Substation equipment operation evaluation method and device based on knowledge graph

Abstract

The present application provides a method and device for evaluating the operation of substation equipment based on a knowledge graph, which relates to the technical field of data processing of power equipment, including: obtaining a multi-source data set of substation equipment operation, and performing text preprocessing on the multi-source data set to obtain knowledge graph entity data, wherein the knowledge graph entity data includes multiple entities related to substation equipment; based on the knowledge graph entity data, using a pre-trained triple model to extract triple information in the multi-source data set, wherein the triple information includes a head entity, a tail entity, and an entity relationship between the head entity and the tail entity; constructing a knowledge graph for evaluating the operation of substation equipment based on the triple information; and determining the operation evaluation results of the target substation equipment according to the knowledge graph. The present application can simplify the standard indicator data for evaluating the operation of substation equipment, and can then quickly locate information related to specific equipment and evaluation indicators, and provide timely decision support.

Description

Technical Field

[0001] The present application relates to the technical field of data processing of power equipment, and in particular to a method and device for evaluating the operation of substation equipment based on a knowledge graph. Background Art

[0002] With the rapid development of the national economy, my country's power system is striding towards large capacity and intelligence, and its annual power generation and installed capacity have leapt to the forefront of the world. The continuous improvement of production and living standards has made users put forward higher requirements for power safety and power supply reliability, and all walks of life are increasingly dependent on the power grid. Once a large-scale power outage and power disaster occurs due to a failure in the power transmission and transformation equipment, it will cause serious losses to the power system and the national economy. At the same time, the expansion of power consumption and the increase in labor costs have made the maintenance expenditure of power transmission and transformation equipment an important part of the total power cost.

[0003] Smart grid is a solid material foundation for achieving high-quality development of power grids. However, the existing smart grid technology standard system can no longer meet the requirements of standardization work in the new era and urgently needs to be optimized and improved. At present, the new power system has problems in the operation evaluation standards of substation equipment, such as insufficient correlation between core indicators and business systems, mismatch between standard application technology and government and user needs, and disconnection between standard digitization and actual application. Summary of the invention

[0004] The present invention provides a method and device for evaluating the operation of substation equipment based on a knowledge graph. Through the knowledge graph, the complex substation equipment operation evaluation standard indicator data and information can be mined and expressed in an orderly manner, so that the complex standard system can become clear and easy to understand, and then the information related to specific equipment and evaluation indicators can be quickly located, and decision support can be provided in a timely manner.

[0005] In a first aspect, a method for evaluating substation equipment operation based on a knowledge graph is provided, comprising:

[0006] Obtain a multi-source data set of substation equipment operation, and perform text preprocessing on the multi-source data set to obtain knowledge graph entity data, which contains multiple entities related to substation equipment;

[0007] Based on the knowledge graph entity data, the pre-trained triple model is used to extract triple information from multi-source data sets. The triple information includes the head entity, the tail entity, and the entity relationship between the head entity and the tail entity.

[0008] Construct a knowledge graph for substation equipment operation evaluation based on triple information;

[0009] Determine the operation evaluation results of the target substation equipment based on the knowledge graph.

[0010] Optionally, the multi-source data set includes structured data, semi-structured data and unstructured data;

[0011] The text preprocessing of the multi-source data set to obtain knowledge graph entity data includes:

[0012] Performing knowledge extraction on the semi-structured data and the unstructured data to obtain structured knowledge extraction results;

[0013] Fusion of the knowledge extraction result and the structured data to obtain knowledge graph text data;

[0014] The knowledge graph text data is segmented based on the point mutual information algorithm to obtain knowledge graph entity data.

[0015] Optionally, the word segmentation processing is performed on the knowledge graph text data based on the point mutual information algorithm to obtain the knowledge graph entity data, including:

[0016] Dividing the knowledge graph text data into multiple words;

[0017] Combining the multiple words based on the point mutual information algorithm to obtain word groups with dependency relationships in the knowledge graph text data;

[0018] The uncombined single words and the combined word groups are determined as knowledge graph entity data related to substation equipment.

[0019] Optionally, combining the multiple words based on the point mutual information algorithm to obtain word groups having dependency relationships in the knowledge graph text data includes:

[0020] Calculating the information entropy between any two adjacent words in the plurality of words;

[0021] Combining two adjacent words whose information entropy is greater than a first preset threshold value to obtain a plurality of word groups having a dependency relationship in the knowledge graph text data;

[0022] Calculating the left information entropy of each word group and a first entity on the left, and the right information entropy of each word group and a second entity on the right, wherein the first entity is a left word group or a left single word adjacent to the word group, and the second entity is a right word group or a right single word adjacent to the word group;

[0023] When it is determined that the left-right entropy product of the left information entropy and the right information entropy is less than a second preset threshold, the word group is combined with the first entity and / or the second entity to form a new word group.

[0024] Optionally, a training method for the triplet model is also included:

[0025] Acquire a sample text and a plurality of sample triples marked corresponding to the sample text, wherein the sample triples include a sample header entity and a sample tail entity related to substation equipment, and an entity relationship between the sample header entity and the sample tail entity;

[0026] Using the sample text as an input feature of the triple model, using the sample triplet as a training label of the triple model, iteratively training the triple model until the loss function of the triple model is less than a third preset threshold, and determining that the training of the triple model is completed;

[0027] Among them, during the training process, the triple model predicts the head entity based on the sample text, and then predicts the tail entity based on the predicted head entity, and then predicts the entity relationship between the predicted head entity and the predicted tail entity based on the predicted head entity and the predicted tail entity.

[0028] Optionally, after constructing the knowledge graph of substation equipment operation evaluation based on the triple information, the method further includes:

[0029] Determine a set of triples in the knowledge graph, and construct a mathematical model between entities and entity relationships based on the set of triples, wherein the data model includes an entity space and an entity relationship space;

[0030] Based on the relationship matrix, the head entity and the tail entity in the entity space are mapped to the entity relationship space to obtain an embedded knowledge graph.

[0031] Optionally, determining the operation evaluation result of the target substation equipment according to the knowledge graph includes:

[0032] Obtain equipment operation data of target substation equipment;

[0033] An operation evaluation is performed on the equipment operation data based on the embedded knowledge graph to obtain an operation evaluation result of the target substation equipment.

[0034] In a second aspect, a substation equipment operation evaluation device based on a knowledge graph is provided, comprising:

[0035] A processing module is used to obtain a multi-source data set of substation equipment operation, and perform text preprocessing on the multi-source data set to obtain knowledge graph entity data, where the knowledge graph entity data includes multiple entities related to substation equipment;

[0036] The extraction module is used to extract triple information from multi-source data sets based on knowledge graph entity data using a pre-trained triple model. The triple information includes a head entity, a tail entity, and the entity relationship between the head entity and the tail entity.

[0037] A construction module for constructing a knowledge graph for substation equipment operation evaluation based on triple information;

[0038] The determination module is used to determine the operation evaluation results of the target substation equipment based on the knowledge graph.

[0039] Optionally, the multi-source data set includes structured data, semi-structured data and unstructured data; the processing module is specifically used to extract knowledge from semi-structured data and unstructured data to obtain structured knowledge extraction results; the knowledge extraction results are integrated with the structured data to obtain knowledge graph text data; and the knowledge graph text data is segmented based on the point mutual information algorithm to obtain knowledge graph entity data.

[0040] Optionally, a processing module is specifically used to divide the knowledge graph text data into multiple single words; combine multiple single words based on the point mutual information algorithm to obtain word groups with dependencies in the knowledge graph text data; and determine the uncombined single words and the combined word groups as knowledge graph entity data related to substation equipment.

[0041] Optionally, the processing module is specifically used to calculate the information entropy between any two adjacent single words among multiple single words; combine two adjacent single words whose information entropy is greater than a first preset threshold value to obtain multiple word groups having a dependency relationship in the knowledge graph text data; calculate the left information entropy of each word group among the multiple word groups and the first entity on the left, and the right information entropy of each word group and the second entity on the right, wherein the first entity is the left word group or the left single word adjacent to the word group, and the second entity is the right word group or the right single word adjacent to the word group; when it is determined that the left-right entropy product of the left information entropy and the right information entropy is less than a second preset threshold value, combine the word group with the first entity and / or the second entity to form a new word group.

[0042] Optionally, the device further comprises: a training module;

[0043] A training module is used to obtain sample text and multiple sample triplets marked corresponding to the sample text, wherein the sample triplet includes a sample head entity, a sample tail entity, and an entity relationship between the sample head entity and the sample tail entity related to the substation equipment; the sample text is used as the input feature of the triplet model, the sample triplet is used as the training label of the triplet model, and the triplet model is iteratively trained until the loss function of the triplet model is less than a third preset threshold, and the training of the triplet model is judged to be completed; wherein, during the training process, the triplet model predicts the head entity based on the sample text, and then predicts the tail entity based on the predicted head entity, and then predicts the entity relationship between the predicted head entity and the predicted tail entity based on the predicted head entity and the predicted tail entity.

[0044] Optionally, the device further includes: a mapping module;

[0045] The construction module is also used to determine the set of triples in the knowledge graph, and to construct a mathematical model between entities and entity relationships based on the set of triples. The data model includes an entity space and an entity relationship space.

[0046] The mapping module is used to map the head entity and the tail entity in the entity space to the entity relationship space based on the relationship matrix to obtain an embedded knowledge graph.

[0047] Optionally, a determination module is specifically used to obtain equipment operation data of the target substation equipment; perform operation evaluation on the equipment operation data based on the embedded knowledge graph to obtain the operation evaluation result of the target substation equipment.

[0048] According to a third aspect, an electronic device is provided, comprising: a processor and a memory, wherein the memory is used to store a computer program, and the processor is used to call and run the computer program stored in the memory to execute the method in the first aspect or its various implementations.

[0049] According to a fourth aspect, a computer-readable storage medium is provided for storing a computer program, wherein the computer program enables a computer to execute the method according to the first aspect or its various implementations.

[0050] Through the technical solution provided by the present invention, after obtaining the multi-source data set of substation equipment operation, the multi-source data set can be pre-processed to obtain knowledge graph entity data, and the knowledge graph entity data contains multiple entities related to substation equipment; then, based on the knowledge graph entity data, the pre-trained triple model can be used to extract triple information in the multi-source data set, and the triple information includes the head entity, the tail entity, and the entity relationship between the head entity and the tail entity; and based on the triple information, a knowledge graph for substation equipment operation evaluation is constructed; finally, the operation evaluation result of the target substation equipment is determined according to the knowledge graph. The technical solution in this application, by constructing a knowledge graph, can mine and orderly express the complex substation equipment operation evaluation standard indicator data and information, so that the complex standard system can become clear and easy to understand, and then it can quickly locate the information related to specific equipment and evaluation indicators, and provide decision support in a timely manner.

[0051] It should be understood that the above general description and the following detailed description are only exemplary and explanatory and cannot limit the present invention. Other features and advantages of the present invention will be described in detail in the following specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0053] Figure 1 An application scenario diagram provided for an embodiment of the present application;

[0054] Figure 2 A schematic diagram of a flow chart of a method for evaluating substation equipment operation based on a knowledge graph provided in an embodiment of the present application;

[0055] Figure 3 A schematic diagram of a flow chart of a method for evaluating substation equipment operation based on a knowledge graph provided in another embodiment of the present application;

[0056] Figure 4 A schematic diagram of the principle flow of embedding processing of a knowledge graph provided in an embodiment of the present application;

[0057] Figure 5 A schematic diagram of the structure of a substation equipment operation evaluation device based on a knowledge graph provided in an embodiment of the present application;

[0058] Figure 6A schematic diagram of the structure of a substation equipment operation evaluation device based on a knowledge graph provided in another embodiment of the present application;

[0059] Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0061] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0062] With the rapid development of the national economy, my country's power system is striding towards large capacity and intelligence, and its annual power generation and installed capacity have leapt to the forefront of the world. The continuous improvement of production and living standards has made users put forward higher requirements for power safety and power supply reliability, and all walks of life are increasingly dependent on the power grid. Once a large-scale power outage and power disaster occurs due to a failure in the power transmission and transformation equipment, it will cause serious losses to the power system and the national economy. At the same time, the expansion of power consumption and the increase in labor costs have made the maintenance expenditure of power transmission and transformation equipment an important part of the total power cost.

[0063] Smart grid is a solid material foundation for achieving high-quality development of power grids. However, the existing smart grid technology standard system can no longer meet the requirements of standardization work in the new era and urgently needs to be optimized and improved. At present, the new power system has problems in the operation evaluation standards of substation equipment, such as insufficient correlation between core indicators and business systems, mismatch between standard application technology and government and user needs, and disconnection between standard digitization and actual application.

[0064] In order to solve the above technical problems, the inventive concept of the present invention is to construct important representations and technical architectures of digital standards by mining the value of data elements of substation equipment operation evaluation standards, connect digital standards for substation equipment operation evaluation with typical business scenarios, establish a correlation mechanism between key technical standards and business processes, and realize the digital upgrade of new power system substation equipment operation evaluation standards, and further promote the development of digital standard systems and intelligent applications. Specifically, by constructing a knowledge graph for substation equipment operation evaluation, the data and information of complex substation equipment operation evaluation standard indicators can be mined and expressed in an orderly manner. Thereafter, the knowledge graph is integrated with the business to achieve a systematic interpretation of the standard structure and indicators.

[0065] It should be understood that the technical solution of the present invention can be applied to the following scenarios, but is not limited to:

[0066] In some possible implementations, Figure 1 An application scenario diagram provided by an embodiment of the present invention, such as Figure 1 As shown, the application scenario may include an electronic device 110 and a network device 120. The electronic device 110 may establish a connection with the network device 120 via a wired network or a wireless network.

[0067] Exemplarily, the electronic device 110 may be a desktop computer, a laptop computer, a tablet computer, etc., but is not limited thereto. The network device 120 may be a terminal device or a server, but is not limited thereto. In one embodiment of the present invention, the electronic device 110 may send a request message to the network device 120, and the request message may be used to request to obtain a multi-source data set of the substation equipment operation. Further, the electronic device 110 may receive a response message sent by the network device 120, and the response message includes the multi-source data set of the substation equipment operation.

[0068] also, Figure 1 An electronic device 110 and a network device 120 are provided as an example, but other numbers of electronic devices and network devices may be included in practice, and the present invention is not limited thereto.

[0069] In some other possible implementations, the technical solution of the present invention may also be executed by the above-mentioned electronic device 110, or the technical solution of the present invention may also be executed by the above-mentioned network device 120, and the present invention does not limit this.

[0070] After introducing the application scenarios of the embodiments of the present invention, the technical solutions of the present invention will be described in detail below:

[0071] Figure 2 A flowchart of a method for evaluating substation equipment operation based on a knowledge graph is provided in an embodiment of the present invention. The method can be performed as follows: Figure 1 The electronic device 110 is shown to perform, but is not limited to, such.

[0072] like Figure 2 As shown, the method may include the following steps:

[0073] Step 210: Obtain a multi-source data set of substation equipment operation, and perform text preprocessing on the multi-source data set to obtain knowledge graph entity data, which includes multiple entities related to the substation equipment.

[0074] Among them, the multi-source data set of substation equipment operation can be data such as national standards, industry standards, enterprise standards and internal enterprise specifications. Because the multi-source data set that can be obtained may not be a machine-readable file, it is necessary to achieve machine-readable through text preprocessing. This application can convert PDF format files into TXT format files by using .replace() in python. Further, through the artificial intelligence technology of natural language processing, the extraction of standard knowledge of substation equipment operation is realized. When extracting knowledge, since the text of the extracted knowledge comes from multiple files, there may be duplication, conflict or inconsistency in the extracted standard knowledge. Therefore, through the method of knowledge fusion, these problems can be eliminated, and the accuracy and completeness of the knowledge graph generated in the end can be ensured. In view of the establishment of a knowledge graph related to substation equipment, it is necessary to have a relevant professional vocabulary as a basis, and use this vocabulary to extract substation operation evaluation knowledge from text data and establish a knowledge graph. Therefore, the knowledge graph text data after knowledge fusion can be segmented to obtain knowledge graph entity data (i.e., professional vocabulary). In the knowledge graph entity data, the entities contained can be single words and word groups with dependencies.

[0075] Step 220: Based on the knowledge graph entity data, use the pre-trained triple model to extract triple information from the multi-source data set.

[0076] The triple information may include a head entity, a tail entity, and an entity relationship between the head entity and the tail entity.

[0077] For the disclosed embodiments, a triple model trained by a triple extraction task can be used to extract triple information from a multi-source data set. In the field of electric power, the extracted entities are electrical equipment, electrical faults, and other contents, all of which come from the above-mentioned knowledge graph entity data. For example, for the sentence "abnormal oil temperature in the transformer" in a multi-source data set, when knowledge is extracted for this sentence through a triple method, it is compared with the knowledge graph entity data. If the knowledge graph entity data contains "transformer", "oil temperature", and other contents, the "transformer" and "oil temperature" in this sentence are extracted through the triple model, and the entity relationship between "transformer" and "oil temperature" is determined.

[0078] Step 230: construct a knowledge graph for substation equipment operation evaluation based on the triple information.

[0079] In a specific application scenario, after extracting the triple information, in order to improve the accuracy of the digitalization of the substation operation evaluation standard, the extracted triple information can be checked, and some of the erroneous triple information can be eliminated. Then, by importing the proofread triple data into Neo4j, the corresponding knowledge graph of the substation operation standard is generated, and a graphical knowledge network is obtained. The knowledge graph can be used to clarify the concepts, attributes and relationships in the substation equipment operation evaluation standard. Among them, the concepts mainly include the type of equipment, operating status, rating indicators, etc. The attributes of the equipment include the model, production date, operating time, etc. of the equipment. The relationship includes the association between the equipment and the indicator, the dependency between the indicators, etc.

[0080] Step 240: Determine the operation evaluation results of the target substation equipment according to the knowledge graph.

[0081] In specific application scenarios, after obtaining a graphical knowledge graph, the function of intelligent search can be realized, and the knowledge network of the operating standard entities and relationships of substation equipment can be quickly obtained, and the visualization function can be realized. When encountering substation equipment operation failure problems, Neo4j can be used to retrieve the fault problems and give the corresponding fault causes, fault solutions and corresponding standard requirements. In this way, appropriate maintenance plans can be implemented intelligently before serious faults occur, which can greatly save power maintenance expenses and reduce total power costs. At the same time, different solutions can be graded according to different reference standards. The standards are graded according to the source of the standards, and the national standards are listed as national level, the industry standards are listed as industry level, and the internal standards of the enterprise and some other specifications are classified as enterprise level. The higher the level of grading, the greater the reference value of the corresponding operation evaluation.

[0082] In summary, according to the substation equipment operation evaluation method based on the knowledge graph provided by the present invention, after obtaining the multi-source data set of the substation equipment operation, the multi-source data set can be preprocessed with text to obtain the knowledge graph entity data, and the knowledge graph entity data contains multiple entities related to the substation equipment; then, based on the knowledge graph entity data, the pre-trained triple model can be used to extract the triple information in the multi-source data set, and the triple information includes the head entity, the tail entity, and the entity relationship between the head entity and the tail entity; and a knowledge graph for substation equipment operation evaluation is constructed based on the triple information; finally, the operation evaluation result of the target substation equipment is determined according to the knowledge graph. The technical solution in this application can mine and orderly express the complex substation equipment operation evaluation standard indicator data and information by constructing a knowledge graph, so that the complex standard system can become clear and easy to understand, and then it can quickly locate the information related to specific equipment and evaluation indicators, and provide decision support in time.

[0083] based on Figure 2 The embodiment shown is a refinement and extension of the above embodiment. In order to fully illustrate the specific implementation process of the method of this embodiment, this embodiment provides the following Figure 3 The specific method shown. Figure 3 based on Figure 2 As shown in 3, the method comprises the following steps:

[0084] Step 310: Acquire a multi-source data set of substation equipment operation, where the multi-source data set includes structured data, semi-structured data, and unstructured data.

[0085] Step 320: extract knowledge from the semi-structured data and the unstructured data to obtain structured knowledge extraction results.

[0086] In specific application scenarios, in order to obtain accurate knowledge from these multi-source heterogeneous data, knowledge extraction is required. Specifically, .replace() in Python can be used to extract knowledge from semi-structured data and unstructured data to obtain structured knowledge extraction results.

[0087] Step 330: Integrate the knowledge extraction results with the structured data to obtain the knowledge graph text data.

[0088] In specific application scenarios, there may be data duplication, conflict or inconsistency between the knowledge extraction results and the structured data. Therefore, the duplicate data can be deleted through knowledge fusion. For data conflicts or inconsistencies, the standards can be graded according to the source of the standards (national standards are listed as national level, industry standards are listed as industry level, and internal enterprise standards and some other specifications are classified as enterprise level), and then the corresponding higher-level conflicting data is retained. Through the above-mentioned knowledge extraction and knowledge fusion methods, the accuracy and completeness of the knowledge graph text data can be ensured.

[0089] Step 340: Perform word segmentation on the knowledge graph text data based on the point mutual information algorithm to obtain knowledge graph entity data.

[0090] For the disclosed embodiment, the word segmentation process of the knowledge graph text data based on the point mutual information algorithm in step 340 of the embodiment to obtain the knowledge graph entity data may include the following steps:

[0091] Step 340-1, divide the knowledge graph text data into multiple single words.

[0092] Step 340-2: Combine multiple words based on the point mutual information algorithm to obtain word groups with dependency relationships in the knowledge graph text data.

[0093] For the disclosed embodiment, the pointwise mutual information (PMI) algorithm can be used to find mutually dependent single words and combine them into a word group. The specific PMI calculation formula is:

[0094]

[0095] In the formula, P(x) represents the probability of word x appearing alone in the text set, P(y) represents the probability of word y appearing alone in the text set, and P(x,y) represents the probability of word x and word y appearing adjacent to each other in the text set. PMI(x,y) represents the information entropy contained when two words x and y appear adjacent to each other. The larger the information entropy value calculated by PMI, the more information the two words contain, so the two adjacent words can be combined. This method can reduce the learning difficulty, that is, when the calculated PMI is less than a certain value, it is considered that the two adjacent words x and y do not contain information, that is, they do not need to be combined, thereby achieving text screening.

[0096] Through PMI, multiple words can be divided into single words and two-word words, that is, by calculating PMI, it is determined which single words can be combined with adjacent single words and contain corresponding information entropy. When the PMI calculated between a word and its two adjacent words on the left and right is too small, the word is divided into single words. When the calculated PMI of a word and its two adjacent words on the left and right are both greater than the minimum value, both are regarded as a combination. In the professional field of electric power, there are some professional words with more than two words, so they need to be integrated into one word through word aggregation. The degree of freedom of a word group is quantified through the method of information entropy. The calculation formula of information entropy is:

[0097] H(x)=-∑p i logp i

[0098] In the formula, p i It represents the probability of i occurring, that is, the probability of word or word group i appearing in the text.

[0099] The larger the left and right information entropy of a word group is, the more contexts this word group exists in, and it should be segmented. Otherwise, it should be aggregated. By introducing the left and right entropy product, we can determine whether two word groups can be combined to achieve the division of multi-word vocabulary. The left and right entropy product calculation formula is:

[0100] H(w)=H(w left )·H(w right )

[0101] In the formula, w represents a word group, w left 、w right Represents a word group consisting of the word group w and the entities on the left and right sides.

[0102] Correspondingly, for the embodiments of the present disclosure, the embodiment steps may include: calculating the information entropy between any two adjacent single words among multiple single words; combining two adjacent single words whose information entropy is greater than a first preset threshold value to obtain multiple word groups having a dependency relationship in the knowledge graph text data; calculating the left information entropy of each word group among the multiple word groups and the first entity on the left, and the right information entropy of each word group and the second entity on the right, wherein the first entity is the left word group or the left single word adjacent to the word group, and the second entity is the right word group or the right single word adjacent to the word group; when it is determined that the left-right entropy product of the left information entropy and the right information entropy is less than a second preset threshold value, combining the word group with the first entity and / or the second entity to form a new word group.

[0103] Step 340-3: Determine the uncombined single words and the combined word groups as knowledge graph entity data related to substation equipment.

[0104] Step 350: Based on the knowledge graph entity data, use the pre-trained triple model to extract triple information from the multi-source data set.

[0105] The triple information includes a head entity, a tail entity, and an entity relationship between the head entity and the tail entity.

[0106] In a specific application scenario, before using the pre-trained triple model to execute the steps of this embodiment, it is necessary to pre-train the triple model, and after judging that it meets the preset training standard, it is repeatedly applied to the triple information extraction task. Among them, the training method of the triple model includes: obtaining a sample text and a plurality of sample triples corresponding to the sample text, the sample triples include a sample head entity, a sample tail entity, and an entity relationship between the sample head entity and the sample tail entity related to the substation equipment; using the sample text as the input feature of the triple model, using the sample triple as the training label of the triple model, iteratively training the triple model until the loss function of the triple model is less than the third preset threshold, and judging that the training of the triple model is completed; wherein, during the training process, the triple model predicts the head entity based on the sample text, and then predicts the tail entity based on the predicted head entity, and then predicts the entity relationship between the predicted head entity and the predicted tail entity based on the predicted head entity and the predicted tail entity.

[0107] For the embodiments of the present disclosure, based on the knowledge graph entity data, a pre-trained triple model can be used to predict the head entity contained in the multi-source data set, and then the tail entity can be predicted based on the predicted head entity. Then, based on the predicted head entity and tail entity, the entity relationship between the head entity and the tail entity can be predicted.

[0108] Step 360: construct a knowledge graph for substation equipment operation evaluation based on the triple information.

[0109] Step 370: Determine the operation evaluation results of the target substation equipment according to the knowledge graph.

[0110] For the embodiments of the present disclosure, after obtaining the equipment operation data of the target substation equipment, an operation evaluation can be performed on the equipment operation data based on the knowledge graph to obtain an operation evaluation result of the target substation equipment.

[0111] In a specific application scenario, before executing the steps of this embodiment, as a preferred method, the knowledge graph can be embedded to obtain an embedded knowledge graph. Through the knowledge graph embedding method, the digitization of the substation equipment operation evaluation standard can be realized, the efficiency of the substation equipment operation evaluation can be further improved, the storage and query efficiency of the knowledge graph can be improved, and information related to specific equipment or evaluation indicators can be quickly located. Secondly, knowledge graph embedding helps to perform more accurate knowledge reasoning. In order to achieve better digitization of the core indicators of the substation equipment operation rating standard and business associations.

[0112] In order to further leverage the advantages of knowledge graphs, we added embedded knowledge graph technology to map entities and relationships to continuous vector space and retain the structural information and some semantic information of the knowledge graph. We use TransR to build mathematical models between entities and relationships in entity space and relationship space respectively. The schematic diagram is shown in the figure below. Figure 4 As shown, h and t exist in the entity space, and r exists in the relationship space. For each relationship, define the relationship matrix M r , embed h and t in the entity space into entities, map h and t into a relational space through the relational matrix, and obtain H and T. The specific formula is:

[0113] H=M,h,T=M r t

[0114] H+r≈T

[0115] Among them, the relationship matrix M r is randomly initialized. During the training process, the goal is to minimize the loss function. r to update.

[0116] The calculation formula of the loss function is:

[0117]

[0118] In the formula, L represents the loss function, H and T represent the entities h and t mapped to the relational space, and S represents the set of triples in the knowledge graph.

[0119] Accordingly, when embedding the knowledge graph to obtain an embedded knowledge graph, the following steps can be included: determining the set of triples in the knowledge graph, constructing a mathematical model between entities and entity relationships based on the set of triples, and the data model includes an entity space and an entity relationship space; based on the relationship matrix, mapping the head entity and the tail entity in the entity space to the entity relationship space to obtain an embedded knowledge graph.

[0120] As a preferred method, for the embodiment of the present disclosure, the operation evaluation result of the target substation equipment can be determined according to the embedded knowledge graph. Accordingly, when determining the operation evaluation result of the target substation equipment according to the knowledge graph in step 370 of the embodiment, the following steps can be included:

[0121] Step 370-1: Obtain equipment operation data of target substation equipment.

[0122] Step 370-2: Perform operation evaluation on the equipment operation data based on the embedded knowledge graph to obtain the operation evaluation results of the target substation equipment.

[0123] In summary, the technical solution in this application can perform text preprocessing on the multi-source data set after obtaining the multi-source data set of substation equipment operation to obtain knowledge graph entity data, and the knowledge graph entity data contains multiple entities related to substation equipment; then, based on the knowledge graph entity data, the pre-trained triple model can be used to extract triple information in the multi-source data set, and the triple information includes the head entity, the tail entity, and the entity relationship between the head entity and the tail entity; and based on the triple information, a knowledge graph for substation equipment operation evaluation is constructed; finally, the operation evaluation result of the target substation equipment is determined according to the knowledge graph. The technical solution in this application can mine and orderly express the complex substation equipment operation evaluation standard indicator data and information by constructing a knowledge graph, so that the complex standard system can become clear and easy to understand, and then it can quickly locate the information related to specific equipment and evaluation indicators, and provide decision support in a timely manner.

[0124] Based on the above Figure 2 , Figure 3 The detailed description of the substation equipment operation evaluation method based on knowledge graph is provided, such as Figure 5 As shown, Figure 5 is a structural block diagram of a substation equipment operation evaluation device based on a knowledge graph according to an exemplary embodiment. Figure 5 As shown, the device comprises:

[0125] The processing module 51 may be used to obtain a multi-source data set of substation equipment operation, and perform text preprocessing on the multi-source data set to obtain knowledge graph entity data, where the knowledge graph entity data includes multiple entities related to the substation equipment;

[0126] An extraction module 52, which can be used to extract triple information in a multi-source data set based on the knowledge graph entity data using a pre-trained triple model, where the triple information includes a head entity, a tail entity, and an entity relationship between the head entity and the tail entity;

[0127] A construction module 53, which can be used to construct a knowledge graph for substation equipment operation evaluation based on triple information;

[0128] The determination module 54 may be used to determine the operation evaluation result of the target substation equipment according to the knowledge graph.

[0129] In some embodiments of the present application, the multi-source data set includes structured data, semi-structured data and unstructured data; the processing module 51 can be specifically used to extract knowledge from semi-structured data and unstructured data to obtain structured knowledge extraction results; integrate the knowledge extraction results with the structured data to obtain knowledge graph text data; perform word segmentation processing on the knowledge graph text data based on the point mutual information algorithm to obtain knowledge graph entity data.

[0130] In some embodiments of the present application, when the knowledge graph text data is segmented based on the point mutual information algorithm to obtain the knowledge graph entity data, the processing module 51 can be specifically used to divide the knowledge graph text data into multiple single words; combine multiple single words based on the point mutual information algorithm to obtain word groups with dependency relationships in the knowledge graph text data; and determine the uncombined single words and the combined word groups as knowledge graph entity data related to substation equipment.

[0131] In some embodiments of the present application, when a plurality of words are combined based on the point mutual information algorithm to obtain word groups with dependency relationships in the knowledge graph text data, the processing module 51 can be specifically used to calculate the information entropy between any two adjacent words among the plurality of words; combine two adjacent words whose information entropy is greater than a first preset threshold to obtain a plurality of word groups with dependency relationships in the knowledge graph text data; calculate the left information entropy of each word group among the plurality of word groups and the first entity on the left, and the right information entropy of each word group and the second entity on the right, wherein the first entity is a left word group or a left word adjacent to the word group, and the second entity is a right word group or a right word adjacent to the word group; when it is determined that the left-right entropy product of the left information entropy and the right information entropy is less than a second preset threshold, combine the word group with the first entity and / or the second entity to form a new word group.

[0132] In some embodiments of the present application, Figure 6 As shown, the device further includes: a training module 55;

[0133] The training module 55 can be used to obtain sample text and multiple sample triplets marked corresponding to the sample text, wherein the sample triples include a sample head entity, a sample tail entity, and an entity relationship between the sample head entity and the sample tail entity related to the substation equipment; the sample text is used as the input feature of the triple model, and the sample triple is used as the training label of the triple model, and the triple model is iteratively trained until the loss function of the triple model is less than a third preset threshold, and the training of the triple model is judged to be completed; wherein, during the training process, the triple model predicts the head entity based on the sample text, and then predicts the tail entity based on the predicted head entity, and then predicts the entity relationship between the predicted head entity and the predicted tail entity based on the predicted head entity and the predicted tail entity.

[0134] In some embodiments of the present application, Figure 6 As shown, the device further includes: a mapping module 56;

[0135] The construction module 53 can also be used to determine a set of triples in the knowledge graph, and to construct a mathematical model between entities and entity relationships based on the set of triples, wherein the data model includes an entity space and an entity relationship space;

[0136] The mapping module 56 can be used to map the head entity and the tail entity in the entity space to the entity relationship space based on the relationship matrix to obtain an embedded knowledge graph.

[0137] In some embodiments of the present application, the determination module 54 can be specifically used to obtain equipment operation data of the target substation equipment; perform operation evaluation on the equipment operation data based on the embedded knowledge graph to obtain the operation evaluation result of the target substation equipment.

[0138] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0139] In the embodiments of the present application, after obtaining a multi-source data set of substation equipment operation, the multi-source data set can be pre-processed with text to obtain knowledge graph entity data, which includes multiple entities related to substation equipment; then, based on the knowledge graph entity data, a pre-trained triple model can be used to extract triple information in the multi-source data set, the triple information includes a head entity, a tail entity, and an entity relationship between the head entity and the tail entity; and a knowledge graph for substation equipment operation evaluation can be constructed based on the triple information; finally, the operation evaluation result of the target substation equipment is determined according to the knowledge graph. The technical solution in the present application can mine and orderly express the complex substation equipment operation evaluation standard indicator data and information by means of a knowledge graph, so that the complex standard system can become clear and easy to understand, and then it can quickly locate the information related to specific equipment and evaluation indicators, and provide decision support in a timely manner.

[0140] In the above, the knowledge graph-based substation equipment operation evaluation device of the embodiment of the present invention is described from the perspective of the functional module in combination with the accompanying drawings. It should be understood that the functional module can be implemented in hardware form, can be implemented by software instructions, and can also be implemented by a combination of hardware and software modules. Specifically, the steps of the embodiment of the substation equipment operation evaluation method based on the knowledge graph in the embodiment of the present invention can be completed by the hardware integrated logic circuit and / or software instructions in the processor, and the steps of the substation equipment operation evaluation method based on the knowledge graph applied in combination with the embodiment of the present invention can be directly embodied as a hardware decoding processor to execute, or a combination of hardware and software modules in the decoding processor to execute. Optionally, the software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory, and the processor reads the information in the memory, and completes the steps in the above-mentioned embodiment of the substation equipment operation evaluation method based on the knowledge graph in combination with its hardware.

[0141] Figure 7 It is a schematic block diagram of an electronic device 700 according to an embodiment of the present invention.

[0142] like Figure 7 As shown, the electronic device 700 may include:

[0143] The memory 710 and the processor 720, the memory 710 is used to store the computer program and transmit the program code to the processor 720. In other words, the processor 720 can call and run the computer program from the memory 710 to implement the method in the embodiment of the present invention.

[0144] For example, the processor 720 may be configured to execute the above method embodiments according to instructions in the computer program.

[0145] In some embodiments of the present invention, the processor 720 may include but is not limited to:

[0146] General-purpose processor, digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc.

[0147] In some embodiments of the present invention, the memory 710 includes but is not limited to:

[0148] Volatile memory and / or non-volatile memory. Among them, the non-volatile memory can be read-only memory (ROM), programmable ROM (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM) or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM) and direct memory bus random access memory (Direct Rambus RAM, DR RAM).

[0149] In some embodiments of the present invention, the computer program may be divided into one or more modules, which are stored in the memory 710 and executed by the processor 720 to complete the method provided by the present invention. The one or more modules may be a series of computer program instruction segments capable of completing specific functions, and the instruction segments are used to describe the execution process of the computer program in the controller.

[0150] like Figure 7 As shown, the electronic device 700 may further include:

[0151] The transceiver 730 may be connected to the processor 720 or the memory 710 .

[0152] The processor 720 may control the transceiver 730 to communicate with other devices, specifically, to send data to other devices or receive data sent by other devices. The transceiver 730 may include a transmitter and a receiver. The transceiver 730 may further include an antenna, and the number of antennas may be one or more.

[0153] It should be understood that the various components in the electronic device are connected via a bus system, wherein the bus system includes not only a data bus but also a power bus, a control bus and a status signal bus.

[0154] The present invention also provides a computer storage medium having a computer program stored thereon, which, when executed by a computer, enables the computer to perform the method of the above method embodiment. In other words, an embodiment of the present invention also provides a computer program product containing instructions, which, when executed by a computer, enables the computer to perform the method of the above method embodiment.

[0155] When software is used for implementation, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, a computer, a server or a data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or a data center that includes one or more available media integration. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a digital video disc (Digital Video Disc, DVD)), or a semiconductor medium (e.g., a solid-state drive (Solid State Disk, SSD)), etc.

[0156] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments applied herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. 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.

[0157] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the module is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0158] The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. For example, each functional module in each embodiment of the present application may be integrated into a processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0159] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in this application, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for evaluating substation equipment operation based on knowledge graph, characterized in that: include: Acquire a multi-source data set of substation equipment operation, and perform text preprocessing on the multi-source data set to obtain knowledge graph entity data, wherein the knowledge graph entity data includes multiple entities related to the substation equipment; Based on the knowledge graph entity data, a pre-trained triple model is used to extract triple information in the multi-source data set, where the triple information includes a head entity, a tail entity, and an entity relationship between the head entity and the tail entity; Constructing a knowledge graph for substation equipment operation evaluation based on the triple information; The operation evaluation results of the target substation equipment are determined according to the knowledge graph.

2. The method according to claim 1, characterized in that The multi-source data set includes structured data, semi-structured data and unstructured data; The text preprocessing of the multi-source data set to obtain knowledge graph entity data includes: Performing knowledge extraction on the semi-structured data and the unstructured data to obtain structured knowledge extraction results; Fusion of the knowledge extraction result and the structured data to obtain knowledge graph text data; The knowledge graph text data is segmented based on the point mutual information algorithm to obtain knowledge graph entity data.

3. The method according to claim 2, characterized in that The word segmentation processing of the knowledge graph text data is performed based on the point mutual information algorithm to obtain the knowledge graph entity data, including: Dividing the knowledge graph text data into multiple words; Combining the multiple words based on the point mutual information algorithm to obtain word groups with dependency relationships in the knowledge graph text data; The uncombined single words and the combined word groups are determined as knowledge graph entity data related to substation equipment.

4. The method according to claim 3, characterized in that The step of combining the multiple words based on the point mutual information algorithm to obtain word groups having dependency relationships in the knowledge graph text data includes: Calculating the information entropy between any two adjacent words in the plurality of words; Combining two adjacent words whose information entropy is greater than a first preset threshold value to obtain a plurality of word groups having a dependency relationship in the knowledge graph text data; Calculating the left information entropy of each word group and a first entity on the left, and the right information entropy of each word group and a second entity on the right, wherein the first entity is a left word group or a left single word adjacent to the word group, and the second entity is a right word group or a right single word adjacent to the word group; When it is determined that the left-right entropy product of the left information entropy and the right information entropy is less than a second preset threshold, the word group is combined with the first entity and / or the second entity to form a new word group.

5. The method according to claim 1, characterized in that Also included is a training method for the triplet model: Acquire a sample text and a plurality of sample triples marked corresponding to the sample text, wherein the sample triples include a sample header entity and a sample tail entity related to substation equipment, and an entity relationship between the sample header entity and the sample tail entity; Using the sample text as an input feature of the triple model, using the sample triplet as a training label of the triple model, iteratively training the triple model until the loss function of the triple model is less than a third preset threshold, and determining that the training of the triple model is completed; Among them, during the training process, the triple model predicts the head entity based on the sample text, and then predicts the tail entity based on the predicted head entity, and then predicts the entity relationship between the predicted head entity and the predicted tail entity based on the predicted head entity and the predicted tail entity.

6. The method according to claim 1, characterized in that After constructing a knowledge graph for substation equipment operation evaluation based on the triple information, the method further includes: Determine a set of triples in the knowledge graph, and construct a mathematical model between entities and entity relationships based on the set of triples, wherein the data model includes an entity space and an entity relationship space; Based on the relationship matrix, the head entity and the tail entity in the entity space are mapped to the entity relationship space to obtain an embedded knowledge graph.

7. The method according to claim 6, characterized in that Determining the operation evaluation result of the target substation equipment according to the knowledge graph includes: Obtain equipment operation data of target substation equipment; An operation evaluation is performed on the equipment operation data based on the embedded knowledge graph to obtain an operation evaluation result of the target substation equipment.

8. A substation equipment operation evaluation device based on knowledge graph, characterized in that: include: A processing module, used to obtain a multi-source data set of substation equipment operation, and perform text preprocessing on the multi-source data set to obtain knowledge graph entity data, wherein the knowledge graph entity data includes multiple entities related to the substation equipment; An extraction module, configured to extract triple information in the multi-source data set based on the knowledge graph entity data and using a pre-trained triple model, wherein the triple information includes a head entity, a tail entity, and an entity relationship between the head entity and the tail entity; A construction module, used to construct a knowledge graph for substation equipment operation evaluation based on the triple information; A determination module is used to determine the operation evaluation results of the target substation equipment based on the knowledge graph.

9. An electronic device, characterized in that: include: A processor and a memory, the memory being used to store a computer program, and the processor being used to call and run the computer program stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, wherein the computer program enables a computer to execute the method according to any one of claims 1 to 7.

Images