Nuclear power secondary circuit equipment intelligent diagnosis method and system based on knowledge graph

Abstract

The invention discloses a nuclear power secondary circuit equipment intelligent diagnosis method and system based on a knowledge graph, and the method comprises the following steps: firstly, constructing a nuclear power secondary circuit knowledge graph; secondly, analyzing the time sequence data and determining fault symptoms; then, entity mapping is carried out on the fault symptom set and the equipment; and finally, searching a possible fault mode according to the fault symptom based on the knowledge graph, and performing reasoning so as to determine the fault mode. According to the method, potential faults of equipment can be captured according to monitoring data, possible fault modes are retrieved in the knowledge graph according to captured fault symptoms, further reasoning is carried out to determine the fault modes, and corresponding fault reasons and fault handling related suggestions are given according to the fault modes. The system can assist operators to realize accurate positioning and rapid troubleshooting of faults, and guarantees the operation safety of secondary circuit equipment.

Description

Technical Field

[0001] The present invention relates to the field of nuclear power secondary circuit system monitoring, and in particular to a knowledge graph-based intelligent diagnosis method and system for nuclear power secondary circuit equipment. Background Art

[0002] The secondary circuit of nuclear power contains many devices, and the operating status of these devices will directly affect the operating safety of nuclear power units. Developing an intelligent diagnostic system for equipment is an important task to ensure the safe operation of nuclear power plants.

[0003] In the existing nuclear power equipment status monitoring system, the monitoring of equipment abnormalities, the treatment methods for equipment abnormalities, and the corresponding treatment procedures are not interconnected, and it is impossible to directly complete the continuous process from equipment failure discovery to diagnosis and then to treatment. For abnormal events occurring in nuclear power equipment, the monitoring system often only captures the abnormal signal for the first time, and manual intervention is required to locate the cause of the fault and formulate the subsequent action plan. In the actual operation process, a certain fault symptom of the equipment usually corresponds to multiple fault modes. The operator needs to rely on experience or consult a large amount of information to analyze the abnormal event in order to determine the fault mode. However, there are many equipment in nuclear power plants, and a large amount of data and the coupling relationship between different equipment will seriously affect the speed of troubleshooting. Although a large number of fault diagnosis methods have emerged to assist manual fault diagnosis, the existing fault diagnosis models are mostly focused on determining the cause of the fault, and are less related to post-processing related knowledge such as operation instruction manuals and fault handling cases.

[0004] Existing knowledge graph applications are mostly focused on the management of knowledge bases such as equipment data and case libraries, and the interaction between nuclear power operators and knowledge graphs is achieved through semantic analysis of manual input. In the process of fault diagnosis, it is necessary to manually determine the signs of the fault and interact with the knowledge graph to complete the fault location and traceability, and the degree of automation is not high. Alternatively, the time series data can be directly input into the graph and compared with the standard value to infer the cause of the abnormality, but the nuclear power operation data has a nonlinear coupling relationship, and only a single measurement point value time series comparison judgment cannot cover all faults or potential faults.

[0005] The current knowledge graph reasoning method performs reasoning through predefined IF-THEN rule templates. The corresponding reasoning templates are defined according to the specific number of faults. However, the actual number of faults is large, and a large number of templates often need to be defined, making the rule design process cumbersome and requiring iteration together with the graph iteration process, or using an external reasoning engine program to complete the reasoning task of the knowledge graph. The use of the graph reasoning engine involves a new algorithm language, which increases the learning and maintenance costs of actual engineering applications. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide an intelligent diagnosis system based on knowledge graphs to address the deficiencies of existing nuclear power secondary circuit equipment fault monitoring systems. The system can quickly locate possible fault causes based on abnormal data associated with a single or multiple measuring points, and provide corresponding fault handling measures, which can assist operators in making equipment maintenance decisions more quickly. At the same time, a knowledge graph reasoning method is proposed based on the graph structure to simplify the graph reasoning rule design process.

[0007] In order to achieve the above object, the present invention adopts the following technical solutions:

[0008] A knowledge graph-based intelligent diagnosis method for nuclear power secondary circuit equipment includes the following steps:

[0009] S1. Construct a knowledge graph and store the knowledge graph in a graph database. The specific method is as follows:

[0010] S11. Model the key equipment of the nuclear power secondary circuit and construct the hierarchical relationship of the graph ontology. The graph hierarchy is divided into multiple entities such as equipment, failure mode, failure symptom, failure cause, failure handling measures, troubleshooting suggestions, and follow-up action suggestions.

[0011] S12. Based on the acquired historical cases of nuclear power equipment fault diagnosis, entity recognition and relationship extraction are performed according to the hierarchical relationship of the graph ontology to obtain diagnostic knowledge triples, and the diagnostic knowledge triples are fused.

[0012] S13. Perform entity recognition and relationship extraction based on the operation instruction manual and operation-related information obtained from the nuclear power operator to obtain operation and maintenance related triples.

[0013] S14. Fusing the triples obtained in steps S12 and S13 to link the operation and maintenance knowledge with the diagnosis knowledge.

[0014] S15. On the basis of step S14, a knowledge graph is constructed based on the obtained diagnostic knowledge triples and operation and maintenance related triples, and the knowledge graph is stored in a graph database.

[0015] S2. Read the monitoring data of the nuclear power secondary circuit equipment to obtain the corresponding time series data. Analyze the time series data of the nuclear power secondary circuit equipment, detect the time series status information of a single measuring point value, and make a joint judgment based on multiple measuring points. If an abnormal state is detected, the fault symptom is determined. The method for determining the fault symptom is as follows: first, cluster the fault data and fault simulation data in the historical operation, determine the centers and boundaries of each cluster, obtain the fault data clustering space, and divide the clustering results into corresponding fault symptom types, and then map the monitored abnormal data to the fault data clustering space to obtain the corresponding fault symptom. If no fault occurs, analyze the time series data of the next time period according to the preset time interval.

[0016] S3. Perform entity mapping on the set of fault symptoms obtained in step S2 and the device to be diagnosed that has related fault symptoms, and match the fault symptoms and the device with the entities in the knowledge graph.

[0017] S4. Based on the entities obtained in step S3 (including the fault symptoms and the equipment with the relevant fault symptoms), the Python language is used to link the graph database, and the possible fault modes are retrieved according to the fault symptoms, and then reasoning is performed to determine the fault mode; the reasoning specifically includes:

[0018] For each fault symptom, all fault mode sets containing the fault symptom are retrieved; the intersection of all fault mode sets is taken, and the obtained results include three cases, namely, an empty set, a set including a single element, and a set including multiple elements.

[0019] If the result is an empty set, the fault mode set is reversely searched to obtain the fault symptom set corresponding to the relevant single fault mode; and all fault symptom sets are combined and traversed to find the fault mode combination that can cover the detected fault symptom set, and the combination is sorted from small to large according to the combination complexity and the corresponding fault mode combination set is output.

[0020] If the result is a set containing a single element, the failure mode set is directly output.

[0021] If the obtained result is a set including multiple elements, the fault mode retrieval results are sorted from small to large according to the combination complexity and output.

[0022] S5. Search the knowledge graph according to the fault mode determined in step S4 to find the corresponding fault cause and fault handling measures, troubleshooting suggestions, and follow-up action suggestions.

[0023] In the above technical solution, further, the failure mode, failure cause, failure handling measures, troubleshooting suggestions, and follow-up action suggestions are fed back to the operating personnel, and the operating personnel can perform subsequent processing.

[0024] Furthermore, in step S11, the key equipment of the nuclear power secondary circuit includes a steam turbine body, a high-pressure heater, a low-pressure heater, a deaerator, a condenser, a high-pressure cylinder, a low-pressure cylinder, etc.

[0025] Furthermore, the nuclear power equipment fault diagnosis historical cases in step S12 specifically include equipment failure cases, equipment degradation assessment reports, etc.

[0026] Furthermore, in step S14, during the fusion process, reference resolution and entity disambiguation are performed on the entities in the diagnosis knowledge triples obtained in step S2 and the operation and maintenance related triples obtained in step S3.

[0027] Further, in step S4, the fault symptom node and the faulty device node are connected via a fault mode.

[0028] The present invention also provides a knowledge graph-based intelligent diagnosis system for nuclear power secondary circuit equipment, the system comprising:

[0029] A knowledge graph construction module is used to construct a knowledge graph and store the knowledge graph in a graph database;

[0030] The fault symptom determination module is used to read the monitoring data of the nuclear power secondary circuit equipment, obtain the corresponding time series data, and analyze the time series data. During the analysis process, not only the time series status information of a single measuring point value is detected, but also a joint judgment is made based on multiple measuring points to determine whether the equipment has an abnormal state; if there is an abnormal state, the fault symptom determination is performed; if no fault occurs, the time series data of the next time period is analyzed according to the preset time interval;

[0031] An entity mapping module, used to perform entity mapping on the set of fault symptoms output by the fault symptom determination module and the device to be diagnosed that has the relevant fault symptoms, so as to correspond the fault symptoms and the device to the entities in the knowledge graph;

[0032] A failure mode determination module is used to link a graph database using Python language according to the entities obtained by the entity corresponding module, retrieve possible failure modes according to the failure symptoms, and then perform reasoning to determine the failure mode;

[0033] The fault cause and handling suggestion determination module is used to search the knowledge graph according to the fault mode output by the fault mode retrieval and reasoning module, and find the corresponding fault cause and fault handling measures, troubleshooting suggestions, and follow-up action suggestions.

[0034] The present invention also provides an electronic device, the electronic system comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned knowledge graph-based intelligent diagnosis method for nuclear power secondary circuit equipment.

[0035] The present invention also provides a computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are used to enable a computer to execute the steps of the above-mentioned knowledge graph-based intelligent diagnosis method for nuclear power secondary circuit equipment.

[0036] Compared with the prior art, the present invention has the following beneficial effects:

[0037] The present invention links the knowledge graph with time series data, and can directly start from the data end, comprehensively analyze the status of nuclear power secondary circuit equipment, determine the signs of equipment failure, locate the list of possible causes of the current equipment failure signs in combination with the knowledge graph, and provide subsequent disposal measures. The entire process does not require manual determination of fault signs and manual retrieval, and the reasoning process is clear and controllable, with good interpretability. After locating possible faults, it can guide operating personnel to further investigate the faults. In terms of knowledge graph reasoning, the graph database is called through the programming language and possible fault modes are retrieved according to the fault symptoms, and then reasoning is performed to determine the fault mode, without the need for other reasoning machine programs, and the maintainability is good. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a flow chart of the intelligent diagnosis method of nuclear power secondary circuit equipment based on knowledge graph of the present invention;

[0039] Figure 2 A knowledge graph constructed for the present invention;

[0040] Figure 3 It is a diagnostic flow chart of the intelligent diagnostic system of the present invention. DETAILED DESCRIPTION

[0041] The technical solution of the present invention is further described below in conjunction with the accompanying drawings.

[0042] The present invention takes the fault diagnosis of the secondary circuit high pressure heater of a nuclear power plant as an example to illustrate the method of the present invention.

[0043] The intelligent diagnosis method of nuclear power secondary circuit equipment based on knowledge graph of the present invention has the following process: Figure 1 As shown, the specific steps include:

[0044] Step S100: construct a knowledge graph and store the knowledge graph in a graph database. Figure 2 shown.

[0045] First, the key equipment of the nuclear power secondary circuit is modeled and the hierarchical relationship of the graph ontology is constructed; the graph hierarchy is divided into entities such as equipment, failure mode, failure symptom, failure cause, failure handling measures, troubleshooting suggestions, and follow-up action suggestions. The key equipment of the secondary circuit includes equipment entities such as the turbine body, high-pressure heater, deaerator, high-pressure cylinder, low-pressure cylinder, low-pressure heater and condenser.

[0046] Then, obtain nuclear power related technical information, nuclear power equipment fault diagnosis cases, etc. Based on the acquired nuclear power equipment fault diagnosis historical cases, perform entity recognition and relationship extraction according to the graph ontology hierarchical relationship, obtain diagnostic knowledge triples, and fuse the diagnostic knowledge triples.

[0047] Next, entity recognition and relationship extraction are performed based on the obtained operation instruction manual and operation-related information of the nuclear power operator to obtain operation and maintenance related triples; the diagnostic knowledge triples and the operation and maintenance related triples are fused to link the operation and maintenance knowledge with the diagnostic knowledge.

[0048] Finally, based on the previous step, a knowledge graph is constructed based on the diagnostic knowledge triples and the operation and maintenance related triples, and the knowledge graph is stored in a graph database for subsequent retrieval.

[0049] This step can integrate fault knowledge into the atlas. The specific equipment in the atlas can be associated with one or more fault modes in the technical data or historical fault cases, and the fault causes, fault symptoms, fault modes, fault resolution measures, troubleshooting suggestions, and follow-up action suggestions corresponding to the fault modes in the technical data or cases can be associated with the fault modes in the atlas.

[0050] Step S200, read the monitoring data of the nuclear power secondary circuit equipment, obtain the corresponding time series data, and analyze the time series data. During the analysis process, not only the time series status information of a single measuring point value is detected, but also a joint judgment is made based on multiple measuring points to determine whether the equipment has an abnormal state; if there is an abnormal state, the fault signs are determined; if no fault occurs, the time series data of the next time period is analyzed according to the preset time interval.

[0051] The method for determining the fault symptoms is as follows: first, cluster the fault data and fault simulation data in historical operation, determine the centers and boundaries of each cluster, obtain the fault data clustering space, and divide the clustering results into corresponding fault symptom types, and then map the monitored abnormal data to the fault data clustering space to obtain the corresponding fault symptoms.

[0052] S300, performing entity mapping on the set of fault symptoms obtained in step S200 and the device to be diagnosed that has related fault symptoms, and matching the fault symptoms and the device with the entities in the knowledge graph;

[0053] S400, based on the entity obtained in step S300, use Python language to link the graph database, retrieve possible fault modes according to the fault symptoms, and then perform reasoning to determine the fault mode; the reasoning is specifically:

[0054] For each fault symptom, all fault mode sets containing the fault symptom are retrieved; the intersection of all fault mode sets is taken, and the obtained results include three cases, namely, an empty set, a set including a single element, and a set including multiple elements.

[0055] If the result is an empty set, the fault mode set is reversely searched to obtain the fault symptom set corresponding to the relevant single fault mode; and all fault symptom sets are combined and traversed to find the fault mode combination that can cover the detected fault symptom set, and the combination is sorted from small to large according to the combination complexity and the corresponding fault mode combination set is output.

[0056] If the result obtained is a set containing a single element, the fault mode set is directly output.

[0057] If the obtained result is a set including multiple elements, the fault mode retrieval results are sorted from small to large according to the combination complexity and output.

[0058] S500. Search the knowledge graph according to the fault mode determined in step S400 to find the corresponding fault cause and fault handling measures, troubleshooting suggestions, and follow-up action suggestions.

[0059] When the operating personnel receive the above diagnostic results, they will give priority to handling the corresponding handling suggestions for the fault mode ranked first.

[0060] Another embodiment of the present invention further provides a knowledge graph-based intelligent diagnosis system for nuclear power secondary circuit equipment, the system comprising:

[0061] A knowledge graph construction module is used to construct a knowledge graph and store the knowledge graph in a graph database;

[0062] The fault symptom determination module is used to read the monitoring data of the nuclear power secondary circuit equipment, obtain the corresponding time series data, and analyze the time series data. During the analysis process, not only the time series status information of a single measuring point value is detected, but also a joint judgment is made based on multiple measuring points to determine whether the equipment has an abnormal state; if there is an abnormal state, the fault symptom determination is performed; if no fault occurs, the time series data of the next time period is analyzed according to the preset time interval;

[0063] An entity mapping module, used to perform entity mapping on the set of fault symptoms output by the fault symptom determination module and the device to be diagnosed that has the relevant fault symptoms, so as to correspond the fault symptoms and the device to the entities in the knowledge graph;

[0064] A failure mode determination module is used to link a graph database using Python language according to the entities obtained by the entity corresponding module, retrieve possible failure modes according to the failure symptoms, and then perform reasoning to determine the failure mode;

[0065] The fault cause and handling suggestion determination module is used to search the knowledge graph according to the fault mode output by the fault mode retrieval and reasoning module, and find the corresponding fault cause and fault handling measures, troubleshooting suggestions, and follow-up action suggestions.

[0066] Specifically, the failure mode retrieval and reasoning module can be deployed to the API server, which can accept time point data and entity data, and ultimately determine the failure mode.

[0067] The intelligent diagnosis system of the present invention is used to diagnose the nuclear power secondary circuit equipment in the knowledge graph. The process is as follows: Figure 3 :

[0068] Start the system and start monitoring the equipment operation status. If abnormal data is detected, cluster analysis is performed to determine what kind of fault sign it is. For example, if the high-voltage heater end difference data is abnormal, first read the high-voltage heater measurement point data, calculate the end difference, compare the end difference data with the fault data cluster space, and the classification result is that the hydrophobic end difference is too large.

[0069] The detected fault signs, equipment and fault time points are sent to the fault mode determination module. The fault mode determination module receives the data and corresponds the incoming fault signs and equipment data to the entities in the knowledge graph. The specific method is: all abnormal detection results are sorted and sorted into standard descriptions of the corresponding entities in the knowledge graph, and then output, such as the faulty equipment is a high-pressure heater, and the fault sign is a large difference in the drain end. Then, based on the obtained entities, the Python language is used to link the graph database, and the possible fault modes are retrieved according to the fault signs, and then reasoning is performed to determine the fault mode; after determining the fault mode, the knowledge graph is searched according to the obtained fault mode to obtain the cause of the fault, fault handling measures, troubleshooting suggestions, and follow-up action suggestions.

[0070] The above fault diagnosis results (including fault causes, fault handling measures, troubleshooting suggestions, and follow-up action suggestions) are transmitted to the front end, which displays the fault diagnosis results at the time of the fault and stores the results in the database. The operator can handle the equipment according to the diagnosis results.

[0071] Another embodiment of the present invention also provides an electronic device, the electronic system comprising: one or more processors; a memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned knowledge graph-based intelligent diagnosis method for nuclear power secondary circuit equipment.

[0072] Another embodiment of the present invention further provides a computer-readable storage medium having computer instructions stored thereon, wherein the computer instructions are used to enable a computer to execute the steps of the above-mentioned knowledge graph-based intelligent diagnosis method for nuclear power secondary circuit equipment.

Claims

1. A knowledge graph-based intelligent diagnosis method for nuclear power secondary circuit equipment, characterized in that: The following steps are involved: S1. Construct a knowledge graph and store the knowledge graph in a graph database; S2. Read the monitoring data of the nuclear power secondary circuit equipment, obtain the corresponding time series data, and analyze the time series data. In the analysis process, not only the time series status information of a single measuring point value is detected, but also a joint judgment is made based on multiple measuring points to determine whether the equipment is in an abnormal state; If there is an abnormal state, the fault symptom is determined; if no fault occurs, the time series data of the next time period is analyzed according to the preset time interval; S3, performing entity mapping on the set of fault symptoms obtained in step S2 and the device to be diagnosed that has the relevant fault symptoms, and matching the fault symptoms and the device with the entities in the knowledge graph; S4. Based on the entity obtained in step S3, use Python language to link the graph database, retrieve possible fault modes based on the fault symptoms, and then perform reasoning to determine the fault mode; the reasoning is specifically: For each fault symptom, all fault mode sets containing the fault symptom are retrieved; the intersection of all fault mode sets is taken, and the obtained results include three cases, namely, an empty set, a set including a single element, and a set including multiple elements; If the result is an empty set, reverse search is performed on the fault mode set to obtain the fault symptom set corresponding to the relevant single fault mode; and all fault symptom sets are combined and traversed to find the fault mode combination that can cover the detected fault symptom set, and the combination is sorted from small to large according to the combination complexity and the corresponding fault mode combination set is output; If the result is a set containing a single element, the fault mode set is directly output; If the result obtained is a set including multiple elements, the fault mode retrieval results are sorted from small to large according to the combination complexity and output; S5. Search the knowledge graph according to the fault mode determined in step S4 to find the corresponding fault cause and fault handling measures, troubleshooting suggestions, and follow-up action suggestions.

2. The intelligent diagnosis method for nuclear power secondary circuit equipment based on knowledge graph according to claim 1 is characterized in that: The step S1 specifically includes the following steps: S11. Model the key equipment of the nuclear power secondary circuit and construct a graph ontology hierarchical relationship; the graph hierarchy is divided into entities such as equipment, failure mode, failure symptom, failure cause, failure handling measures, troubleshooting suggestions, and follow-up action suggestions; S12, based on the acquired historical cases of nuclear power equipment fault diagnosis, entity recognition and relationship extraction are performed according to the hierarchical relationship of the graph ontology to obtain diagnostic knowledge triples, and the diagnostic knowledge triples are fused; S13, perform entity recognition and relationship extraction based on the operation instruction manual and operation-related information of the nuclear power operator to obtain operation and maintenance related triples; S14, fusing the diagnosis knowledge triplet and the operation and maintenance related triplet, thereby linking the operation and maintenance knowledge with the diagnosis knowledge; S15. On the basis of step S14, a knowledge graph is constructed based on the diagnostic knowledge triples and the operation and maintenance related triples, and the knowledge graph is stored in a graph database.

3. The intelligent diagnosis method for nuclear power secondary circuit equipment based on knowledge graph according to claim 2 is characterized in that: In step S11, the key equipment of the nuclear power secondary circuit includes a steam turbine body, a high-pressure heater, a low-pressure heater, a deaerator, a high-pressure cylinder, a low-pressure cylinder and a condenser.

4. The intelligent diagnosis method for nuclear power secondary circuit equipment based on knowledge graph according to claim 2 is characterized in that: In step S12, the nuclear power equipment fault diagnosis historical cases include equipment failure cases and equipment degradation assessment reports.

5. The intelligent diagnosis method for nuclear power secondary circuit equipment based on knowledge graph according to claim 1 is characterized in that: In step S2, the determination of the fault symptom specifically includes the following steps: Firstly, the fault data and fault simulation data in historical operation are clustered to determine the centers and boundaries of each cluster to obtain the fault data clustering space. The clustering results are divided into corresponding fault symptom types, and then the monitored abnormal data are mapped to the fault data clustering space to obtain the corresponding fault symptoms.

6. The intelligent diagnosis method for nuclear power secondary circuit equipment based on knowledge graph according to claim 1 is characterized in that: In step S14, during the fusion process, reference resolution and entity disambiguation are performed on the entities in the diagnostic knowledge triples and the operation and maintenance related triples.

7. A knowledge graph-based intelligent diagnosis system for nuclear power secondary circuit equipment, characterized in that: include A knowledge graph construction module is used to construct a knowledge graph and store the knowledge graph in a graph database; The fault symptom determination module is used to read the monitoring data of the nuclear power secondary circuit equipment, obtain the corresponding time series data, and analyze the time series data. During the analysis process, not only the time series status information of a single measuring point value is detected, but also a joint judgment is made based on multiple measuring points to determine whether the equipment is in an abnormal state; If there is an abnormal state, the fault symptom is determined; if no fault occurs, the time series data of the next time period is analyzed according to the preset time interval; An entity mapping module, used to perform entity mapping on the set of fault symptoms output by the fault symptom determination module and the device to be diagnosed that has the relevant fault symptoms, so as to correspond the fault symptoms and the device to the entities in the knowledge graph; A fault mode retrieval and reasoning module is used to link the graph database using Python language according to the entities obtained by the entity corresponding module, retrieve possible fault modes according to the fault symptoms, and then perform reasoning to determine the fault mode; The fault cause and handling suggestion determination module is used to search the knowledge graph according to the fault mode output by the fault mode retrieval and reasoning module, and find the corresponding fault cause and fault handling measures, troubleshooting suggestions, and follow-up action suggestions.

8. An electronic device, characterized in that: include: one or more processors; A memory for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having computer instructions stored thereon, characterized in that: The computer instructions are used to enable a computer to execute the steps of the method according to any one of claims 1 to 6.

Images