A knowledge graph-driven equipment fault diagnosis method and device

Abstract

The present invention provides a method and apparatus for equipment fault diagnosis based on a knowledge graph. The method utilizes the relationships between equipment structures through knowledge graph technology, obtains entities and relationships related to equipment faults through feature matching of fault data, and deduces possible fault causes through the relationships between equipment structures, enabling more concise and accurate fault diagnosis and reasoning. The method includes: acquiring and processing equipment operating data, using ontology modeling technology to represent relevant knowledge ontology as entities and relationships in a graph, and obtaining a knowledge graph for equipment fault diagnosis; using feature extraction technology to convert sensor fault data into feature vectors, and obtaining entities and relationships related to equipment faults through feature matching; inferring the cause of the fault through the relationships between equipment structures to obtain the fault type and location; and analyzing the diagnosis results and maintenance methods based on the relevant knowledge and rules in the knowledge graph to obtain maintenance recommendations for the equipment fault.

Description

Technical Field

[0001] The present invention relates to the field of optical communication technology, and in particular to a knowledge graph-driven equipment fault diagnosis method and apparatus. Background Art

[0002] In the steel process industry, various equipment failures cause production line shutdowns each year, resulting in low production efficiency and huge economic losses. Therefore, finding an effective means of monitoring equipment status and operation plays a vital role in ensuring safe, stable production operations and reducing costs. Traditional equipment fault diagnosis methods typically rely on expert experience and rule bases, which struggle to cover all possible fault conditions. As equipment monitoring data pools grow larger, accurate identification of fault characteristics requires manual analysis by experienced maintenance personnel, placing high demands on the quality of on-site maintenance personnel and achieving extremely low efficiency. The introduction of knowledge graphs provides prior knowledge and new research methods for tracing equipment faults and rapidly troubleshooting them. Through methods such as automated reasoning and uncertainty reasoning, knowledge graphs provide a deeper analysis of equipment failures, uncovering patterns and patterns, enabling better prediction and diagnosis of equipment failures and promoting intelligent upgrades to equipment repair and maintenance.

[0003] Many existing equipment fault diagnosis methods utilize knowledge graph technology, each employing different technical approaches tailored to specific needs. Chinese Patent Publication No. CN 114491037 A proposes a knowledge graph-based fault diagnosis method that determines the similarity between the feature vector of the current faulty device and the feature vectors of each historically faulty device in the equipment fault knowledge graph. This method requires a large amount of data computation. This present invention proposes a new knowledge graph-driven equipment fault diagnosis method, offering a novel approach that enables more concise and accurate fault diagnosis and reasoning. Summary of the Invention

[0004] In order to solve the technical problems in the background technology, the present invention provides a knowledge graph-driven equipment fault diagnosis method and device, which utilizes the relationship between equipment structures through knowledge graph technology, obtains entities and relationships related to equipment faults through feature matching of fault data, and deduces possible fault causes through the relationship between equipment structures, so that fault diagnosis and reasoning can be performed more concisely and accurately.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] A knowledge graph-driven equipment fault diagnosis method includes the following steps:

[0007] Step S1, acquiring and processing the operating data of the equipment: converting the original data into a data format suitable for knowledge graph construction and analysis;

[0008] Step S2: Construct a knowledge graph for equipment faults: Using ontology modeling technology, represent the relevant knowledge ontology as entities and relationships in the graph to obtain a knowledge graph for equipment fault diagnosis;

[0009] Step S3: Using feature extraction technology, the sensor fault data is converted into feature vectors. The entities and relationships related to the equipment fault are obtained through feature matching. Matching the feature data with the entities and relationships in the knowledge graph includes the following:

[0010] 1) Convert the equipment fault data into a feature vector f = (f1, f2, ... f n ), the knowledge graph is represented as G(V,E), and the entity relationship in the graph is e i,j ; Among them, f1,f2,…f n represents n fault data; V represents the set of entities and concepts; E represents the set of relationships between entities and concepts, i, j represents entity numbers;

[0011] 2) Use the similarity function to measure the similarity between the feature vector and the entity relationship, and find the entity v with the greatest correlation in the knowledge graph i , the calculation method is as follows:

[0012]

[0013] Among them, similarity(f,v i ) is the feature vector f and entity v i The similarity, v i Represents an entity, v i, k represents entity v i The kth eigenvalue of k is the kth fault data;

[0014] 3) Through entity v i Obtain the corresponding equipment component entity and fault type entity;

[0015] 4) Relate the feature vector to the feature relationship e in the knowledge graph i,j Perform matching to find the relationship between the corresponding equipment components and fault types;

[0016] Step S4, reasoning about the cause of the fault through the relationship between the equipment structures to obtain the fault type and location;

[0017] Step S5, fault diagnosis and repair: Based on the relevant knowledge and rules in the knowledge graph, analyze the diagnosis results and repair methods to obtain repair suggestions for equipment failures.

[0018] Furthermore, the step S4 specifically includes:

[0019] 1) Before observing the device data, assume that the device has N possible fault states s1, s2, ...s N When each fault state has a corresponding probability P(s1), P(s2), ...P(s N ),in

[0020] 2) According to the device status, there are M measurement parameters, x1, x2, ... x M Description, for fault status s i , the device state is represented by an M-dimensional vector X i =[x i1 ,x i2 ,...x iM ] is used to describe the state equation of the device;

[0021] 3) According to the state equation of the equipment, the probability distribution function f of each parameter under each fault state is obtained i,k (x k ), xk is the kth measurement parameter;

[0022] 4) Obtain the actual value of the monitoring data based on the monitoring parameters and monitoring data of the measuring point;

[0023] 5) According to Bayes' theorem, given the actual measurement value of the device, the posterior probability of device failure is:

[0024]

[0025] in, Indicates that in fault state S i Next, the jth monitoring parameter The probability distribution function of f i,k (x k ) is brought into the monitoring parameter model to calculate, X i is the state vector, Indicates the actual measured value of the monitoring data; P(S i ) indicates fault status S i probability;

[0026] Determine the fault type and location of the equipment based on the size of the posterior probability. The maximum fault state S i That is the fault type of the equipment, and the fault location of the machine is determined by analyzing the state vector Xi The abnormal values ​​of each parameter in the

[0027] Furthermore, in the inference process, the influence of measurement error is considered and the monitoring data is preprocessed using the Kalman filter method.

[0028] Furthermore, a mathematical model of the equipment is constructed, which contains information including the structure, dynamic characteristics and operating parameters of the machine; the state equation of the machine is established based on this information to describe the response and vibration characteristics of the machine under different fault conditions; the vibration, sound and temperature parameters of the machine are monitored and analyzed to obtain the required monitoring data.

[0029] Furthermore, the step S2 specifically includes the following:

[0030] 1) Using ontology modeling technology, establish an ontology concept set for the equipment, including: a concept set for the equipment structure, a concept set for each measurement point used to monitor the equipment's operating status, a concept set for various phenomena exhibited when the equipment fails, a concept set for the causes of equipment failures, and a concept set for maintenance recommendations;

[0031] 2) Establish the ontological relationship of the equipment and express the relationship between the various ontological concepts of the equipment, including: the equipment component has a fault phenomenon, the sensor monitors the operating status of the equipment, the fault causes other phenomena to occur, the cause of the equipment failure is found, and repair measures are taken;

[0032] 3) Based on the established equipment fault diagnosis knowledge ontology representation, information is extracted from IT-side data through data configuration, entity labels are defined, and an equipment triple relationship table is established to identify the relationships. The equipment knowledge is represented as a triple structure: entity-relationship-entity. A graph database is used to store triple knowledge data. The nodes in the graph database represent entities, corresponding to the entity relationships in the triples. Directed edges represent the relationships between knowledge, corresponding to the relationships in the triples. The constructed equipment knowledge graph includes partial relationships and node visualization diagrams.

[0033] Furthermore, step S1 includes: timely processing of data missing, abnormal values, and duplicate value problems in fault diagnosis through data cleaning, feature extraction, and data preprocessing, extracting the spectral characteristics, time domain characteristics, and wavelet transform characteristics of the equipment from data such as vibration signals, sound signals, and temperature signals, and normalizing and standardizing the feature data.

[0034] Furthermore, the maintenance suggestion in step S5 specifically includes:

[0035] 1) By combining the actual operation conditions and maintenance records of the machinery, the relevant knowledge and rules in the knowledge graph are continuously updated so that the knowledge graph can adapt to the actual operation conditions of the machinery;

[0036] 2) Based on the real-time operating status of the equipment, obtain knowledge graph-driven fault diagnosis results, determine the maintenance method of the machinery, and prepare relevant maintenance tools and materials for maintenance.

[0037] Furthermore, the present invention also provides a diagnostic system for equipment fault diagnosis method based on knowledge graph drive, comprising:

[0038] Data acquisition and processing module: collects sensor data, pre-processes the collected data, and obtains data in a form suitable for knowledge graph construction and analysis;

[0039] Knowledge graph construction module: Using ontology modeling technology, the relevant knowledge ontology is represented as entities and relationships in the graph to obtain the knowledge graph for equipment fault diagnosis;

[0040] Fault feature data and entity and relationship matching module in the knowledge graph: Using feature extraction technology, the sensor fault data is converted into feature vectors, and the entities and relationships related to the equipment fault are obtained through feature matching;

[0041] Fault type and location analysis module: Infers the cause of the fault through the relationship between equipment structures to obtain the fault type and location;

[0042] Fault diagnosis and maintenance suggestion module: Based on the relevant knowledge and rules in the knowledge graph, the diagnosis results and maintenance methods are analyzed to obtain maintenance suggestions for equipment failures.

[0043] The present invention also provides a device for implementing the equipment fault diagnosis method driven by knowledge graph, comprising a processor and a memory;

[0044] Wherein, the processor is configured to execute the device fault diagnosis method driven by knowledge graph;

[0045] The memory is used to store executable instructions of the processor.

[0046] The present invention also provides a computer storable medium on which a computer program is stored. The computer program is executed by a processor to implement the equipment fault diagnosis method driven by knowledge graph.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention provides a knowledge graph-driven equipment fault diagnosis method and device, which utilizes the relationship between equipment structures through knowledge graph technology, obtains entities and relationships related to equipment faults through feature matching of fault data, and deduces possible fault causes through the relationship between equipment structures, enabling more concise and accurate fault diagnosis and reasoning. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a flow chart of a device fault diagnosis method based on knowledge graph drive of the present invention;

[0050] Figure 2 Schematic diagram of the knowledge graph used in the present invention;

[0051] Figure 3 The process of constructing a knowledge graph designed for an embodiment of the present invention;

[0052] Figure 4 A device body model designed for an embodiment of the present invention;

[0053] Figure 5 This is an example of a motor knowledge graph according to an embodiment of the present invention. DETAILED DESCRIPTION

[0054] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings:

[0055] Figure 1 This is a flow chart of a device fault diagnosis method based on knowledge graph drive according to an embodiment of the present invention. Figure 2 Schematic diagram of the knowledge graph used in this invention. Figure 3 The construction process of the knowledge graph designed for this invention is combined with Figure 1 、 Figure 2 and Figure 3 The fault diagnosis method of the present application is described. The device fault diagnosis method based on knowledge graph driving of the present application includes the following steps:

[0056] Step S1, obtain and process the operating data of the equipment: convert the original data into a data form suitable for knowledge graph construction and analysis.

[0057] Step S2, constructing a knowledge graph for equipment failure: using ontology modeling technology, the relevant knowledge ontology is represented as entities and relationships in the graph to obtain a knowledge graph for equipment fault diagnosis.

[0058] Step S3: Using feature extraction technology, the sensor fault data is converted into feature vectors. The entities and relationships related to the equipment fault are obtained through feature matching. Matching the feature data with the entities and relationships in the knowledge graph includes the following:

[0059] 1) Convert the equipment fault data into a feature vector f = (f1, f2, ... f n ), the knowledge graph is represented as G(V,E), and the entity relationship in the graph is e i,j ; Among them, f1,f2,…f n Represents n fault data; V represents the set of entities and concepts; E represents the set of relationships between entities and concepts, and i, j represent entity numbers.

[0060] 2) Use the similarity function to measure the similarity between the feature vector and the entity relationship, and find the entity v with the greatest correlation in the knowledge graph i , the calculation method is as follows:

[0061]

[0062] Among them, similarity(f,v i ) is the feature vector f and entity v i The similarity, v i Represents an entity, v i, k represents entity v i The kth eigenvalue of k is the kth fault data.

[0063] 3) Through entity v i Get the corresponding equipment component entity and fault type entity.

[0064] 4) Relate the feature vector to the feature relationship e in the knowledge graph i,j Perform matching to find the relationship between the corresponding equipment components and fault types.

[0065] Step S4: Infer the cause of the fault through the relationship between the equipment structures to obtain the fault type and location.

[0066] Step S5, fault diagnosis and repair: Based on the relevant knowledge and rules in the knowledge graph, analyze the diagnosis results and repair methods to obtain repair suggestions for equipment failures.

[0067] Furthermore, the step S4 specifically includes:

[0068] 1) Before observing the device data, assume that the device has N possible fault states s1, s2, ...s N When each fault state has a corresponding probability P(s1), P(s2), ...P(s N ),in

[0069] 2) According to the device status, there are M measurement parameters, x1, x2, ... x MDescription, for fault status s i , the device state is represented by an M-dimensional vector X i =[x i1 ,x i2 ,...x iM ] to describe, which is the state equation of the device;

[0070] 3) According to the state equation of the equipment, the probability distribution function f of each parameter under each fault state is obtained i,k (x k ), xk is the kth measurement parameter;

[0071] 4) Obtain the actual value of the monitoring data based on the monitoring parameters and monitoring data of the measuring point;

[0072] 5) According to Bayes' theorem, given the actual measurement value of the device, the posterior probability of device failure is:

[0073]

[0074] in, Indicates that in fault state S i Next, the jth monitoring parameter The probability distribution function of f i,k (x k ) is brought into the monitoring parameter model to calculate, X i is the state vector, Indicates the actual measured value of the monitoring data; P(S i ) indicates fault status S i probability;

[0075] Determine the fault type and location of the equipment based on the size of the posterior probability. The maximum fault state S i That is the fault type of the equipment, and the fault location of the machine is determined by analyzing the state vector X i The abnormal values ​​of each parameter in the

[0076] Furthermore, in the inference process, the influence of measurement error is considered and the monitoring data is preprocessed using the Kalman filter method.

[0077] Furthermore, a mathematical model of the equipment is constructed, which contains information including the structure, dynamic characteristics and operating parameters of the machine; the state equation of the machine is established based on this information to describe the response and vibration characteristics of the machine under different fault conditions; the vibration, sound and temperature parameters of the machine are monitored and analyzed to obtain the required monitoring data.

[0078] Furthermore, the step S2 specifically includes the following:

[0079] 1) Using ontology modeling technology, establish an ontology concept set for the equipment, including: a concept set for the equipment structure, a concept set for each measurement point used to monitor the equipment's operating status, a concept set for various phenomena exhibited when the equipment fails, a concept set for the causes of equipment failures, and a concept set for maintenance recommendations;

[0080] 2) Establish the ontological relationship of the equipment and express the relationship between the various ontological concepts of the equipment, including: the equipment component has a fault phenomenon, the sensor monitors the operating status of the equipment, the fault causes other phenomena to occur, the cause of the equipment failure is found, and repair measures are taken;

[0081] 3) Based on the established equipment fault diagnosis knowledge ontology representation, information is extracted from IT-side data through data configuration, entity labels are defined, and an equipment triple relationship table is established to identify the relationships. The equipment knowledge is represented as a triple structure: entity-relationship-entity. A graph database is used to store triple knowledge data. The nodes in the graph database represent entities, corresponding to the entity relationships in the triples. Directed edges represent the relationships between knowledge, corresponding to the relationships in the triples. The constructed equipment knowledge graph includes partial relationships and node visualization diagrams.

[0082] Furthermore, step S1 includes: timely processing of data missing, abnormal values, and duplicate value problems in fault diagnosis through data cleaning, feature extraction, and data preprocessing, extracting the spectral characteristics, time domain characteristics, and wavelet transform characteristics of the equipment from data such as vibration signals, sound signals, and temperature signals, and normalizing and standardizing the feature data.

[0083] Furthermore, the maintenance suggestion in step S5 specifically includes:

[0084] 1) By combining the actual operation conditions and maintenance records of the machinery, the relevant knowledge and rules in the knowledge graph are continuously updated so that the knowledge graph can adapt to the actual operation conditions of the machinery;

[0085] 2) Based on the real-time operating status of the equipment, obtain knowledge graph-driven fault diagnosis results, determine the maintenance method of the machinery, and prepare relevant maintenance tools and materials for maintenance.

[0086] The present invention also provides a diagnostic system for equipment fault diagnosis method based on knowledge graph drive, comprising:

[0087] Data acquisition and processing module: collects sensor data, pre-processes the collected data, and obtains data in a form suitable for knowledge graph construction and analysis;

[0088] Knowledge graph construction module: Using ontology modeling technology, the relevant knowledge ontology is represented as entities and relationships in the graph to obtain the knowledge graph for equipment fault diagnosis;

[0089] Fault feature data and entity and relationship matching module in the knowledge graph: Using feature extraction technology, the sensor fault data is converted into feature vectors, and the entities and relationships related to the equipment fault are obtained through feature matching;

[0090] Fault type and location analysis module: Infers the cause of the fault through the relationship between equipment structures to obtain the fault type and location;

[0091] Fault diagnosis and maintenance suggestion module: Based on the relevant knowledge and rules in the knowledge graph, the diagnosis results and maintenance methods are analyzed to obtain maintenance suggestions for equipment failures.

[0092] Example 1:

[0093] Taking steel plant motors as an example, a motor fault diagnosis knowledge graph is constructed, and the test case design is explained through motor winding short-circuit faults to verify the feasibility of the method.

[0094] Step S1, acquiring and processing the operating data of the equipment;

[0095] Convert raw data into a data form suitable for knowledge graph construction and analysis;

[0096] Among them, the problems such as missing data, abnormal values, repeated values, etc. in motor fault diagnosis are preprocessed, and the corresponding spectral features, time domain features, and wavelet transform features are extracted from vibration signals, sound signals, temperature signals and other data, and the feature data are normalized and standardized.

[0097] Step S2, constructing a knowledge graph for motor faults;

[0098] Using ontology modeling technology, a set of motor ontology concepts is established, including: a concept set of motor structure, a concept set of measurement points for monitoring motor operating status, a concept set of various phenomena exhibited when a motor fails, a concept set of motor failure causes, and a concept set of maintenance suggestions;

[0099] Establish the ontological relationship of the motor and express the relationship between the various ontological concepts of the motor, including: the cause of the motor failure, the sensor monitoring the operating status of the motor, the failure leading to other phenomena, the motor failure phenomenon, and the measures taken to repair it;

[0100] According to the established motor fault diagnosis knowledge ontology representation, information is extracted from IT side data through data configuration, entity labels are defined, a motor triple relationship table is established, relationships are identified, and motor knowledge is represented as a triple structure: entity-relationship-entity. The graph database Neo4j is used to store triple knowledge data. The nodes in the graph database represent entities, corresponding to the entity relationships in the triples. The directed edges represent the relationships between knowledge, corresponding to the relationships in the triples. The constructed motor knowledge graph has some relationship and node visualization diagrams, such as Figure 4 As shown;

[0101] Motor failures mainly occur in the stator, rotor, and bearing components. Sensors monitor the status of each component at each monitoring point and detect motor heating, abnormal noise and other fault phenomena through the measurement point type. The motor failure phenomena are caused by wear, insulation and other fault reasons, such as Figure 5 shown.

[0102] Step S3, matching the feature data with entities and relationships in the knowledge graph;

[0103] Convert the motor fault data into a feature vector f=(f1,f2,…f n ), the knowledge graph can be represented as G(V,E)

[0104] Among them, V represents the set of entities and concepts; E represents the set of relationships between entities and concepts, and v i Represents an entity, e i,j Represents a relationship, i and j represent entity numbers. The motor fault data includes motor current, vibration, temperature and other data.

[0105] Use the similarity function to measure the similarity between the feature vector and the entity relationship, and find the entity v with the greatest correlation in the knowledge graph. i , the calculation method is as follows:

[0106]

[0107] Among them, v i,k Represents entity v i The kth eigenvalue of ;

[0108] Through entity v i Obtain the corresponding motor component entity and fault type entity;

[0109] The feature vector is related to the feature relationship in the knowledge graph i,j Match and find the relationship between the corresponding motor components and fault types.

[0110] Step S4, reasoning about the cause of the fault based on the relationship between motor structures to obtain the fault type and location;

[0111] When a motor winding short circuit occurs, the possible symptoms include: increased motor temperature, abnormal current, motor vibration, and abnormal sound. Only when the above symptoms appear at the same time can it be determined that the motor winding short circuit has occurred.

[0112] Based on the motor's response and vibration characteristics under different fault conditions, fault cause inference uses an inference matching method to monitor and analyze the motor's vibration, sound, temperature, and current parameters. The fault type and location are inferred based on specific judgment rules, including:

[0113] Before observing the motor data, assume that the motor has N possible fault states s1, s2, ...s N When each fault state has a corresponding probability P(s1), P(s2), ...P(s N ),in

[0114] According to the motor state, there are M measurement parameters, x1, x2, ... x M Description, for fault status s i , the motor state can be represented by an M-dimensional vector X i =[x i1 ,x i2 ,...x iM ] to describe;

[0115] According to the state equation of the motor, the probability distribution function f of each parameter under each fault state is obtained ik (x k );

[0116] Based on the monitoring parameters y1, y2, ...y6 and monitoring data of the six measuring points, the actual values ​​of the 48 measurement parameters are obtained. The calculation formula is as follows:

[0117]

[0118]

[0119] in, represents the measurement error, y j Indicates the value of the jth parameter.

[0120] According to Bayes' theorem, given the actual measured values ​​of the motor, the posterior probability of motor failure is:

[0121]

[0122] in, Indicates that in fault state S iUnder this condition, the probability distribution function of the jth monitoring parameter can be obtained by transforming f i,k (x k ) is brought into the monitoring parameter model to calculate, Indicates the actual measurement value;

[0123] The fault type and location of the motor are determined based on the size of the posterior probability. The maximum fault state S i This is the fault type of the motor, and the fault location of the machine is determined by analyzing the state vector X i The abnormal values ​​of each parameter in the

[0124] In the inference process, considering the influence of measurement error, the Kalman filter method is often used to preprocess the monitoring data.

[0125] Step S5, fault diagnosis and repair;

[0126] Based on the relevant knowledge and rules in the knowledge graph, the diagnostic results and maintenance methods are analyzed to obtain maintenance recommendations for equipment failures.

[0127] The maintenance recommendations specifically include:

[0128] By combining the actual operation conditions of the machinery and maintenance records, the relevant knowledge and rules in the knowledge graph are continuously updated, so that the knowledge graph can adapt to the actual operation conditions of the machinery;

[0129] According to the real-time operating status of the equipment, obtain knowledge graph-driven fault diagnosis results, determine the maintenance method of the machinery, and prepare relevant maintenance tools and materials for maintenance.

[0130] The present invention also provides a device for implementing the equipment fault diagnosis method driven by knowledge graph, comprising a processor and a memory; wherein the processor is configured to execute the equipment fault diagnosis method driven by knowledge graph; and the memory is used to store executable instructions of the processor.

[0131] The present invention also provides a computer storable medium on which a computer program is stored. The computer program is executed by a processor to implement the equipment fault diagnosis method driven by knowledge graph.

[0132] In summary, the present invention discloses a method and device for equipment fault diagnosis driven by knowledge graph. Through ontology modeling technology, relevant knowledge ontology is represented as entities and relationships in the graph to obtain a knowledge graph for equipment fault diagnosis. The data is processed to obtain data suitable for knowledge graph construction and analysis. The sensor data is converted into feature vectors. The entities and relationships related to the equipment fault are obtained through feature matching. The cause of the fault is inferred from the relationship between the equipment structures to obtain the type and location of the fault. According to the relevant knowledge and rules in the knowledge graph, the diagnosis results and maintenance methods are analyzed to obtain maintenance suggestions for the equipment fault. By constructing a knowledge graph to diagnose equipment faults, compared with traditional methods, the present invention improves the accuracy of diagnosis and provides more accurate guidance for equipment maintenance.

[0133] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiment of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal translation scripting language JavaScript, etc.

[0134] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0135] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0136] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0137] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0138] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. A device fault diagnosis method based on knowledge graph drive, characterized in that: The steps include: Step S1, acquiring and processing the operating data of the equipment: converting the original data into a data format suitable for knowledge graph construction and analysis; Step S2: Construct a knowledge graph for equipment faults: Using ontology modeling technology, represent the relevant knowledge ontology as entities and relationships in the graph to obtain a knowledge graph for equipment fault diagnosis; Step S3: Using feature extraction technology, the equipment fault data detected by the sensor is converted into a feature vector. Through feature matching, entities and relationships related to the equipment fault are obtained. The feature vector is matched with entities in the knowledge graph to obtain corresponding equipment component entities and fault type entities. The feature vector is matched with relationships in the knowledge graph to find the corresponding relationships between equipment components and fault types. Step S4, reasoning about the cause of the fault through the relationship between the equipment structures to obtain the fault type and location; Step S5, fault diagnosis and repair: Based on the relevant knowledge and rules in the knowledge graph, the diagnosis results and repair methods are analyzed to obtain repair suggestions for the equipment fault; The step S4 specifically includes: 1) Before observing the device data, assume that the device has N possible fault states s1, s2, ...s N When each fault state has a corresponding probability P(s1), P(s2), ...P(s N ),in 2) According to the device status, there are M measurement parameters, x1, x2, ... x M Description, for fault status s i , the device state is represented by an M-dimensional vector X i =[x i1 ,x i2 ,...x iM ] to describe, which is the state equation of the device; 3) According to the state equation of the equipment, the probability distribution function f of each parameter under each fault state is obtained i,k (x k ), x k is the kth measurement parameter, f i,k (x k ) is the probability distribution function of the kth parameter under the i-th fault state; 4) Obtain the actual value of the monitoring data based on the monitoring parameters and monitoring data of the measuring point; 5) According to Bayes' theorem, given the actual measurement value of the device, the posterior probability of device failure is: in, Indicates that in fault state S i Next, the jth monitoring parameter The probability distribution function of f i,k (x k ) is brought into the monitoring parameter model to calculate; X i is the state vector, Indicates the actual measured value of the monitoring data; P(S i ) indicates fault status S i probability; Determine the fault type and location of the equipment based on the size of the posterior probability. The maximum fault state S i That is the fault type of the equipment, and the fault location of the machine is determined by analyzing the state vector X i The abnormal values ​​of each parameter in the 2. The device fault diagnosis method based on knowledge graph drive according to claim 1 is characterized in that: The step S3 includes the following: 1) Convert the equipment fault data into a feature vector f = (f1, f2, ... f n ), the knowledge graph is represented as G(V,E), and the entity relationship in the graph is e i,j ; Among them, f1,f2,…f n represents n fault data; V represents the set of entities and ontology concepts; E represents the set of relationships between entities and ontology concepts, and i, j represent entity numbers; 2) Use the similarity function to measure the similarity between the feature vector and the entity relationship, and find the entity v with the greatest correlation in the knowledge graph i , the calculation method is as follows: Among them, similarity(f,v i ) is the feature vector f and entity v i The similarity, v i Represents an entity, v i,k Represents entity v i The kth eigenvalue of k is the kth fault data; 3) Through entity v i Obtain the corresponding equipment component entity and fault type entity; 4) Relate the feature vector to the feature relationship e in the knowledge graph i,j Perform matching to find the relationship between the corresponding equipment components and fault types.

3. The device fault diagnosis method based on knowledge graph drive according to claim 1 is characterized in that: In the inference process, the influence of measurement error is considered and the monitoring data is preprocessed using the Kalman filter method.

4. The device fault diagnosis method based on knowledge graph drive according to claim 1 is characterized in that: Construct a mathematical model of the equipment, which includes information on the mechanical structure, dynamic characteristics and operating parameters; This information is used to establish the mechanical state equation to describe the response and vibration characteristics of the machine under different fault conditions; the vibration, sound, and temperature parameters of the machine are monitored and analyzed to obtain the required monitoring data.

5. The device fault diagnosis method based on knowledge graph drive according to claim 1 is characterized in that: The step S2 specifically includes the following: 1) Using ontology modeling technology, establish an ontology concept set for the equipment, including: a concept set for the equipment structure, a concept set for each measurement point used to monitor the equipment's operating status, a concept set for various phenomena exhibited when the equipment fails, a concept set for the causes of equipment failures, and a concept set for maintenance recommendations; 2) Establish the ontological relationship of the equipment and express the relationship between the various ontological concepts of the equipment, including: the equipment component has a fault phenomenon, the sensor monitors the operating status of the equipment, the fault causes other phenomena to occur, the cause of the equipment failure is found, and repair measures are taken; 3) Based on the established equipment fault diagnosis knowledge ontology representation, information is extracted from IT-side data through data configuration, entity labels are defined, and an equipment triple relationship table is established to identify the relationships. The equipment knowledge is represented as a triple structure: entity-relationship-entity. A graph database is used to store triple knowledge data. The nodes in the graph database represent entities, corresponding to the entity relationships in the triples. Directed edges represent the relationships between knowledge, corresponding to the relationships in the triples. The constructed equipment knowledge graph includes partial relationships and node visualization diagrams.

6. The device fault diagnosis method based on knowledge graph drive according to claim 1 is characterized in that: The step S1 includes: timely handling of data missing, abnormal value, and duplicate value problems in fault diagnosis through data cleaning, feature extraction, and data preprocessing, extracting the equipment's spectral characteristics, time domain characteristics, and wavelet transform characteristics from vibration signals, sound signals, and temperature signal data, and normalizing and standardizing the feature data.

7. The device fault diagnosis method based on knowledge graph drive according to claim 1 is characterized in that: The maintenance suggestion in step S5 specifically includes: 1) By combining the actual operation conditions and maintenance records of the machinery, the relevant knowledge and rules in the knowledge graph are continuously updated so that the knowledge graph can adapt to the actual operation conditions of the machinery; 2) Based on the real-time operating status of the equipment, obtain knowledge graph-driven fault diagnosis results, determine the maintenance method of the machinery, and prepare relevant maintenance tools and materials for maintenance.

8. The diagnostic system for equipment fault diagnosis method based on knowledge graph drive according to claim 1, characterized in that: include: Data acquisition and processing module: collects sensor data, pre-processes the collected data, and obtains data in a form suitable for knowledge graph construction and analysis; Knowledge graph construction module: Using ontology modeling technology, the relevant knowledge ontology is represented as entities and relationships in the graph to obtain the knowledge graph for equipment fault diagnosis; Fault feature data and entity and relationship matching module in the knowledge graph: Using feature extraction technology, the sensor fault data is converted into feature vectors, and the entities and relationships related to the equipment fault are obtained through feature matching; Fault type and location analysis module: Infers the cause of the fault through the relationship between equipment structures to obtain the fault type and location; Fault diagnosis and maintenance suggestion module: Based on the relevant knowledge and rules in the knowledge graph, the diagnosis results and maintenance methods are analyzed to obtain maintenance suggestions for equipment failures.

9. A device for implementing the knowledge graph-driven equipment fault diagnosis method according to any one of claims 1 to 7, characterized in that: including a processor and a memory; Wherein, the processor is configured to execute a device fault diagnosis method based on knowledge graph drive as described in any one of claims 1 to 7; The memory is used to store executable instructions of the processor.

10. A computer storable medium, characterized in that: A computer program is stored thereon, and the computer program is executed by a processor to implement a device fault diagnosis method based on knowledge graph drive as described in any one of claims 1 to 7.

Images