Knowledge graph construction method and system based on equipment fault data

Abstract

The invention is suitable for the technical field of knowledge graph construction, and provides a knowledge graph construction method and system based on equipment fault data, and the method comprises the steps: obtaining the multi-source heterogeneous data of an equipment fault, the semi-structured data of a multi-source heterogeneous data packet, the structured rule data and the equipment operation monitoring data; performing data analysis on the multi-source heterogeneous data to determine target key data; performing knowledge extraction according to the target key data to obtain a first entity type, a second entity type, a first relationship type and a second relationship type; and based on the first entity type, the second entity type, the first relationship type and the second relationship type, constructing a knowledge graph of the equipment fault data. According to the method, the key information is extracted from the heterogeneous data, and then the entity type and the relationship type are set according to the key information, so that the precision of the knowledge graph is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of knowledge graph construction, and in particular to a method and system for constructing a knowledge graph based on equipment failure data. Background Art

[0002] Traditional physical asset management of power grid equipment has strict management measures for bidding, procurement, arrival acceptance, waste materials, etc. With the acceleration of the pace of power grid construction, the avalanche of physical assets of power grid equipment has not only increased the workload of material management personnel, but also posed a huge challenge to the existing physical equipment asset management model. A large amount of data and expert knowledge is generated and accumulated at various stages of the life cycle of power grid equipment. The sources and storage media of these data are different, the data forms are different, and the structures are diverse, making the equipment asset management process cumbersome and difficult, and it is easy to have errors in equipment management data. Summary of the invention

[0003] The embodiment of the present application provides a method for constructing a knowledge graph based on equipment failure data, which is used to solve the problem of errors in equipment management data.

[0004] A first aspect of an embodiment of the present application provides a method for constructing a knowledge graph based on equipment failure data, comprising:

[0005] Acquire multi-source heterogeneous data of equipment failure, wherein the multi-source heterogeneous data includes semi-structured data, structured rule data, and equipment operation monitoring data;

[0006] Performing data analysis on the multi-source heterogeneous data to determine target key data;

[0007] Performing knowledge extraction according to the target key data to obtain a first entity type, a second entity type, a first relationship type, and a second relationship type;

[0008] A knowledge graph of equipment failure data is constructed based on the first entity type, the second entity type, the first relationship type, and the second relationship type.

[0009] Furthermore, the performing data analysis on the multi-source heterogeneous data to determine target key data includes:

[0010] The semi-structured data, structured rule data and equipment operation monitoring data are cleaned and preprocessed to extract target key data.

[0011] Furthermore, the extracting knowledge according to the target key data includes:

[0012] Extracting structured knowledge from the semi-structured data by using flag characters or stop words, and then merging the extraction results;

[0013] The knowledge is extracted from the structured rule data by direct dumping.

[0014] Furthermore, the extracting knowledge according to the target key data includes:

[0015] Random sampling is performed based on the target key data extracted from the equipment operation monitoring data to generate a data subset for decision tree training;

[0016] Construct a decision tree and perform standardized processing on the equipment operation monitoring data based on the constructed decision tree;

[0017] Knowledge extraction is performed based on the result of the normalization process.

[0018] Furthermore, the expression for standardizing the equipment operation monitoring data according to the construction of the decision tree is as follows:

[0019]

[0020] Where: γ is the difference parameter in the data, f is the sampling condition of abnormal data samples, is the sampling mean of the data samples within a unit time, is a Gaussian distribution vector, and s′ is a dimensionless parameter.

[0021] Furthermore, the determining the first entity type, the second entity type, the first relationship type, and the second relationship type includes:

[0022] Setting entity type and relationship type priorities, the first entity type priority is greater than the second entity type priority;

[0023] The first relationship type and the second relationship type are determined according to the Euclidean distance between the first entity type and the second entity type.

[0024] Furthermore, the expression of the Euclidean distance between the first entity type and the second entity type is as follows:

[0025]

[0026] Where: i k and j k are the coordinates of the first entity type and the second entity type in the m-dimensional space of the data set.

[0027] A second aspect of an embodiment of the present application provides a knowledge graph construction system based on equipment failure data, including:

[0028] A multi-source heterogeneous data acquisition unit, used to acquire multi-source heterogeneous data of equipment failure, wherein the multi-source heterogeneous data includes semi-structured data, structured rule data, and equipment operation monitoring data;

[0029] A target key data determination unit, configured to perform data analysis on the multi-source heterogeneous data to determine the target key data;

[0030] A knowledge extraction unit, configured to extract knowledge according to the target key data to obtain a first entity type, a second entity type, a first relationship type, and a second relationship type;

[0031] A knowledge graph construction unit is used to construct a knowledge graph of equipment failure data based on the first entity type, the second entity type, the first relationship type and the second relationship type.

[0032] A third aspect of an embodiment of the present application provides a computer device, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein when the processor executes the computer-readable instructions, the steps of the method for constructing a knowledge graph based on equipment fault data as described in any one of the above items are implemented.

[0033] It can be seen from the above technical solutions that the embodiments of the present application have the following advantages:

[0034] This application first obtains multi-source heterogeneous data of equipment failure, then performs data analysis on the multi-source heterogeneous data to determine the target key data; then performs knowledge extraction on the multi-source heterogeneous data to obtain the first entity type, the second entity type, the first relationship type and the second relationship type, and finally constructs a knowledge graph of the equipment failure data based on the first entity type, the second entity type, the first relationship type and the second relationship type. The present invention extracts key information based on the type of heterogeneous data, and then sets the entity type and relationship type according to the key information, so that the knowledge can be further divided, the accuracy of the knowledge graph can be improved, and a more accurate data foundation can be provided for the subsequent use of the knowledge graph analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of an embodiment of a method for constructing a knowledge graph based on equipment fault data in the present invention;

[0036] Figure 2 This is a structural block diagram of an embodiment of a system for building a knowledge graph based on equipment fault data in the present invention;

[0037] Figure 3 The figure is a structural block diagram of a computer device in the present invention. DETAILED DESCRIPTION

[0038] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0039] In this embodiment, the knowledge graph construction method based on equipment failure data is used to reduce data errors in equipment failure data management. In this embodiment, the implementation method can be implemented in the system, can be implemented in the server, and can also be implemented in the terminal, without specific limitation.

[0040] Embodiment 1

[0041] See also Figure 1 , an embodiment of a method for constructing a knowledge graph based on equipment failure data in the present invention comprises the following steps:

[0042] S11. Acquire multi-source heterogeneous data of equipment failure, where the multi-source heterogeneous data includes semi-structured data, structured rule data, and equipment operation monitoring data;

[0043] In this embodiment, the fault data is classified from the multi-source heterogeneous data of the equipment fault, which can be divided into a limited number of categories, such as semi-structured data, structured rule data, and equipment operation monitoring data. Among them, semi-structured data is the data model structure that does not conform to the relational database; structured rule data is data that can be represented and stored using a relational database, or can be logically expressed using a two-dimensional table, and the storage and arrangement of these data conform to certain rules; equipment operation monitoring data is data obtained in real time by the system-related power grid monitoring platform. The equipment operation monitoring data is large in quantity, wide in scope, unclear in entity, unclear in relationship, and incomplete in attributes. Here, the rules of these data are determined with the help of mining technology, and most of the equipment operation monitoring data are in the form of unstructured data.

[0044] S12. Perform data analysis on multi-source heterogeneous data to determine target key data;

[0045] In this embodiment, data analysis is performed on multi-source heterogeneous data to determine target key data, including:

[0046] Perform data cleaning and preprocessing on semi-structured data, structured rule data and equipment operation monitoring data to extract target key data.

[0047] Specifically, semi-structured data is used to extract structured knowledge through flag characters or stop words. Semi-structured data is generally in the form of a table. Most of the data in the table is standardized and formatted. However, some semi-structured data needs to be cleaned when extracting knowledge. Long sentences are directly extracted and merged through flag characters or stop words. Under the guidance of the ontology, the merged results are summarized and sorted, and then the entities, relationships and attributes in the results are matched with the ontology.

[0048] For structured data, they are directly dumped. The column names in the relational database correspond to the labels, attributes or relationships in the graph database, and should be set according to the information and functions contained: some column names in the table can be used as attributes or labels, or relationships at the same time; the values ​​in a certain column of the table can be used as nodes, and can also be repeated as attribute values ​​of other nodes; in addition, the types of edge relationships should be as few as possible to reduce the difficulty of reasoning.

[0049] For the semi-structured data and structured rule data, data cleaning can be performed or directly dumped into graph data. For unstructured data, that is, equipment operation monitoring data, the specific knowledge extraction process is as follows:

[0050] 1. Randomly sample the target key data extracted from the equipment operation monitoring data to generate a data subset for decision tree training;

[0051] 2. Construct a decision tree and standardize the equipment operation monitoring data according to the constructed decision tree;

[0052] 3. Extract knowledge based on the results of standardization processing.

[0053] First, after extracting the target key data from the acquired equipment operation monitoring data, random sampling is performed based on the key data to generate a data subset for the decision tree. After determining the pruned objects in the decision tree, a decision tree is constructed based on the mining depth, and then the equipment operation monitoring data is standardized according to the decision tree. The expression for the standardized processing is as follows:

[0054]

[0055] Where: γ is the difference parameter in the data, f is the sampling condition of abnormal data samples, is the sampling mean of the data samples within a unit time, is a Gaussian distribution vector, s ′ is a dimensionless parameter.

[0056] S13. Perform knowledge extraction based on the target key data to obtain a first entity type, a second entity type, a first relationship type, and a second relationship type;

[0057] S14. Construct a knowledge graph of equipment failure data based on the first entity type, the second entity type, the first relationship type, and the second relationship type.

[0058] Specifically, the priority of the entity type and the relationship type is set, and the priority of the first entity type is greater than the priority of the second entity type;

[0059] The first relationship type and the second relationship type are determined according to the Euclidean distance between the first entity type and the second entity type.

[0060] The first entity type here is the direct fault data in the fault data, such as equipment name, fault function, fault model, fault impact, single point fault, and fault location, which are data that directly show the fault. The second entity type is the indirect fault data, such as fault stage, equipment working mode, equipment fault degree, equipment fault influencing factors, and corrective measures. After determining the entity type, the relationship type is determined by calculating the Euclidean distance between the entity types. The specific relationship types are divided into relationship types that meet the conditions, i.e., the first relationship type, and relationship types that do not meet the conditions, i.e., the second relationship type. The condition set here is to set a threshold. When the calculated Euclidean distance is within the threshold range, the condition is met. The expression of the Euclidean distance is as follows:

[0061]

[0062] Where: i k and j k are the coordinates of the first entity type and the second entity type in the m-dimensional space of the data set.

[0063] In summary, after determining the first entity type, the second entity type, the first relationship type, and the second relationship type, the knowledge graph of the equipment fault data is reconstructed. The above-mentioned knowledge graph construction method extracts key information based on the type of heterogeneous data, and then sets the entity type and relationship type according to the key information, so that the knowledge can be further divided, the accuracy of the knowledge graph can be improved, and a more accurate data foundation can be provided for the subsequent use of the knowledge graph analysis.

[0064] Embodiment 2

[0065] See also Figure 2 , an embodiment of a knowledge graph construction system based on equipment fault data of the present invention comprises the following steps:

[0066] A multi-source heterogeneous data acquisition unit 101 is used to acquire multi-source heterogeneous data of equipment failure, wherein the multi-source heterogeneous data packet includes semi-structured data, structured rule data, and equipment operation monitoring data;

[0067] A target key data determination unit 102 is used to perform data analysis on the multi-source heterogeneous data to determine target key data;

[0068] A knowledge extraction unit 103, configured to extract knowledge according to the target key data to obtain a first entity type, a second entity type, a first relationship type, and a second relationship type;

[0069] The knowledge graph construction unit 104 is used to construct a knowledge graph of equipment fault data based on the first entity type, the second entity type, the first relationship type and the second relationship type.

[0070] For the specific definition of the system, please refer to the definition of the method above, which will not be repeated here. Each module in the above system can be implemented in whole or in part by software, hardware and a combination thereof. The above modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0071] Embodiment 3

[0072] See also Figure 3 The present invention provides a computer device, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor. When the processor executes the computer-readable instructions, the steps of the above method are implemented.

[0073] Those of ordinary skill in the art will appreciate that the units of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition of each example has been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0074] In the embodiments provided by the present invention, it should be understood that the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units can be combined into one unit, one unit can be split into multiple units, or some features can be ignored. In addition, each functional unit in each embodiment of the present invention can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units.

[0075] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-0nlyMemory), random access memory (RAM, RandomAccessMemory), mobile hard disk, magnetic disk or optical disk, etc., which can store program code.

[0076] It can be understood that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein by equivalents. These modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be included in the scope of the claims and specification of the present invention.

Claims

1. A method for constructing a knowledge graph based on equipment failure data, characterized in that: include: Acquire multi-source heterogeneous data of equipment failure, wherein the multi-source heterogeneous data includes semi-structured data, structured rule data, and equipment operation monitoring data; Performing data analysis on the multi-source heterogeneous data to determine target key data; Performing knowledge extraction according to the target key data to obtain a first entity type, a second entity type, a first relationship type, and a second relationship type; A knowledge graph of equipment failure data is constructed based on the first entity type, the second entity type, the first relationship type, and the second relationship type.

2. The method for constructing a knowledge graph based on equipment failure data according to claim 1, characterized in that: The performing data analysis on the multi-source heterogeneous data to determine target key data includes: The semi-structured data, structured rule data and equipment operation monitoring data are cleaned and preprocessed to extract target key data.

3. The method for constructing a knowledge graph based on equipment failure data according to claim 2, characterized in that: The extracting knowledge according to the target key data includes: Extracting structured knowledge from the semi-structured data by using flag characters or stop words, and then merging the extraction results; The knowledge is extracted from the structured rule data by direct dumping.

4. The method for constructing a knowledge graph based on equipment failure data according to claim 3 is characterized in that: The extracting knowledge according to the target key data includes: Random sampling is performed based on the target key data extracted from the equipment operation monitoring data to generate a data subset for decision tree training; Construct a decision tree and perform standardized processing on the equipment operation monitoring data based on the constructed decision tree; Knowledge extraction is performed based on the result of the normalization process.

5. The method for constructing a knowledge graph based on equipment failure data according to claim 4, characterized in that: The expression for standardizing the equipment operation monitoring data according to the construction of the decision tree is as follows: Where: γ is the difference parameter in the data, f is the sampling condition of abnormal data samples, is the sampling mean of the data samples within a unit time, is a Gaussian distribution vector, and s′ is a dimensionless parameter.

6. The method for constructing a knowledge graph based on equipment failure data according to claim 1, characterized in that: The determining of the first entity type, the second entity type, the first relationship type, and the second relationship type includes: Setting entity type and relationship type priorities, wherein the priority of the first entity type is greater than the priority of the second entity type; The first relationship type and the second relationship type are determined according to the Euclidean distance between the first entity type and the second entity type.

7. The method for constructing a knowledge graph based on equipment failure data according to claim 6, characterized in that: The expression of the Euclidean distance between the first entity type and the second entity type is as follows: Where: i k and j k are the coordinates of the first entity type and the second entity type in the m-dimensional space of the data set.

8. A knowledge graph construction system based on equipment failure data, characterized in that: include: A multi-source heterogeneous data acquisition unit, used to acquire multi-source heterogeneous data of equipment failure, wherein the multi-source heterogeneous data includes semi-structured data, structured rule data, and equipment operation monitoring data; A target key data determination unit, configured to perform data analysis on the multi-source heterogeneous data to determine the target key data; A knowledge extraction unit, configured to extract knowledge according to the target key data to obtain a first entity type, a second entity type, a first relationship type, and a second relationship type; A knowledge graph construction unit is used to construct a knowledge graph of equipment failure data based on the first entity type, the second entity type, the first relationship type and the second relationship type.

9. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, characterized in that: When the processor executes the computer-readable instructions, it implements the steps of the method for constructing a knowledge graph based on equipment fault data as described in any one of claims 1 to 7.

10. A readable storage medium storing computer-readable instructions, characterized in that: When the computer-readable instructions are executed by a processor, the steps of the method for constructing a knowledge graph based on equipment failure data as described in any one of claims 1 to 7 are implemented.

Images