A power distribution network technical standard knowledge graph construction method and system

Abstract

A power distribution network technical standard knowledge graph construction method and system, comprising the following steps: obtaining power distribution network knowledge text, obtaining a data set after preprocessing the power distribution network knowledge text, and training a power distribution network standard text classification model based on the data set; obtaining text information related to power distribution network standard vocabulary, and obtaining a power distribution network professional field vocabulary set in combination with the trained power distribution network standard text classification model; obtaining related text data of each vocabulary based on the power distribution network professional field vocabulary set, and jointly forming a power distribution network standard reference database in combination with the obtained power distribution network knowledge text; performing syntax analysis on the data in the power distribution network standard reference database to obtain power distribution network standard segmentation entity relationships; and constructing a power distribution network standard knowledge graph based on the vocabulary in the power distribution network professional field vocabulary set and the power distribution network standard segmentation entity relationships. The present application can efficiently and accurately obtain the required vocabulary sentences for knowledge graph construction.

Description

Technical Field

[0001] This invention relates to the technical field of power distribution technology standards, and more specifically, to a method and system for constructing a knowledge graph of power distribution network technology standards. Background Technology

[0002] With the rapid advancement of my country's new power system construction, the functional positioning of the distribution network, as the end point of the power system, is also changing. The large-scale integration of distributed photovoltaic, transformer substation energy storage, electric vehicle charging piles, and other devices has transformed the distribution network into a new type of distribution system with diversified source-grid-load-storage interaction.

[0003] In existing methods for constructing knowledge graphs in the power sector, it is common to build them by acquiring large amounts of data. However, this approach increases the complexity of data processing and is not conducive to obtaining keywords. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for constructing a knowledge graph of distribution network technology standards, which can solve the problems of being overly complex in the construction process of the knowledge graph of distribution network technology standards, incomplete vocabulary, and inaccurate vocabulary relationships in existing technologies.

[0005] The present invention adopts the following technical solution.

[0006] A method for constructing a knowledge graph of power distribution network technical standards includes the following steps:

[0007] The distribution network knowledge text is obtained, and after preprocessing, a training dataset is obtained. The standard class text classification model of the distribution network is then trained based on the training dataset.

[0008] Obtain textual information related to standard vocabulary of distribution network, and combine it with the trained standard text classification model of distribution network to obtain a vocabulary set of distribution network professional field;

[0009] Based on the vocabulary set of the power distribution network professional field, relevant text data of each term is obtained, and together with the power distribution network knowledge text obtained in step 1, a power distribution network standard reference database is constructed.

[0010] Syntactic analysis was performed on the data in the distribution network standard reference database to obtain the distribution network standard word segmentation entity relations;

[0011] A knowledge graph of distribution network standards is constructed based on the vocabulary in the distribution network professional domain vocabulary set and the entity relationships of distribution network standard word segmentation.

[0012] Preferably, the step of obtaining distribution network knowledge text, preprocessing the distribution network knowledge text to obtain a training dataset, and training a distribution network standard class text classification model based on the training dataset specifically includes the following steps:

[0013] Obtain standard knowledge texts about power distribution networks to construct an initial dataset;

[0014] The acquired distribution network standard knowledge text is preprocessed to obtain the training dataset required for training.

[0015] The standard text classification model for power distribution networks is trained using a CNN neural network and a training dataset to obtain the trained standard text classification model for power distribution networks.

[0016] Preferably, in the process of acquiring distribution network standard knowledge text to construct the initial dataset, the acquired distribution network knowledge text includes technical dictionaries, technical articles, and distribution network standard news texts;

[0017] The acquisition of news text related to power distribution network standards also includes: using search tools to search for news related to power distribution network standards, using Python programs to crawl target URLs to obtain webpage source code, and using BeautifulSoup parsing tools to parse the source code to obtain news text related to power distribution network standards.

[0018] Preferably, the preprocessing of the acquired distribution network standard knowledge text includes:

[0019] For the obtained technical articles and power distribution network standard news texts, remove all non-Chinese characters and increase the length of all texts to the maximum text length, and fill blank text with specified symbols;

[0020] The preprocessed technical articles and power distribution network standard news texts are split into data segments of equal length to obtain a set of data segments;

[0021] The similarity of each data segment is compared sequentially. If k consecutive Chinese characters are identical, the two are considered similar, where k is a pre-set integer.

[0022] Similar data segments are grouped into the same set to obtain m sets. Each set contains at least two data segments. If a data segment does not have a similar data segment, it is directly assigned to the training dataset required for training and labeled as the first type of dataset.

[0023] Sort the m sets according to the number of data segments in each set, and filter the data segments in each set to remove completely duplicate data segments to avoid the training dataset being too large. Label the filtered data segments as the second type dataset, the third type dataset, up to the Nth type dataset according to the number of data segments in the set to which they belong. The Nth type dataset has the most data segments in its set.

[0024] The filtered and labeled data segments are divided into the training dataset required for training, and each data segment is weighted, with the weights of the data segments in the first class dataset to the Nth class dataset increasing sequentially.

[0025] Preferably, the step of obtaining text information related to standard vocabulary of the distribution network and combining it with the trained standard text classification model of the distribution network to obtain a vocabulary set for the distribution network professional field specifically includes:

[0026] By searching for relevant terms related to power distribution networks on the internet and crawling text information, the obtained text information is input into a trained standard text classification model for power distribution networks to obtain a vocabulary set for the power distribution network professional field, as well as the importance of each term.

[0027] Preferably, the step of obtaining relevant text data for each term based on a vocabulary set in the field of power distribution networks, and combining it with the power distribution network knowledge text obtained in step 1 to form a power distribution network standard reference database, specifically includes:

[0028] By using the vocabulary from the obtained distribution network professional field vocabulary set, relevant term text data is crawled from the Internet. The term text data and distribution network knowledge text together form a distribution network standard reference database. Among them, the importance of each term is taken into account, and the more important the term, the more term text sentences are required.

[0029] Preferably, the step of performing syntactic analysis on the data in the power distribution network standard reference database to obtain the power distribution network standard word segmentation entity relations specifically includes the following steps:

[0030] For the power distribution network standard reference database, sentence components are obtained through syntactic analysis, forming text triples containing phrase 1, relation, and phrase 2;

[0031] By using dependency parsing, the neutral words in a sentence are identified, resulting in a set of neutral words.

[0032] By determining the dependency distance between the central word and other words, the entity relationships between words in the standard word segmentation of the power distribution network can be identified.

[0033] This invention also proposes a knowledge graph construction system for distribution network technology standards using the aforementioned knowledge graph construction method, comprising:

[0034] The data acquisition module is used to collect relevant data of the power distribution network, including power distribution network knowledge texts and power distribution network related terms;

[0035] The preprocessing module is used to preprocess the collected power distribution network knowledge text to obtain the training dataset;

[0036] The training module can train a standard text classification model for power distribution networks based on a training dataset.

[0037] The database construction module can obtain a vocabulary set for the power distribution network professional field based on the trained power distribution network standard text classification model, and obtain relevant text data of each word based on the vocabulary set for the power distribution network professional field, and construct a power distribution network standard reference database by combining power distribution network knowledge text.

[0038] The syntactic analysis module is used to perform syntactic analysis on the data in the power distribution network standard reference database to obtain the power distribution network standard word segmentation entity relations;

[0039] The knowledge graph construction module can construct a standard knowledge graph of the power distribution network based on the vocabulary in the power distribution network professional domain vocabulary set and the entity relationships of the power distribution network standard word segmentation.

[0040] The present invention also proposes a terminal, including a processor and a storage medium;

[0041] The storage medium is used to store instructions;

[0042] The processor is configured to operate according to the instructions to execute the steps of the knowledge graph construction method.

[0043] The present invention also proposes a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the knowledge graph construction method.

[0044] The beneficial effects of this invention are as follows: Compared with the prior art, the method for constructing a knowledge graph of distribution network technical standards provided by this invention, through crawling massive amounts of distribution network knowledge text from the Internet, training a text classification model for distribution network standards, extracting word segmentation entity relationships of distribution network standards, and constructing a knowledge graph of distribution network standards, provides learners and users with search and query functions for related word segmentation and relationships of distribution network standards, facilitating comprehensive and effective acquisition of relevant technical standards and clarifying the interrelationships between words in the standards. The knowledge graph construction method proposed in this invention can ensure the efficient implementation of distribution network technical standards, facilitate learning, application, and evaluation by grassroots staff, and assist searchers in comprehensively and effectively acquiring relevant technical standards and clarifying the interrelationships between words in the standards. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the method for constructing a knowledge graph of power distribution network technical standards in this invention;

[0046] Figure 2 This is a schematic diagram of a knowledge graph constructed based on the method proposed in this invention;

[0047] Figure 3 This is a schematic diagram of the knowledge graph construction system for power distribution network technical standards in this invention. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this invention. The embodiments described in this application are merely some embodiments of this invention, and not all embodiments. Based on the spirit of this invention, other embodiments obtained by those skilled in the art without creative effort are all within the protection scope of this invention.

[0049] like Figure 1 As shown, this invention proposes a method for constructing a knowledge graph of power distribution network technical standards. This method specifically includes the following steps:

[0050] The distribution network knowledge text is obtained, and after preprocessing, a dataset is obtained. Based on the dataset, a standard distribution network text classification model is trained.

[0051] Specifically, the process of acquiring distribution network knowledge text, preprocessing the distribution network knowledge text to obtain a dataset, and training a standard distribution network text classification model based on the dataset also includes the following steps:

[0052] An initial dataset was constructed by acquiring standard knowledge texts of the power distribution network.

[0053] Since the terminology of distribution network technical standards is a glossary document approved by a recognized body, this application selects technical dictionaries, technical articles, and news texts on distribution network standards published by recognized bodies when obtaining distribution network knowledge texts.

[0054] The acquisition of news text related to power distribution network standards also includes: using search tools to search for news related to "power distribution network standards", crawling target URLs using Python programs to obtain the webpage source code, and using BeautifulSoup parsing tools to parse the source code to obtain news text related to power distribution network standards.

[0055] The acquired distribution network standard knowledge text is preprocessed to obtain the training dataset required for training.

[0056] Specifically, preprocessing of distribution network standard knowledge text includes:

[0057] For the obtained technical articles and power distribution network standard news texts, all non-Chinese characters, including punctuation marks, English characters, and numbers, are removed. Each Chinese character is then separated. To speed up the network training process, the length of all texts is increased to the maximum text length, and blank text is filled with specified symbols.

[0058] Furthermore, considering the numerous citations in technical articles and news texts related to distribution network standards, the acquired technical articles and news texts related to distribution network standards need further screening after preprocessing, as follows:

[0059] The technical articles and power distribution network standard news texts that have undergone the above preprocessing are split into data segments of the same length to obtain a set of data segments;

[0060] The similarity of each data segment is compared sequentially. If k consecutive Chinese characters are identical, the two are considered similar, where k is an integer preset by the technicians.

[0061] Similar data segments are grouped into the same set to obtain m sets. Each set contains at least two data segments. If a data segment does not have a similar data segment, it is directly assigned to the training dataset required for training and labeled as the first type of dataset.

[0062] Sort the m sets according to the number of data segments in each set, and filter the data segments in each set to remove completely duplicate data segments to avoid the training dataset being too large. Label the filtered data segments as the second type dataset, the third type dataset, ... the Nth type dataset according to the number of data segments in the set to which they belong. The Nth type dataset has the most data segments in its set.

[0063] The filtered and labeled data segments are divided into the training dataset required for training, and weights are assigned to each data segment, with the weights of the data segments in the first to Nth datasets increasing sequentially.

[0064] A standard text classification model for power distribution networks was trained using a CNN neural network and a training dataset.

[0065] The training dataset is divided into training, validation, and test sets in a 7:1:2 ratio. The CNN network parameters are configured according to actual needs, including word vector dimension, sequence length, number of convolutional kernels, number of neurons in fully connected layers, and number of iterations.

[0066] The data segments and their corresponding weights from the training set, validation set, and test set are used as inputs for model training. The text classification model network is trained by calling cnn_model.py under the Python-tensorflow framework.

[0067] Preferably, multiple k values ​​can be preset, including k1, k2, ..., k m And k1 to k m Increment by increment, and then check if there are consecutive k values ​​in the data segment. pAll Chinese characters are identical, p∈[1,m]. The data segment is assigned a second weight based on the value of k, and this weight is used as a reference parameter for model training.

[0068] Obtain textual information related to standard vocabulary of distribution network, and combine it with the trained standard text classification model of distribution network to obtain a vocabulary set of distribution network professional field;

[0069] The specific steps for constructing a vocabulary set for the power distribution network professional field include: searching for power distribution networks on the internet, crawling text information from the searched terms using a Python program, inputting the text information into the power distribution network standard text classification model in step 1, and obtaining a vocabulary set for the power distribution network professional field, as well as the importance of each term.

[0070] Based on the vocabulary set of the power distribution network professional field, relevant text data of each term is obtained, and together with the power distribution network knowledge text obtained in step 1, a power distribution network standard reference database is constructed.

[0071] The construction of the distribution network standard reference database specifically includes: using the obtained distribution network professional field terminology to connect the word "standard" to crawl the term text data in the search engine, and the term text data and the above-mentioned distribution network knowledge text together form the distribution network standard reference database.

[0072] Preferably, those skilled in the art pre-construct standard system table text data for the distribution network, obtain relevant internal documents of the enterprise, and incorporate the standard system table text data and the data from the relevant internal documents of the enterprise into the distribution network standard reference database.

[0073] Further preferred, when acquiring term text data, the text data is constructed based on the obtained word importance, with words with higher importance corresponding to more text sentences.

[0074] Syntactic analysis was performed on the data in the distribution network standard reference database to obtain the distribution network standard word segmentation entity relations;

[0075] Specifically, performing syntactic analysis on the data in the power distribution network standard reference database to obtain the power distribution network standard word segmentation entity relations also includes the following steps:

[0076] Based on the obtained power distribution network standard reference database, sentence components are obtained through syntactic analysis, forming text triples containing phrase 1, relation, and phrase 2.

[0077] By using dependency parsing, the neutral words in a sentence are identified, resulting in a set of neutral words.

[0078] By determining the dependency distance between the central word and other words, the entity relationships between words in the standard word segmentation of the power distribution network can be identified.

[0079] A knowledge graph of standard distribution networks is constructed based on the vocabulary set of the distribution network professional field and the entity relationship of standard distribution network words.

[0080] Specifically, by storing the entities and word segmentation relationships of the distribution network standard reference database into the graph database neo4j, a knowledge graph of distribution network standards can be constructed. The knowledge graph supports search and query functions for word segmentation and relationships related to distribution network standards.

[0081] like Figure 2 As shown, Figure 2 Based on the method proposed in this invention to construct a knowledge graph of standard categories for power distribution networks, inputting "power distribution automation" yields related knowledge graphs such as... Figure 2 As shown in Table 1 below:

[0082] Table 1

[0083]

[0084]

[0085] like Figure 3 As shown, this invention also proposes a knowledge graph construction system for power distribution network technology standards. The above method can be implemented based on this system, which includes a data acquisition module, a preprocessing module, a training module, a database construction module, a syntactic analysis module, and a knowledge graph construction module.

[0086] The data acquisition module is used to collect relevant data of the power distribution network, including power distribution network knowledge texts and power distribution network related terms.

[0087] The preprocessing module is used to preprocess the collected power distribution network knowledge text to obtain the training dataset;

[0088] The training module trains a standard text classification model for power distribution networks based on the training dataset.

[0089] The database construction module obtains a vocabulary set for the power distribution network professional field based on the trained power distribution network standard text classification model, and obtains relevant text data for each term based on the vocabulary set. It then combines the power distribution network knowledge text to construct a power distribution network standard reference database.

[0090] The syntactic analysis module is used to perform syntactic analysis on the data in the power distribution network standard reference database to obtain the power distribution network standard word segmentation entity relations;

[0091] The knowledge graph construction module constructs a standard knowledge graph of the power distribution network based on the vocabulary in the power distribution network professional domain vocabulary set and the entity relationships of the power distribution network standard word segmentation.

[0092] The beneficial effects of this invention are that, compared with the prior art, the knowledge graph construction method proposed in this invention can ensure the efficient implementation of power distribution network technical standards, facilitate the learning, application and evaluation of grassroots staff, assist searchers in comprehensively and effectively obtaining the corresponding technical standards, and clarify the interrelationships between words in the standards.

[0093] This disclosure can be a system, method, and / or computer program product. A computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of this disclosure.

[0094] Computer-readable storage media can be tangible devices capable of holding and storing instructions for use by an instruction execution device. Computer-readable storage media can be, for example—but not limited to—electrical storage devices, magnetic storage devices, optical storage devices, electromagnetic storage devices, semiconductor storage devices, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital multifunction disc (DVD), memory sticks, floppy disks, mechanical encoding devices, such as punch cards or recessed protrusions storing instructions thereon, and any suitable combination of the foregoing. The computer-readable storage media used herein are not to be construed as transient signals themselves, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., light pulses through fiber optic cables), or electrical signals transmitted through wires.

[0095] The computer-readable program instructions described herein can be downloaded from computer-readable storage media to various computing / processing devices, or downloaded via a network, such as the Internet, local area network, wide area network, and / or wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards them to the computer-readable storage media in the respective computing / processing device.

[0096] Computer program instructions used to perform the operations of this disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, status setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The computer-readable program instructions may execute entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing the status information of the computer-readable program instructions to implement various aspects of this disclosure.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A method for constructing a knowledge graph of power distribution network technical standards, characterized in that, Includes the following steps: The distribution network knowledge text is obtained, and after preprocessing, a training dataset is obtained. The standard class text classification model of the distribution network is then trained based on the training dataset. Preprocessing of the acquired distribution network knowledge text includes: For the obtained technical articles and power distribution network standard news texts, remove all non-Chinese characters and increase the length of all texts to the maximum text length, and fill blank text with specified symbols; The preprocessed technical articles and power distribution network standard news texts are split into data segments of equal length to obtain a set of data segments; The similarity of each data segment is compared sequentially. If k consecutive Chinese characters are identical, the two are considered similar, where k is a pre-set integer. Similar data segments are grouped into the same set to obtain m sets. Each set contains at least two data segments. If a data segment does not have a similar data segment, it is directly assigned to the training dataset required for training and labeled as the first type of dataset. Sort the m sets according to the number of data segments in each set, and filter the data segments in each set to remove completely duplicate data segments to avoid the training dataset being too large. Label the filtered data segments according to the number of data segments in their respective sets as the second type dataset, the third type dataset, up to the Nth type dataset, with the Nth type dataset having the most data segments in its respective set. The filtered and labeled data segments are divided into the training dataset required for training, and each data segment is assigned a weight. The weights of the data segments in the first class dataset to the Nth class dataset increase sequentially. Obtain textual information related to standard vocabulary of distribution network, and combine it with the trained standard text classification model of distribution network to obtain a vocabulary set of distribution network professional field; Based on the vocabulary set of the power distribution network professional field, relevant text data of each term is obtained, and together with the obtained power distribution network knowledge text, a power distribution network standard reference database is constructed. Syntactic analysis was performed on the data in the distribution network standard reference database to obtain the distribution network standard word segmentation entity relations; A knowledge graph of distribution network standards is constructed based on the vocabulary in the distribution network professional domain vocabulary set and the entity relationships of distribution network standard word segmentation.

2. The method for constructing a knowledge graph of power distribution network technical standards as described in claim 1, characterized in that, The process of obtaining distribution network knowledge text, preprocessing the distribution network knowledge text to obtain a training dataset, and training a standard distribution network text classification model based on the training dataset specifically includes the following steps: Obtain knowledge text about power distribution networks to construct an initial dataset; The acquired power distribution network knowledge text is preprocessed to obtain the training dataset required for training. The standard text classification model for power distribution networks is trained using a CNN neural network and a training dataset to obtain the trained standard text classification model for power distribution networks.

3. The method for constructing a knowledge graph of power distribution network technical standards as described in claim 2, characterized in that, The initial dataset constructed by acquiring distribution network knowledge text includes technical dictionaries, technical articles, and news texts related to distribution network standards. The acquisition of news text related to power distribution network standards also includes: using search tools to search for news related to power distribution network standards, using Python programs to crawl target URLs to obtain webpage source code, and using BeautifulSoup parsing tools to parse the source code to obtain news text related to power distribution network standards.

4. The method for constructing a knowledge graph of power distribution network technical standards as described in claim 1, characterized in that, The process of acquiring text information related to standard vocabulary in the power distribution network and combining it with a trained standard text classification model for the power distribution network to obtain a vocabulary set for the power distribution network professional field specifically includes: By searching for relevant terms related to power distribution networks on the internet and crawling text information, the obtained text information is input into a trained standard text classification model for power distribution networks to obtain a vocabulary set for the power distribution network professional field, as well as the importance of each term.

5. The method for constructing a knowledge graph of power distribution network technical standards as described in claim 4, characterized in that, The process involves acquiring relevant text data for each term based on a vocabulary set specific to the power distribution network field, and combining this data with the acquired power distribution network knowledge text to form a power distribution network standard reference database. Specifically, this includes: By using the vocabulary from the obtained distribution network professional field vocabulary set, relevant term text data is crawled from the Internet. The term text data and distribution network knowledge text together form a distribution network standard reference database. Among them, the importance of each term is taken into account, and the more important the term, the more term text sentences are required.

6. The method for constructing a knowledge graph of power distribution network technical standards as described in claim 1, characterized in that, The process of performing syntactic analysis on the data in the power distribution network standard reference database to obtain the power distribution network standard word segmentation entity relations specifically includes the following steps: For the power distribution network standard reference database, sentence components are obtained through syntactic analysis, forming text triples containing phrase 1, relation, and phrase 2; Dependency parsing is used to determine the head words of a sentence, thus obtaining a set of head words; By determining the dependency distance between the central word and other words, the entity relationships between words in the standard word segmentation of the power distribution network can be identified.

7. A knowledge graph construction system for power distribution network technical standards using the knowledge graph construction method according to any one of claims 1-6, characterized in that, include: The data acquisition module is used to collect relevant data of the power distribution network, including power distribution network knowledge texts and power distribution network related terms; The preprocessing module is used to preprocess the collected power distribution network knowledge text to obtain the training dataset; Preprocessing of the acquired distribution network knowledge text includes: For the obtained technical articles and power distribution network standard news texts, remove all non-Chinese characters and increase the length of all texts to the maximum text length, and fill blank text with specified symbols; The preprocessed technical articles and power distribution network standard news texts are split into data segments of equal length to obtain a set of data segments; The similarity of each data segment is compared sequentially. If k consecutive Chinese characters are identical, the two are considered similar, where k is a pre-set integer. Similar data segments are grouped into the same set to obtain m sets. Each set contains at least two data segments. If a data segment does not have a similar data segment, it is directly assigned to the training dataset required for training and labeled as the first type of dataset. Sort the m sets according to the number of data segments in each set, and filter the data segments in each set to remove completely duplicate data segments to avoid the training dataset being too large. Label the filtered data segments according to the number of data segments in their respective sets as the second type dataset, the third type dataset, up to the Nth type dataset, with the Nth type dataset having the most data segments in its respective set. The filtered and labeled data segments are divided into the training dataset required for training, and each data segment is assigned a weight. The weights of the data segments in the first class dataset to the Nth class dataset increase sequentially. The training module can train a standard text classification model for power distribution networks based on a training dataset. The database construction module can obtain a vocabulary set for the power distribution network professional field based on the trained power distribution network standard text classification model, and obtain relevant text data of each word based on the vocabulary set for the power distribution network professional field, and construct a power distribution network standard reference database by combining power distribution network knowledge text. The syntactic analysis module is used to perform syntactic analysis on the data in the power distribution network standard reference database to obtain the power distribution network standard word segmentation entity relations; The knowledge graph construction module can construct a standard knowledge graph of the power distribution network based on the vocabulary in the power distribution network professional domain vocabulary set and the entity relationships of the power distribution network standard word segmentation.

8. A terminal, comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the steps of the method according to any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the program implements the steps of the method according to any one of claims 1-6.

Images