a computer program stored on the memory, characterized in that, The processor executes the computer program to implement the method described in any one of claims 1-4 or claims 5-7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the method described in any one of claims 1-4 or claims 5-7.

Images