Method for processing insulation status data of power equipment

The method addresses inefficiencies in relational databases by converting unstructured power equipment data into a graph database through named entity recognition and coreference resolution, improving data processing efficiency and scalability.

CN112559739BActive Publication Date: 2024-12-03GUANGZHOU POWER SUPPLY BUREAU GUANGDONG POWER GRID CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202011384232.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-12-01
Publication Date
2024-12-03
Estimated Expiration
2040-12-01

AI Technical Summary

Technical Problem

Traditional relational databases struggle with processing complex relationships in large-scale, unstructured power equipment insulation status data, leading to reduced efficiency, high memory consumption, and poor data scalability, especially when handling text data.

Method used

A method involving named entity recognition, coreference resolution, data classification, and normalization to convert unstructured insulation status data into a graph database for efficient data aggregation and knowledge query.

Benefits of technology

Enhances data processing efficiency, reduces manual workload, and improves data scalability and sharing by utilizing natural language processing and graph databases for power equipment insulation status data management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112559739B_ABST
    Figure CN112559739B_ABST
Patent Text Reader

Abstract

The present application relates to a method for processing insulation status data of electric equipment. The method comprises: obtaining text data of insulation status of electric equipment, and identifying and extracting named entities of the text data of insulation status of electric equipment to obtain entity vocabulary; performing co-reference resolution processing on the entity vocabulary to determine the final entity library; classifying the association relationship between entity vocabulary in the final entity library to obtain classified data; normalizing the classified data, and outputting normalized data for importing into a graph database. The use of this method can reduce the manual workload of the data processing process, has good applicability, and greatly increases the processing efficiency of insulation status data of electric equipment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of power operation and maintenance management, and in particular to a method for processing insulation status data of power equipment. Background Art

[0002] Partial discharge is an electrical discharge in which the insulation between conductors is only partially bridged. It occurs in power equipment with internal insulation faults. Partial discharge accelerates the insulation degradation inside power equipment and affects the safe and stable operation of the power system. The relevant detection and research of partial discharge is of great significance. During the partial discharge detection process at the substation, the equipment status is usually recorded in the form of a test report, which will accumulate a large amount of unstructured power equipment insulation status data.

[0003] At present, the data in the power grid is characterized by being diverse, large-scale, and highly correlated, and higher requirements are placed on data processing technology. Faced with a wide variety of unstructured data, traditional relational databases have exposed their inherent disadvantage of being slow when processing complex relationships. Because they use association tables to store the relationships between entities, when the number of entities increases and the relationships become complex, the efficiency of relational databases will be greatly reduced, and a lot of memory consumption will also be caused. In addition, relational databases have poor data scalability and sharing. Due to the large amount of text data, a lot of labor costs will also be incurred when processing the information contained in the text data.

[0004] During the implementation process, the inventors found that there are at least the following problems in the traditional technology: Currently, the storage of insulation status information of power equipment mainly uses traditional relational databases, and the relationship between the time of each entity is stored through an associated table. It is necessary to manually organize the insulation status text data of the power equipment accumulated in the data center, and then import the data into the relational database, and the processing efficiency is very low. Summary of the invention

[0005] Based on this, it is necessary to provide a method for processing insulation status data of power equipment in response to the above technical problems, including:

[0006] Acquire the text data of the insulation status of the power equipment, and perform named entity recognition and extraction on the text data of the insulation status of the power equipment to obtain entity vocabulary;

[0007] Perform coreference resolution on entity vocabulary to determine the final entity library;

[0008] Classify the associations between entity words in the final entity library to obtain classified data;

[0009] Normalize the classified data and output the normalized data for importing into the graph database.

[0010] In one of the embodiments, the electric power equipment insulation status text data is unstructured text data;

[0011] The steps of identifying and extracting named entities from the text data of insulation status of power equipment to obtain entity vocabulary include:

[0012] Convert the insulation status text data of the power equipment into text form and perform script processing to obtain the data to be processed;

[0013] Based on the auxiliary dictionary, a preset statistical model is used to perform text segmentation and delete stop words on the data to be processed to obtain the segmentation result;

[0014] Screen the word segmentation results to obtain entity vocabulary.

[0015] In one embodiment, the preset statistical model includes a hidden Markov model; the auxiliary dictionary includes any one or any combination of the following dictionaries: a power system professional dictionary, a stop word dictionary, and a common word dictionary.

[0016] In one embodiment, the step of performing coreference resolution processing on entity vocabulary to determine a final entity library includes:

[0017] According to the preset word vector dimension, determine the word vector of each entity vocabulary;

[0018] Obtaining the multiplication results of the entity vocabulary; the multiplication result is obtained by multiplying the normalized word vector of the entity vocabulary with the word vector of another entity vocabulary;

[0019] Sort the multiplication results, and determine the synonyms of the entity vocabulary according to the sorting results;

[0020] Filter and delete synonyms in each synonym, update the word vector of each entity word, and obtain the final entity library.

[0021] In one embodiment, the association relationship includes one or more of a containment relationship, an instance relationship, an attribute relationship, and no relationship;

[0022] The steps of classifying the associations between entity words in the final entity library to obtain classified data include:

[0023] Perform difference processing on the word vectors of any two entity words in the final entity library to obtain the difference processing result;

[0024] Based on the difference processing results, a classifier is used to learn, classify and filter the associations between entity words in the final entity library to obtain classified data.

[0025] In one embodiment, the step of normalizing the classified data and outputting the normalized data for importing into a graph database includes:

[0026] The entity words, the associations between entity words, and the attributes of entity words in the classified data are written into the corresponding format files according to the specified format of the corresponding graph database to complete the data normalization processing and output the normalized data for batch import into the corresponding graph database.

[0027] In one embodiment, the graph database includes a Neo4j graph database.

[0028] A device for processing insulation status data of electric power equipment, the device comprising:

[0029] The recognition module is used to obtain the text data of the insulation status of the power equipment, and to recognize and extract the named entities in the text data of the insulation status of the power equipment to obtain entity vocabulary;

[0030] The coreference resolution module is used to perform coreference resolution on entity vocabulary and determine the final entity library;

[0031] The classification module is used to classify the associations between entity words in the final entity library to obtain classified data;

[0032] The normalization processing module is used to normalize the classified data and output the normalized data for importing into the graph database.

[0033] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0034] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0035] One of the above technical solutions has at least the following advantages and beneficial effects:

[0036] The present application obtains the insulation status text data of the substation power equipment to identify and extract named entities, performs co-reference resolution on the obtained entity vocabulary, and classifies the association relationship between the entity vocabulary after the co-reference resolution, obtains the classified data and performs normalization processing, and outputs the normalized data for importing into the graph database; thus, the present application realizes data aggregation in the field of partial discharge, constructs an efficient and convenient data matching and knowledge query method, has good scalability and sharing, reduces the manual workload of the data processing process, has good applicability, and greatly increases the processing efficiency of the insulation status data of the power equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the conventional technology, the drawings required for use in the embodiments or the conventional technology descriptions are 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.

[0038] Figure 1 It is a flowchart of a method for processing insulation status data of power equipment in one embodiment;

[0039] Figure 2 A flowchart of a step of obtaining entity vocabulary in one embodiment;

[0040] Figure 3 A schematic diagram of a process for determining a final entity library step in an embodiment;

[0041] Figure 4 A flow chart of the entity library established for an example;

[0042] Figure 5 A schematic diagram of a flow chart of a step of obtaining classified data in one embodiment;

[0043] Figure 6 A schematic diagram of the process of implementing graph database storage for an entity library in an example;

[0044] Figure 7 A flowchart of a method for processing insulation status data of power equipment in an example;

[0045] Figure 8 The figure is a structural block diagram of a device for processing insulation status data of electric power equipment in one embodiment. DETAILED DESCRIPTION

[0046] In order to facilitate understanding of the present application, the present application will be described more fully below with reference to the relevant drawings. Embodiments of the present application are provided in the drawings. However, the present application can be implemented in many different forms and is not limited to the embodiments described herein. On the contrary, the purpose of providing these embodiments is to make the disclosure of the present application more thorough and comprehensive.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application.

[0048] When used herein, the singular forms "a", "an", and "said / the" may also include plural forms, unless the context clearly indicates otherwise. It should also be understood that the terms "include / comprise" or "have" and the like specify the presence of stated features, wholes, steps, operations, components, parts, or combinations thereof, but do not exclude the possibility of the presence or addition of one or more other features, wholes, steps, operations, components, parts, or combinations thereof. At the same time, the term "and / or" used in this specification includes any and all combinations of the relevant listed items.

[0049] In one embodiment, a method for processing insulation status data of power equipment is provided. This embodiment uses the method applied to a terminal as an example for illustration. It is understandable that the method can also be applied to a server, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. The terminal can be, but is not limited to, various personal computers, laptops, smart phones, and tablet computers, etc. The server can be implemented as an independent server or a server cluster composed of multiple servers. In this embodiment, Figure 1 As shown, the method may include the following steps:

[0050] Step 202, obtaining the text data of the insulation status of the power equipment, and performing named entity recognition and extraction on the text data of the insulation status of the power equipment to obtain entity vocabulary;

[0051] Among them, during the local discharge detection process at the substation, the status of the power equipment is usually recorded in the form of a test report, thereby accumulating a large amount of unstructured power equipment insulation status data; named entities include personal names, organization names, place names, and all other entities identified by names, and broader entities also include numbers, dates, currencies, addresses, etc.

[0052] Specifically, the insulation status text data of the power equipment is obtained, and the natural language processing technology is used to automatically identify the named entities in the insulation status text data of the power equipment, and the identified named entities are extracted to complete the recognition and extraction of entity vocabulary in the corpus.

[0053] Step 204, performing coreference resolution processing on entity vocabulary to determine a final entity library;

[0054] Among them, the co-reference phenomenon is the phenomenon of using pronouns, titles and abbreviations to refer to the full names of the entities mentioned above in order to avoid repetition. The so-called co-reference resolution is the process of merging different descriptions of the same entity in the real world. Co-reference resolution is very critical for natural language processing, machine translation, information extraction, information retrieval and other fields, and plays the role of hyperlinks in natural language. In the process of knowledge fusion, co-reference resolution and entity disambiguation are usually required to remove synonyms in the entity library and distinguish between synonyms in the entity library. For highly professional domain knowledge graphs, there are usually no synonyms.

[0055] Specifically, the insulation status text data of the power equipment is obtained and the named entities are recognized and extracted. After the entity vocabulary is obtained, the entity vocabulary is subjected to coreference resolution and entity disambiguation processing, the synonyms in the entity vocabulary are removed, and the synonyms in the entity vocabulary are distinguished, so as to update the entity library to obtain the final entity library.

[0056] Step 206, classifying the associations between entity words in the final entity library to obtain classified data;

[0057] After obtaining the discrete entity words in the final entity library, in order to obtain semantic information, it is also necessary to extract the association relationship between the entity words from the corpus, and connect the discrete entity words through the association relationship to form a network knowledge structure. The association relationship between entity words includes one or more of the following: inclusion relationship, instance relationship, attribute relationship, and no relationship.

[0058] Specifically, when the final entity library is obtained, it is necessary to classify the associations between entity words in the final entity library based on the associations between entity words in the final entity library, that is, the network knowledge structure formed by the associations between entity words and entity words, to obtain classified data, from which the semantic information constructed between entity words can be obtained.

[0059] Step 208, normalize the classified data and output the normalized data for importing into the graph database.

[0060] Among them, graph database is a type of NoSQL (Not Only Structured Query Language) database, which is a non-relational database that uses graph theory to store relationship information between entities. Relational databases are not very effective in storing "relational" data. Their queries are complex, slow, and beyond expectations, and the unique design of graph databases just makes up for this defect. That is, compared with the various association tables in relational databases, the relationships in graph databases can provide a richer relationship display method through the function that relationships can contain attributes. Graph databases are specially used to process large-scale, multi-category, and highly correlated data, which can greatly increase the processing efficiency of insulation status data of power equipment. Graph databases include Neo4j graph database, FlockDB graph database, AllegroGrap graph database, GraphDB graph database, InfiniteGraph graph database, and HugeGraph graph database.

[0061] In one example, the classified data is normalized and the normalized data can be output for importing into a Neo4j graph database.

[0062] Specifically, the insulation status data of power equipment is obtained, and the named entities are identified and extracted. After the entity vocabulary is obtained, the entity vocabulary is subjected to coreference resolution, the synonyms in the entity library are removed, and the synonyms in the entity library are distinguished, so as to determine the final entity library; in order to obtain semantic information, the association relationships between the entity vocabulary in the final entity library, such as inclusion relationship, instance relationship, attribute relationship and no relationship, are classified. When the classified data is obtained, the classified data is normalized so that the normalized data can be output for importing into the Neo4j graph database.

[0063] The method for processing the insulation status data of electric power equipment of the present application performs named entity recognition and extraction and co-reference resolution processing on a large amount of unstructured insulation status data of electric power equipment accumulated in a substation, determines a final entity library, classifies the association relationship between entity words in the final entity library to obtain classified data, and normalizes the classified data to output normalized data for importing into a graph database; that is, the present application uses natural language processing and machine learning methods to construct an entity library of insulation status data of electric power equipment, and uses a graph database to store data of the insulation status of electric power equipment. The advantages of the graph database lie in the processing of unstructured batch data, the query and calculation of complex relationships, and a more intuitive display, and the data stored on the web (World Wide Web) is easier to expand and share; thereby, an efficient data matching and knowledge query method is constructed, which has good scalability and sharing, and improves the efficiency of data processing.

[0064] In a specific embodiment, the electric power equipment insulation status text data may be unstructured text data;

[0065] like Figure 2 As shown, the steps of identifying and extracting named entities from the text data of insulation status of electric power equipment to obtain entity vocabulary may include:

[0066] Step 302, converting the insulation status text data of the power equipment into text form and performing script processing to obtain data to be processed;

[0067] Step 304, based on the auxiliary dictionary, using a preset statistical model to perform text segmentation and delete stop words on the data to be processed to obtain a segmentation result;

[0068] Step 306, screening the word segmentation results to obtain entity vocabulary.

[0069] In a specific example, the preset statistical model includes a hidden Markov model; the auxiliary dictionary includes any one or any combination of the following dictionaries: a power system professional dictionary, a stop word dictionary, and a common word dictionary.

[0070] Among them, the text data of the insulation status of power equipment can be unstructured text data. Unstructured data is data with irregular or incomplete data structure, no predefined data model, and is not convenient to be represented by a two-dimensional logical table of a database. It includes office documents, texts, pictures, various reports, images, and audio / video information in all formats. Unstructured text data is unstructured data in the form of text (such as characters, numbers, punctuation, various printable symbols, etc.).

[0071] Specifically, the insulation status text data of the power equipment is obtained, the text data to be processed is converted into text form, and imported into Python for script processing to obtain the data to be processed; the power system professional dictionary, stop word dictionary and common word dictionary are imported or directly used as auxiliary, and the hidden Markov model is used for text segmentation. During the word segmentation, the power system professional dictionary and the common word dictionary are matched to perform text segmentation on the data to be processed, and the stop word dictionary is matched to delete the stop words to obtain the word segmentation result; the word segmentation result is reviewed and screened to obtain the entity vocabulary.

[0072] This application utilizes natural language processing methods to automatically identify named entities in text data, uses auxiliary dictionaries such as power system professional dictionaries, stop word dictionaries, and common word dictionaries as auxiliary, and performs text segmentation based on hidden Markov models, so as to achieve more accurate recognition and extraction of named entities in text data on the insulation status of power equipment.

[0073] In a specific embodiment, Figure 3 As shown, the steps of performing coreference resolution processing on entity vocabulary and determining the final entity library may include:

[0074] Step 402, confirming the word vector of each entity vocabulary according to the preset word vector dimension;

[0075] Step 404, obtaining the multiplication results of the entity vocabulary; the multiplication result is obtained by multiplying the normalized word vector of the entity vocabulary with the word vector of another entity vocabulary;

[0076] Step 406, sorting the multiplication results, and determining the synonyms of the entity vocabulary according to the sorting results;

[0077] Step 408, filter and delete synonymous words in each synonym, update the word vector of each entity word, and obtain the final entity library.

[0078] Among them, word embedding, also known as word embedded natural language processing (NLP), is a general term for a set of language modeling and feature learning techniques in which words or phrases from the vocabulary are mapped to vectors of real numbers. Conceptually, it involves mathematical embedding from a one-dimensional space for each word to a continuous vector space with a lower dimension. The numerical value of the word vector reflects the characteristics of the vocabulary in the corpus, and entity words with similar meanings have relatively similar word vectors.

[0079] Specifically, when entity vocabulary is obtained, the word vector of each entity vocabulary is confirmed according to the preset word vector dimension. For example, the preset word vector dimension can be 300 dimensions, then the word vector of each entity vocabulary can be obtained and confirmed; the word vectors of any two entity vocabulary are normalized and multiplied. The larger the result obtained, the closer the word vectors of the two entity vocabulary are, that is, the closer the meanings of the two entity vocabulary are; the multiplication results obtained for each entity vocabulary are sorted, and the synonyms of each entity vocabulary can be determined according to the sorting results. For example, the multiplication results obtained for each entity vocabulary can be sorted from large to small, and the top ten entity vocabulary ranked with the multiplication results of each entity vocabulary, that is, the ten entity vocabulary closest to each entity vocabulary, are recorded, confirmed as synonyms of the entity vocabulary and stored; the synonyms of each entity vocabulary are further screened to obtain synonyms that need to be coreference resolved, the synonyms of each entity vocabulary are deleted from the entity library, the word vectors of each entity vocabulary are recalculated and updated for storage, and the final entity library is obtained.

[0080] The present application obtains the multiplication results by normalizing the word vectors of the entity vocabulary and then multiplying them, and sorts the multiplication results to determine the synonyms of the entity vocabulary, and then performs coreference resolution on the synonyms, deletes the synonyms to obtain the final entity library, that is, in the process of knowledge fusion, coreference resolution and entity disambiguation are performed, thereby ensuring that the extraction of entity vocabulary is more accurate, enhancing the readability of the text, and is also conducive to mining more useful entity information and attribute information.

[0081] In one example, if Figure 4 The figure shows the flow chart of establishing the entity library. The specific process is as follows: obtain the insulation status data of the equipment, convert it into text form and import it into Python, import the power professional dictionary, stop word dictionary and common word dictionary as auxiliary dictionaries, complete the word segmentation based on the hidden Markov model, review the required entity vocabulary after obtaining the word segmentation result, obtain the entity library, and obtain the word vector data of the entity vocabulary; then normalize the word vectors of the entity vocabulary in the entity library and multiply them two by two to obtain the synonyms of the entity vocabulary, review the synonyms, delete the synonyms in the synonyms, calculate and update the word vector data again, and update the entity library to obtain the final entity library.

[0082] In a specific embodiment, Figure 5 As shown, the steps of classifying the associations between entity words in the final entity library to obtain classified data may include:

[0083] Step 502, performing difference processing on the word vectors of any two entity words in the final entity library to obtain a difference processing result;

[0084] Step 504: Based on the difference processing result, a classifier is used to learn, classify and filter the association relationship between entity words in the final entity library to obtain classified data.

[0085] Among them, the classifier can adopt a semi-supervised learning classifier.

[0086] Specifically, after obtaining the final entity library, it is necessary to perform difference processing on the word vectors of any two entity words in the final entity library. The difference can reflect the degree of relationship between the two entity words, and the difference result is obtained and stored; based on the difference result of the entity words, several groups of entity word vectors can be preset as labeled samples for the relationship between each entity word. The preset number can be 30 groups. Since the number of training samples is small, a semi-supervised learning classifier can be used to learn and classify the association relationship between entity words and further screen them to obtain classified data; and the entity words are stored in the form of "entity-relationship-entity" to facilitate import into the graph database.

[0087] This application performs difference processing on the word vectors of any two entity words in the final entity library, and uses a semi-supervised learning classifier to learn and classify the association relationship between the entity words based on the difference results, and obtains classified data, thereby accurately extracting the association relationship between entity words from the corpus and obtaining accurate semantic information.

[0088] In a specific embodiment, the step of normalizing the classified data and outputting the normalized data for importing into a graph database includes:

[0089] The entity words, the associations between entity words, and the attributes of entity words in the classified data are written into the corresponding format files according to the specified format of the corresponding graph database to complete the data normalization processing and output the normalized data for batch import into the corresponding graph database.

[0090] The graph database may adopt the Neo4j graph database, the prescribed format of the corresponding graph database is the prescribed format of the Neo4j graph database, and the corresponding format file may be a csv (Comma-Separated Values, comma separated value file format) file.

[0091] Specifically, when the classified data is obtained, the Neo4j graph database can be used to store the networked data, that is, the entity words in the classified data, the relationship between the entity words, and the attributes of the entity words are written into different csv files according to the specified format of the Neo4j graph database, and the data is normalized, and the normalized data is output for batch import into the Neo4j graph database.

[0092] In one example, if Figure 6 The figure shows the flow chart of graph database storage implemented by entity library. The general process is as follows: after the entity library is established, the relationship types between entity words in the entity library are defined, and labeled samples are made for the relationship between each entity word. The word vectors of any two entity words in the entity library are calculated for difference, and the relationship classification based on the semi-supervised learning classifier is completed. The entity words, the attributes of entity words, and the associations between entity words are then normalized. Finally, they are written into a csv file in the format specified by the Neo4j graph database, and the normalized data for batch import into the Neo4j graph database is output to complete the data storage.

[0093] like Figure 7 The figure shows a flow chart of the method for processing the insulation status data of the power equipment of the present application in an example. The general process is to obtain the insulation status data of the power equipment and then identify the named entities thereof, identify and extract the entity vocabulary and then perform co-reference resolution to establish an entity library, classify and learn the associations between the entity vocabulary in the entity library to obtain the classified data, and perform normalized data processing on the entity vocabulary of the classified data, the associations between the entity vocabulary and the attributes of the entity vocabulary, so that the normalized data can be output for importing into the graph database for storage and use.

[0094] As mentioned above, this application will use natural language processing methods to complete the automatic recognition of named entities in text data, and use auxiliary dictionaries such as power system professional dictionaries, stop word dictionaries, and common word dictionaries as auxiliary, and perform text segmentation based on hidden Markov models, which can realize the accurate recognition and extraction of named entities; and perform co-reference elimination processing on the extracted entity vocabulary, which ensures the accuracy of the extraction of entity vocabulary, and is also conducive to mining more useful entity information and attribute information, and constructing an entity library of power equipment insulation status data; using a semi-supervised learner to learn and classify the association between entity vocabulary, it can deeply mine the relationship between entity vocabulary; output standardized data for batch import into graph databases, which can intelligently store power equipment insulation status data. This application realizes data aggregation in the field of partial discharge, builds an efficient and convenient data matching and knowledge query platform, has good scalability, sharing and applicability, reduces the manual workload in the data processing process, and greatly improves the efficiency of data processing.

[0095] It should be understood that although Figure 1-Figure 7 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1-Figure 7At least part of the steps may include multiple steps or multiple stages. These steps or stages are not necessarily performed at the same time, but can be performed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed in turn or alternately with other steps or at least part of the steps or stages in other steps.

[0096] In one embodiment, Figure 8 As shown, a device for processing insulation status data of electric power equipment is provided, and the device may include:

[0097] The recognition module 110 is used to obtain the text data of the insulation status of the power equipment, and to recognize and extract the named entities in the text data of the insulation status of the power equipment to obtain entity vocabulary;

[0098] A coreference resolution module 120 is used to perform coreference resolution processing on entity vocabulary to determine a final entity library;

[0099] The classification module 130 is used to classify the association relationship between entity words in the final entity library to obtain classified data;

[0100] The normalization processing module 140 is used to perform normalization processing on the classified data and output the normalized data for importing into the graph database.

[0101] For the specific limitations of the insulation status data processing device for electric equipment, please refer to the limitations of the insulation status data processing method for electric equipment mentioned above, which will not be repeated here. Each module in the above-mentioned insulation status data processing device for electric equipment can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned 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. It should be noted that the division of modules in the embodiment of the present application is schematic and is only a logical function division. There may be other division methods in actual implementation.

[0102] In one embodiment, a computer device is provided, which may include a memory and a processor. The memory stores a computer program, and the processor implements the steps in the above method embodiments when executing the computer program.

[0103] In one embodiment, a computer-readable storage medium is further provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0104] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0105] In the description of this specification, the description with reference to the terms "some embodiments", "other embodiments", "ideal embodiments", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic descriptions of the above terms do not necessarily refer to the same embodiment or example.

[0106] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0107] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for processing insulation status data of electric power equipment, characterized in that: include: Acquire the text data of the insulation status of the electric power equipment, and perform named entity recognition and extraction on the text data of the insulation status of the electric power equipment to obtain entity vocabulary; The entity vocabulary is subjected to coreference resolution processing to determine a final entity library, including: According to the preset word vector dimension, confirm the word vector of each entity vocabulary; obtain each multiplication result of the entity vocabulary; the multiplication result is obtained by normalizing the word vector of the entity vocabulary and the word vector of another entity vocabulary and multiplying them; sort the multiplication results, and determine the synonyms of the entity vocabulary according to the sorting results; filter and delete the synonyms in each synonym, update the word vector of each entity vocabulary, and obtain the final entity library; Classify the associations between entity words in the final entity library to obtain classified data, including: Performing difference processing on the word vectors of any two entity words in the final entity library to obtain a difference processing result; based on the difference processing result, using a classifier to learn, classify and screen the association relationship between the entity words in the final entity library to obtain the classified data; wherein the association relationship includes one or more of a containment relationship, an instance relationship, an attribute relationship and no relationship; The classified data is normalized and outputted as normalized data for importing into a graph database.

2. The method for processing insulation status data of electric power equipment according to claim 1, characterized in that: The electric power equipment insulation status text data is unstructured text data; The step of identifying and extracting named entities from the insulation status text data of the power equipment to obtain entity vocabulary includes: Convert the insulation status text data of the power equipment into text form and perform script processing to obtain data to be processed; Based on the auxiliary dictionary, a preset statistical model is used to perform text segmentation and delete stop words on the data to be processed to obtain a segmentation result; The word segmentation results are screened to obtain the entity vocabulary.

3. The method for processing insulation status data of electric power equipment according to claim 2, characterized in that: The preset statistical model includes a hidden Markov model; the auxiliary dictionary includes any one or any combination of the following dictionaries: a power system professional dictionary, a stop word dictionary, and a common word dictionary.

4. The method for processing insulation status data of electric power equipment according to claim 1, characterized in that: The step of normalizing the classified data and outputting the normalized data for importing into a graph database comprises: The entity words, the associations between the entity words, and the attributes of the entity words in the classified data are written into corresponding format files according to the prescribed format of the corresponding graph database to complete the data normalization processing and output the normalized data for batch import into the corresponding graph database.

5. The method for processing insulation status data of electric power equipment according to claim 1, characterized in that: The graph database includes a Neo4j graph database.

6. A device for processing insulation status data of electric power equipment, characterized in that: The device comprises: The recognition module is used to obtain the text data of the insulation status of the power equipment, and to recognize and extract the named entities of the text data of the insulation status of the power equipment to obtain entity vocabulary; A coreference resolution module is used to perform coreference resolution processing on the entity vocabulary to determine the final entity library, including: confirming the word vector of each entity vocabulary according to a preset word vector dimension; obtaining each multiplication result of the entity vocabulary; the multiplication result is obtained by multiplying the word vector of the entity vocabulary with the word vector of another entity vocabulary after normalization; sorting each of the multiplication results, and determining the synonyms of the entity vocabulary according to the sorting results; screening and deleting synonyms in each of the synonyms, updating the word vector of each of the entity vocabulary, and obtaining the final entity library; A classification module, used for classifying the association relationship between entity words in the final entity library to obtain classified data, including: performing difference processing on the word vectors of any two entity words in the final entity library to obtain a difference processing result; based on the difference processing result, using a classifier to learn, classify and screen the association relationship between the entity words in the final entity library to obtain the classified data; wherein the association relationship includes one or more of a containment relationship, an instance relationship, an attribute relationship and no relationship; The normalization processing module is used to perform normalization processing on the classified data and output the normalized data for importing into the graph database.

7. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 5 are implemented.

Citation Information

Cited By

  • Insulation deterioration diagnosis method and system oriented to impedance spectrum data mining

    CN122796429A