Hierarchical identification method, device and equipment for nuclear power plant maintenance and medium

Abstract

The application relates to the field of nuclear power equipment maintenance management, and in particular to a hierarchical identification method and device for nuclear power equipment maintenance, equipment and a medium. The method obtains a real-time maintenance hierarchical table and a latest version of a maintenance manual, carries out word segmentation on each cell text of the maintenance hierarchical table, calculates the occurrence frequency and the inverse document frequency of the word segmentation in the maintenance manual, filters out the word segmentation with the inverse document frequency of zero and the occurrence frequency higher than a value, splits the filtered text to form a word group, calculates a word group feature vector expression, generates a maintenance hierarchical vector table, converts a user input maintenance scheme description into a target vector, queries a matching result in the maintenance hierarchical vector table according to the vector, and determines a corresponding maintenance hierarchy. Therefore, the maintenance manual and the maintenance description of the user are automatically matched, so that accurate hierarchical identification of nuclear power equipment maintenance is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of nuclear power plant maintenance management, and in particular to a hierarchical identification method and device for nuclear power plant maintenance, equipment and medium. BACKGROUND

[0002] The safe and stable operation of nuclear power equipment is the key to ensuring the normal power generation of nuclear power plants and the safety of the surrounding environment and personnel. By classifying maintenance work, appropriate resource allocation, safety measures and management processes can be taken according to different maintenance levels, improving maintenance efficiency and reducing maintenance costs, while minimizing the impact of maintenance work on the normal operation of nuclear power plants. However, most of the current nuclear power equipment maintenance classification methods still have many defects.

[0003] Most of the current nuclear power equipment maintenance classification methods mainly rely on the manual experience of maintenance personnel. Maintenance experts rely on their years of work experience to classify maintenance work based on factors such as equipment failure phenomena, historical maintenance records, etc. However, different experts' judgment standards may differ, resulting in a lack of consistency and accuracy in the classification results. Moreover, the accumulation of manual experience takes a long time, and it is difficult for new maintenance personnel to quickly and accurately classify maintenance. And with the continuous development of nuclear power technology and the upgrading of equipment, new types of failures and maintenance needs are constantly emerging, and manual experience may not be able to cover all situations in a timely manner, thereby affecting the effectiveness of maintenance classification.

[0004] Therefore, how to automatically match the maintenance manual with the user's maintenance description to achieve accurate classification and identification of nuclear power equipment maintenance has become a problem to be solved. SUMMARY

[0005] Therefore, the embodiments of the present application provide a hierarchical identification method, device, equipment and medium for nuclear power equipment maintenance to solve the problem of how to automatically match the maintenance manual with the user's maintenance description to achieve accurate classification and identification of nuclear power equipment maintenance.

[0006] In a first aspect, the embodiments of the present application provide a hierarchical identification method for nuclear power equipment maintenance, comprising:

[0007] Obtain a real-time maintenance classification table and at least one latest version of a maintenance manual, perform word segmentation on the cell text in each cell in the maintenance classification table to obtain N word segmentation corresponding to the cell, N is an integer greater than zero;

[0008] For any cell, calculate the first occurrence frequency and the first inverse document frequency of each word segmentation in the cell in all maintenance manuals;

[0009] From the cell text, filter out the word segments whose first inverse document frequency is zero and whose first occurrence frequency is greater than the first frequency threshold to obtain the filtered text corresponding to the cell;

[0010] The word segments in the filtered text are split to form at least one first word group consisting of a first word segment and consecutive words, wherein the first word segment is any word segment in the filtered text, and the consecutive words segment is the first word segment combined with another word segment in the filtered text that is consecutive to the first word segment.

[0011] Calculate the second occurrence frequency and the second inverse document frequency of the continuous word segment in all maintenance manuals. Based on the first occurrence frequency and the second inverse document frequency of the first word segment, and the second word segment and the second occurrence frequency of the continuous word segment, obtain the feature vector expression of the corresponding first word group.

[0012] Traverse all cells of the maintenance classification table and the first phrase of each cell to obtain the feature vector expression of all cells. Write the feature vector expression of each cell into the corresponding cell to obtain the maintenance classification vector table.

[0013] Obtain the repair solution description input by the user, perform vector transformation on the repair solution description according to all repair manuals to obtain the target vector expression, query the repair classification vector table according to the target vector expression to obtain the matching result, and determine the repair classification corresponding to the matching result.

[0014] Secondly, according to an embodiment of this application, a classification identification device for nuclear power equipment maintenance includes:

[0015] The word segmentation module is used to obtain a real-time maintenance classification table and at least one latest version of the maintenance manual, and to segment the cell text in each cell of the maintenance classification table to obtain N words for the corresponding cell, where N is an integer greater than zero.

[0016] The frequency calculation module is used to calculate the first occurrence frequency and the first inverse document frequency of each word in the cell in all maintenance manuals for any given cell.

[0017] The filtering module is used to filter the word segments from the cell text that have a first inverse document frequency of zero and a first occurrence frequency greater than a first frequency threshold, so as to obtain the filtered text corresponding to the cell.

[0018] The word segmentation module is used to segment the words in the filtered text to form at least one first word group consisting of a first word segment and consecutive words, wherein the first word segment is any word in the filtered text, and the consecutive words segment is the first word segment combined with another word in the filtered text that is consecutive to the first word segment.

[0019] The vector representation module is used to calculate the second occurrence frequency and the second inverse document frequency of the continuous word segment in all maintenance manuals, and to obtain the feature vector representation of the corresponding first word group based on the first occurrence frequency and the second inverse document frequency of the first word segment, as well as the second word segment and the second occurrence frequency of the continuous word segment.

[0020] The vector table acquisition module is used to traverse all cells of the maintenance classification table and the first phrase of each cell to obtain the feature vector expression of all cells. The feature vector expression of each cell is written into the corresponding cell to obtain the maintenance classification vector table.

[0021] The matching module is used to obtain the repair solution description input by the user, perform vector transformation on the repair solution description according to all repair manuals to obtain the target vector expression, query the repair classification vector table according to the target vector expression, obtain the matching result, and determine the repair classification corresponding to the matching result.

[0022] Thirdly, embodiments of this application provide a computer device, the computer device including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the graded identification method for nuclear power equipment maintenance as described in the first aspect.

[0023] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the graded identification method for nuclear power equipment maintenance as described in the first aspect.

[0024] The beneficial effects of the embodiments in this application compared with the prior art are:

[0025] This application obtains a real-time maintenance grading table and at least one latest version of a maintenance manual. The text in each cell of the maintenance grading table is segmented into words, resulting in N words for each cell, where N is a positive integer. For any cell, the first occurrence frequency and the first inverse document frequency of each word in all maintenance manuals are calculated. From the cell text, words with a first inverse document frequency of zero and a first occurrence frequency greater than a first frequency threshold are filtered out to obtain the filtered text for the corresponding cell. The words in the filtered text are then split to form at least one first word group consisting of a first word and consecutive words. The first word is any word in the filtered text, and the consecutive words are the first word and another word consecutive to the first word in the filtered text. The process involves word segmentation and merging, calculating the second occurrence frequency and second inverse document frequency of consecutive word segments in all repair manuals, and obtaining the feature vector expression of the corresponding first word group based on the first occurrence frequency and second inverse document frequency of the first word segment, as well as the second occurrence frequency of consecutive word segments. This process iterates through all cells of the repair grading table and the first word group of each cell, obtaining the feature vector expression of all cells. The feature vector expression of each cell is written into the corresponding cell to obtain the repair grading vector table. The user-input repair solution description is then obtained. Based on all repair manuals, the repair solution description is vectorized to obtain the target vector expression. The repair grading vector table is then queried based on the target vector expression to obtain the matching result, and the repair grading corresponding to the matching result is determined. Alternatively, by obtaining the real-time repair grading table and the latest version of the repair manual, the text in each cell of the repair grading table is segmented, and the occurrence frequency and inverse document frequency of each word segment in the repair manual are calculated. Word segments with an inverse document frequency of zero and an occurrence frequency higher than a certain value are filtered out. The filtered text is split into word groups, and the word group feature vector expression is calculated to generate the repair grading vector table. The user-input repair solution description is converted into a target vector, and the matching result is queried in the repair grading vector table based on this vector to determine the corresponding repair grading. This automatically matches maintenance manuals with user maintenance descriptions to achieve accurate classification and identification of nuclear power equipment maintenance. Attached Figure Description

[0026] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0027] Figure 1 This is a schematic diagram of an application environment for a graded identification method for nuclear power equipment maintenance provided in Embodiment 1 of this application;

[0028] Figure 2 This is a flowchart illustrating a graded identification method for nuclear power equipment maintenance provided in Embodiment 2 of this application;

[0029] Figure 3 This is a flowchart illustrating a graded identification method for nuclear power equipment maintenance provided in Embodiment 3 of this application;

[0030] Figure 4 This is a flowchart illustrating a graded identification method for nuclear power equipment maintenance provided in Embodiment 4 of this application;

[0031] Figure 5 This is a flowchart illustrating a graded identification method for nuclear power equipment maintenance provided in Embodiment 5 of this application;

[0032] Figure 6 This is a schematic diagram of the structure of a graded identification device for nuclear power equipment maintenance provided in Embodiment Six of this application;

[0033] Figure 7 This is a schematic diagram of the structure of a computer device provided in Embodiment 7 of this application. Detailed Implementation

[0034] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0035] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0036] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0037] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0038] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0039] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0040] It should be understood that the sequence number of each step in the following embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0041] To illustrate the technical solution of this application, specific embodiments are described below.

[0042] The classification and identification method for nuclear power equipment maintenance provided in Embodiment 1 of this application can be applied to, for example... Figure 1 In this application environment, the client and server communicate with each other. Users can provide conditions, requirements, and operation instructions for the graded identification of nuclear power equipment maintenance through the client. The server generates control instructions for the graded identification method of nuclear power equipment maintenance based on the relevant content sent by the client.

[0043] The client side includes, but is not limited to, PDAs, desktop computers, laptops, ultra-mobile personal computers (UMPCs), netbooks, cloud terminal devices, and personal digital assistants (PDAs). The server side can be a standalone server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0044] See Figure 2 This is a flowchart illustrating a classification and identification method for nuclear power equipment maintenance provided in Embodiment 2 of this application. The classification and identification method for nuclear power equipment maintenance may include the following steps:

[0045] Step S201: Obtain the real-time maintenance classification table and at least one latest version of the maintenance manual. Perform word segmentation on the cell text in each unit of the maintenance classification table to obtain N words for the corresponding cell, where N is an integer greater than zero.

[0046] The maintenance classification table serves as the basis for classifying different maintenance situations in nuclear power equipment maintenance. It can include various maintenance scenarios, maintenance items, and corresponding classification information. "Real-time" emphasizes that the obtained classification table is up-to-date and reflects the current situation, because the maintenance classification standards for nuclear power equipment may be continuously updated due to factors such as technological advancements and regulatory requirements.

[0047] Maintenance manuals are detailed guides to the maintenance of nuclear power equipment, containing a wealth of information such as equipment structure, maintenance procedures, and technical requirements. Obtaining the latest version of the maintenance manual ensures that the knowledge used for subsequent analysis is the most accurate and up-to-date. At least one maintenance manual is required; different manuals may cover different aspects of maintenance knowledge, and multiple manuals can provide more comprehensive information.

[0048] Maintenance classification tables are mostly presented in tabular form, with each cell containing specific text information such as the name and description of the maintenance item. Word segmentation is the process of dividing a continuous sequence of text into individual words according to certain rules. For example, the cell text "nuclear power equipment fault maintenance" can be segmented into the individual words "nuclear power equipment," "fault," and "maintenance." The purpose of word segmentation is to transform text information into basic units that computers can process, facilitating subsequent operations such as frequency calculation and semantic analysis.

[0049] Each cell, after word segmentation, will yield a certain number of words, represented by N, and N must be greater than zero. This is because even if the text in a cell is short, it will still contain at least one word. For example, if the cell text is "maintenance", it will yield one word after segmentation. If the text is "routine inspection of nuclear power equipment", it may yield three words: "nuclear power equipment", "routine", and "inspection".

[0050] Step S202: For any cell, calculate the first occurrence frequency and the first inverse document frequency of each word in the cell in all maintenance manuals.

[0051] The first occurrence frequency refers to how frequently a particular word appears across all maintenance manuals. Specifically, it's calculated by dividing the number of times the word appears in all maintenance manuals by the total number of words in all manuals. A high first occurrence frequency indicates that a word appears frequently in maintenance manuals, potentially representing a common concept or term. For example, in a nuclear power equipment maintenance manual, words like "nuclear power" and "equipment" might appear frequently because they are fundamental terms closely related to nuclear power equipment maintenance.

[0052] The first inverse document frequency (INV) reflects the rarity or uniqueness of a word segment. It measures how many maintenance manuals contain that word. The calculation method is to divide the total number of maintenance manuals by the number of manuals containing that word, and then take the logarithm. A high INV indicates that a word appears only in a few maintenance manuals, possessing strong uniqueness and potentially representing a specific maintenance scenario, technology, or problem. For example, proprietary terms for certain types of nuclear power equipment may only appear in maintenance manuals covering that type of equipment, resulting in a relatively high INV.

[0053] Step S203: From the cell text, filter out the word segments whose first inverse document frequency is zero and whose first occurrence frequency is greater than the first frequency threshold to obtain the filtered text corresponding to the cell.

[0054] The first inverse document frequency (INV) measures the rarity of a word within a document set. A INV of zero for a word means that it appears in all maintenance manuals. This lack of uniqueness may not be very useful for distinguishing different maintenance levels. For example, terms like "equipment maintenance" might appear frequently in all nuclear power equipment maintenance manuals, and their INV might be zero.

[0055] The first frequency of occurrence reflects how frequently a word appears in a document set. The first frequency threshold is a pre-set value used to determine whether a word's frequency of occurrence is too high. If a word's first frequency of occurrence is greater than this threshold, it means that it appears very frequently in the maintenance manual. Combined with the condition that the first inverse document frequency is zero, such a word not only appears in all manuals but also appears very frequently, and is likely some commonly used, undifferentiated words. Words that simultaneously satisfy the condition of zero first inverse document frequency and a first frequency of occurrence greater than the first frequency threshold are identified from the cell text and then removed from the cell text. After the filtering operation, the text composed of the remaining words is the filtered text for the corresponding cell. The words retained in this filtered text are more likely to be words that are important for maintenance classification and have distinctiveness, and can more accurately reflect the maintenance situation represented by the cell.

[0056] Step S204: The word segments in the filtered text are split to form at least one first word group consisting of a first word segment and consecutive words, wherein the first word segment is any word segment in the filtered text, and the consecutive words segment is the first word segment combined with another word segment in the filtered text that is consecutive to the first word segment.

[0057] Optionally, step S204, which involves splitting the word segments in the filtered text to form at least one first word group consisting of a first word segment and consecutive word segments, may include the following steps:

[0058] According to a preset word segmentation library, the filtered text is segmented in the word segmentation library to obtain the segmented text;

[0059] Based on semantic logic, the split text is combined to form at least one first word group consisting of a first word segment and consecutive word segments.

[0060] The first segment is any segment in the filtered text. Within the filtered text, every segment can potentially be used as the first segment.

[0061] Continuous word segmentation is formed by combining a first word segment with another word segmented consecutively with the first word segment in the filtered text. For example, if the filtered text is "nuclear power equipment failure", and "nuclear power" is the first word segment, and "equipment" is the word segmented consecutively with it, then "nuclear power equipment" is a continuous word segment, and "nuclear power" and "nuclear power equipment" together form a first word phrase.

[0062] The filtered text is segmented based on a pre-built vocabulary library, which contains common terms and technical jargon used in the field of nuclear power equipment maintenance. This pre-built vocabulary library is constructed based on domain knowledge and previous text data to assist in more accurate segmentation of the filtered text.

[0063] The filtered text is compared with a word segmentation database, and then segmented according to the vocabulary rules in the database. For example, if the filtered text is "nuclear power unit maintenance", and the word segmentation database contains the words "nuclear power unit" and "maintenance", then the filtered text will be split into two parts: "nuclear power unit" and "maintenance", resulting in the split text.

[0064] Semantic logic refers to the appropriate collocation and semantic association between words in the context of nuclear power equipment maintenance. For example, the combination of "nuclear power" and "equipment" into "nuclear power equipment" is semantically reasonable because it accurately describes the concept of equipment related to nuclear power. Based on semantic logic, the word segments in the split text are combined to form at least one first phrase. There can be various ways to combine them, taking into account the semantic relationships between words and the actual maintenance scenario. For example, splitting the text into "nuclear power equipment failure" can form first phrases such as "nuclear power equipment," "equipment failure," and "nuclear power equipment failure."

[0065] Step S205: Calculate the second occurrence frequency and second inverse document frequency of the continuous word segment in all maintenance manuals. Based on the first occurrence frequency and second inverse document frequency of the first word segment, and the second word segment and second occurrence frequency of the continuous word segment, obtain the feature vector expression of the corresponding first word group.

[0066] Optionally, step S205, which involves obtaining the feature vector representation of the corresponding first word group based on the first occurrence frequency and second inverse document frequency of the first word segmentation, and the second word segmentation and second occurrence frequency of the continuous word segmentation, may include the following steps:

[0067] The first occurrence frequency, the second inverse document frequency, and the second occurrence frequency are calculated to obtain the frequency calculation results;

[0068] Based on the frequency calculation results, determine the first weight of the first word group;

[0069] Based on the first weight, the feature vector representation of the corresponding first word group is obtained.

[0070] Similar to the first frequency of occurrence, the second frequency of occurrence refers to the frequency with which the consecutive word appears in all repair manuals. It is calculated by dividing the number of times the consecutive word appears in all repair manuals by the total number of words in all repair manuals.

[0071] Similar to the first inverse document frequency (IVF), the second inverse document frequency reflects the rarity or uniqueness of consecutive word segments. The calculation method is to divide the total number of repair manuals by the number of repair manuals containing the given consecutive word segment, then take the logarithm to obtain the feature vector representation of the corresponding first word group.

[0072] The first occurrence frequency, the second inverse document frequency, and the second occurrence frequency are calculated, for example, by weighted summation. The specific calculation method can be adjusted according to actual needs and experimental results. The first weight of the first term is determined based on the frequency calculation results. The frequency calculation results reflect the comprehensive importance and uniqueness of the first term in the maintenance manual. The first weight of the first term can be determined based on this result. Feature vector representation represents the first term in vector form for processing in machine learning or other algorithms. The first weight can be an important element in the feature vector. In more complex cases, the feature vector may contain multiple dimensions, including other information related to the term besides the first weight.

[0073] Step S206: Traverse all cells of the maintenance classification table and the first phrase of each cell to obtain the feature vector expression of all cells. Write the feature vector expression of each cell into the corresponding cell to obtain the maintenance classification vector table.

[0074] The maintenance classification table contains multiple cells, each potentially recording different information related to nuclear power equipment maintenance, such as fault description, maintenance procedures, and maintenance level. Each cell in the table needs to be accessed individually to ensure no information is missed.

[0075] The first phrase of each cell is traversed. For each cell, the filtered text has been split into at least one first phrase in step S204. Here, it is necessary to traverse all the first phrases in each cell because each first phrase has its corresponding feature vector representation, and these vectors will together constitute the feature vector representation of that cell.

[0076] Based on the method for calculating the feature vector representation of the first word group in step S205, the feature vectors of all first word groups in each cell are integrated. Common integration methods include concatenating or weighted summing of these feature vectors.

[0077] Write the calculated feature vector representation of each cell back to the corresponding cell in the maintenance grading table. Originally, the cells stored text information; now, they are replaced with feature vector representations, thus transforming the maintenance grading table from a text table into a table containing vector information.

[0078] After the above operations, the entire maintenance classification table becomes a maintenance classification vector table. Each cell in this vector table is represented by a feature vector, which more accurately reflects the semantics and features of the maintenance information contained in the cell. The maintenance classification vector table can be used as input for subsequent machine learning algorithms, data analysis, and other operations for tasks such as maintenance classification identification, classification, and clustering.

[0079] Step S207: Obtain the repair solution description input by the user; perform vector transformation on the repair solution description according to all repair manuals to obtain the target vector expression; query the repair classification vector table according to the target vector expression to obtain the matching result; and determine the repair classification corresponding to the matching result.

[0080] In actual nuclear power equipment maintenance scenarios, users will provide a maintenance plan description based on the equipment malfunction and their own maintenance experience. This description is usually presented in natural language and includes key information in the maintenance process, such as the malfunction symptoms, maintenance steps, and tools used.

[0081] Based on all repair manuals, the repair plan descriptions are vectorized. First, similar to the processing of text in the repair grading table cells, the repair plan descriptions are segmented into individual words. Referring to all repair manuals, the frequency of occurrence (FOC) and inverse document frequency (IVF) of these words are calculated. The FOC reflects how frequently the words appear in the repair plan descriptions, while the IVF reflects their rarity across all repair manuals. The words are then combined into phrases, and the feature vectors of these phrases are calculated. Finally, the feature vectors of these phrases are integrated to obtain the target vector representation of the repair plan description. This target vector representation transforms the semantic information in the repair plan description into a computer-processable numerical form.

[0082] Using a suitable similarity metric (such as cosine similarity or Euclidean distance), the target vector representation is compared with the feature vector representation of each cell in the maintenance classification vector table. The similarity calculation results reflect the degree of similarity between the maintenance plan description and the maintenance situation represented by each cell in the maintenance classification table. The cells in the maintenance classification vector table are then sorted according to their similarity scores, and the cells(s) with the highest similarity to the target vector representation are identified; these cells represent the matching results.

[0083] Each cell in the maintenance grading vector table corresponds to a specific maintenance grading. Once a match is found, the maintenance grading level is determined based on the matching cell's position in the maintenance grading table and pre-defined grading rules. For example, if the matching cell is located in the "Advanced Maintenance" area of ​​the maintenance grading table, then the maintenance plan can be determined to correspond to an advanced maintenance grading.

[0084] This application obtains a real-time maintenance grading table and at least one latest version of a maintenance manual. The text in each cell of the maintenance grading table is segmented into words, resulting in N words for each cell, where N is a positive integer. For any cell, the first occurrence frequency and the first inverse document frequency of each word in all maintenance manuals are calculated. From the cell text, words with a first inverse document frequency of zero and a first occurrence frequency greater than a first frequency threshold are filtered out to obtain the filtered text for the corresponding cell. The words in the filtered text are then split to form at least one first word group consisting of a first word and consecutive words. The first word is any word in the filtered text, and the consecutive words are the first word and another word consecutive to the first word in the filtered text. The process involves word segmentation and merging, calculating the second occurrence frequency and second inverse document frequency of consecutive word segments in all repair manuals, and obtaining the feature vector expression of the corresponding first word group based on the first occurrence frequency and second inverse document frequency of the first word segment, as well as the second occurrence frequency of consecutive word segments. This process iterates through all cells of the repair grading table and the first word group of each cell, obtaining the feature vector expression of all cells. The feature vector expression of each cell is written into the corresponding cell to obtain the repair grading vector table. The user-input repair solution description is then obtained. Based on all repair manuals, the repair solution description is vectorized to obtain the target vector expression. The repair grading vector table is then queried based on the target vector expression to obtain the matching result, and the repair grading corresponding to the matching result is determined. Alternatively, by obtaining the real-time repair grading table and the latest version of the repair manual, the text in each cell of the repair grading table is segmented, and the occurrence frequency and inverse document frequency of each word segment in the repair manual are calculated. Word segments with an inverse document frequency of zero and an occurrence frequency higher than a certain value are filtered out. The filtered text is split into word groups, and the word group feature vector expression is calculated to generate the repair grading vector table. The user-input repair solution description is converted into a target vector, and the matching result is queried in the repair grading vector table based on this vector to determine the corresponding repair grading. This automatically matches maintenance manuals with user maintenance descriptions to achieve accurate classification and identification of nuclear power equipment maintenance.

[0085] See Figure 3 This is a flowchart illustrating a graded identification method for nuclear power equipment maintenance provided in Embodiment 3 of this application. Figure 3As shown, the step S207 above, which involves vectorizing the repair plan description based on all repair manuals to obtain the target vector expression, may include the following steps:

[0086] Step S301: The words in the repair solution description are split to form at least one descriptive word group consisting of descriptive words and connecting words, wherein the descriptive words are any words in the repair solution description, and the connecting words are the descriptive words and another word in the repair solution description that is continuous with the descriptive words.

[0087] Step S302: Calculate the third occurrence frequency and third inverse document frequency of the connective segment in all maintenance manuals. Based on the fourth occurrence frequency and fourth inverse document frequency of the descriptive segment, and the third segment and third inverse document frequency of the connective segment, perform vector transformation on the maintenance solution description to obtain the target vector expression.

[0088] In this context, the descriptive segmentation refers to any single segmentation word within the repair plan description. After the repair plan description undergoes segmentation, each segmentation word can potentially serve as a descriptive segmentation word.

[0089] Connective word segmentation refers to the combination of a descriptive word and another word that is consecutive to the descriptive word in the repair solution description. For example, if the repair solution description is "replace the motor belt", and "motor" is the descriptive word and "belt" is the word that is consecutive to it, then "motor belt" is the connective word, and "motor" and "motor belt" constitute a descriptive word phrase.

[0090] By breaking down the repair plan description into descriptive phrases, the semantic information within the repair plan can be captured more comprehensively. A single word may not be able to fully express the meaning of the repair operation, while phrases can more accurately describe the specific objects and actions in the repair process.

[0091] The third frequency of occurrence refers to the frequency of the connecting word in all repair manuals. The calculation method is to divide the number of times the connecting word appears in all repair manuals by the total number of words (including consecutive words and single words) in all repair manuals.

[0092] The third inverse document frequency (3IVF) reflects the rarity or uniqueness of a connective term. It is calculated by dividing the total number of repair manuals by the number of manuals containing the connective term, and then taking the logarithm. For example, if there are 20 repair manuals, and 4 of them contain the connective term "motor belt," then the 3IVF of "motor belt" is...

[0093] This involves comprehensively considering the fourth occurrence frequency and fourth inverse document frequency of descriptive word segments, as well as the relevant frequency information of connecting word segments. A mathematical model or algorithm is typically used to integrate and calculate this frequency information. For example, a weighted summation of the frequency information of descriptive and connecting word segments can be performed.

[0094] These calculations transform the textual information in the maintenance plan description into numerical form, ultimately yielding a target vector representation. This target vector reflects the semantic features of the maintenance plan description, facilitating subsequent comparison and matching with vectors in the maintenance classification vector table.

[0095] In this embodiment, by comparing the target vector with the vectors in the maintenance classification vector table, the best-matching maintenance classification can be found quickly and accurately. This helps improve the accuracy and efficiency of maintenance classification, provides scientific guidance for maintenance work, rationally allocates maintenance resources, and ensures the safe and stable operation of nuclear power equipment.

[0096] See Figure 4 This is a flowchart illustrating a graded identification method for nuclear power equipment maintenance provided in Embodiment 4 of this application. Figure 4 As shown, step S207, which involves querying the maintenance classification vector table based on the target vector expression to obtain a matching result and determining the maintenance classification corresponding to the matching result, may include the following steps:

[0097] Step S401: Traverse the maintenance classification vector table, calculate the similarity between the target vector expression and all vectors in the maintenance classification vector table, and obtain the similarity result;

[0098] Step S402: Based on the similarity result, query the target vector in the maintenance classification vector table to obtain the matching result and determine the maintenance classification corresponding to the matching result.

[0099] Optionally, based on the similarity results, the matching results are sorted in descending order of similarity to obtain a vector sorting result;

[0100] Based on a preset candidate range, the vector sorting results are filtered to obtain the vector selection results;

[0101] Based on the vector selection results, the maintenance level corresponding to the matching result is determined.

[0102] The maintenance classification vector table contains feature vector representations for each cell, representing different maintenance conditions and classifications. Each vector in the table needs to be accessed to ensure no possible matches are missed.

[0103] Use an appropriate similarity metric to calculate the similarity between the target vector representation and all vectors in the maintenance classification vector table. Available similarity metrics include cosine similarity and Euclidean distance. Cosine similarity measures the similarity between two vectors by calculating the cosine of the angle between them. A cosine value closer to 1 indicates that the two vectors are more similar in direction, meaning higher similarity; a cosine value closer to -1 indicates that the two vectors are more opposite in direction, meaning lower similarity.

[0104] Euclidean distance refers to the straight-line distance between two vectors. The closer the distance, the more similar the two vectors are. After calculation, a set of similarity values ​​is obtained, each value corresponding to the similarity between the target vector and a vector in the maintenance grading vector table. Based on the calculated similarity results, the vector with the highest similarity to the target vector is found in the maintenance grading vector table. The cell contents corresponding to these vectors are the matching results. Each cell in the maintenance grading vector table is associated with a specific maintenance grading. Once a matching result is found, the maintenance grading corresponding to that matching result is determined based on the position of the matching cell in the table and pre-defined grading rules.

[0105] Arrange all similarity results in descending order to obtain vector sorting results. This makes it clear which vectors have a high similarity to the target vector and which have a low similarity.

[0106] The preset candidate range is a pre-defined standard, such as selecting only the top k vectors with the highest similarity. This filtering process reduces the workload of subsequent processing while focusing on the most likely matches, resulting in vector selection. From the filtered vector selection results, the corresponding matching results are determined, and the final maintenance level is determined based on these matching results.

[0107] In this embodiment of the application, by calculating similarity and filtering matching results, the most matching level with the user-input maintenance plan can be found quickly and accurately from a large amount of maintenance classification information, avoiding the tediousness and inaccuracy of manual search and judgment, and improving the efficiency and accuracy of maintenance classification.

[0108] See Figure 5 This is a flowchart illustrating a graded identification method for nuclear power equipment maintenance provided in Embodiment 5 of this application. Figure 5 As shown, after querying the maintenance classification vector table according to the target vector expression to obtain the matching result and determine the maintenance classification corresponding to the matching result in step S207, it may further include the following steps:

[0109] Step S501: Obtain a preset table template and extract the title field from the table template;

[0110] Step S502: Based on the title field, classify the maintenance level to obtain classification information;

[0111] Step S503: Fill the table template with the classification information to obtain the display table.

[0112] The preset table template is a pre-designed table format that specifies the structure and style for displaying maintenance classification information. This template can be designed according to the needs and standards of nuclear power equipment maintenance management, and may include different information areas such as "Maintenance Classification," "Maintenance Content," "Maintenance Personnel Requirements," and "Maintenance Tools."

[0113] The header fields in a table template are the names of the various information areas, defining the information categories represented by different columns or rows in the table. Extracting these header fields clarifies the basis for subsequent classification of maintenance grading information. For example, header fields such as "Maintenance Grading," "Fault Description," and "Maintenance Procedures" can be extracted from the table template.

[0114] Based on the previously determined maintenance classification results, the various information contained therein is sorted and analyzed. For example, maintenance classification information may include the specific maintenance operations, the equipment components involved, and the required maintenance skills. According to the extracted title fields, the maintenance classification information is assigned to the corresponding categories. For instance, if the title field includes "fault description," then the descriptions related to the fault are extracted from the maintenance classification information and categorized under the "fault description" category; if it includes "maintenance steps," then the steps related to the specific maintenance operations are extracted from the maintenance classification information and categorized under the "maintenance steps" category. This process yields categorized information, making the originally somewhat disorganized maintenance classification information more organized.

[0115] The categorized information is then filled into the corresponding positions in the table template according to the correspondence of the title fields. For example, information categorized as "Maintenance Classification" is filled into the cells under the "Maintenance Classification" title in the table template; information categorized as "Fault Description" is filled into the cells corresponding to the "Fault Description" title. After filling, the table template becomes a display table containing detailed maintenance classification information. This table presents the relevant content of maintenance classification in a structured and visual way, making it convenient for maintenance personnel and managers to quickly view and understand the specific details of maintenance classification, providing an intuitive basis for maintenance decisions and resource allocation.

[0116] In this embodiment of the application, complex maintenance classification information is presented in a clear and easy-to-understand form to facilitate communication and collaboration among different personnel.

[0117] Corresponding to the graded identification method for nuclear power equipment maintenance in the above embodiments,Figure 6 A structural block diagram of a classification identification device for nuclear power equipment maintenance provided in Embodiment Six of this application is shown. For ease of explanation, only the parts relevant to the embodiments of this application are shown.

[0118] See Figure 6 The classification and identification device for nuclear power equipment maintenance includes:

[0119] The word segmentation module 61 is used to obtain a real-time maintenance classification table and at least one latest version of the maintenance manual, and to segment the cell text in each cell of the maintenance classification table to obtain N words for the corresponding cell, where N is an integer greater than zero.

[0120] The frequency calculation module 62 is used to calculate, for any cell, the first occurrence frequency and the first inverse document frequency of each word in the cell in all maintenance manuals;

[0121] The filtering module 63 is used to filter the word segments from the cell text that have a first inverse document frequency of zero and a first occurrence frequency greater than a first frequency threshold, so as to obtain the filtered text corresponding to the cell.

[0122] The word segmentation module 64 is used to segment the words in the filtered text to form at least one first word group consisting of a first word segment and consecutive words, wherein the first word segment is any word in the filtered text, and the consecutive words segment is the first word segment and another word in the filtered text that is consecutive to the first word segment.

[0123] The vector representation module 65 is used to calculate the second occurrence frequency and the second inverse document frequency of the continuous word segment in all maintenance manuals, and to obtain the feature vector representation of the corresponding first word group based on the first occurrence frequency and the second inverse document frequency of the first word segment, as well as the second word segment and the second occurrence frequency of the continuous word segment.

[0124] The vector table is obtained by module 66, which is used to traverse all cells of the maintenance classification table and the first phrase of each cell to obtain the feature vector expression of all cells, and write the feature vector expression of each cell into the corresponding cell to obtain the maintenance classification vector table.

[0125] The matching module 67 is used to obtain the repair solution description input by the user, perform vector transformation on the repair solution description according to all repair manuals to obtain the target vector expression, query the repair classification vector table according to the target vector expression to obtain the matching result, and determine the repair classification corresponding to the matching result.

[0126] Optionally, the word segmentation module 64 includes:

[0127] The text segmentation unit is used to segment the filtered text according to a preset word segmentation library to obtain segmented text;

[0128] A text combination unit is used to combine the split text according to semantic logic to form at least one first word group consisting of a first word segment and consecutive word segments.

[0129] Optionally, the vector representation module 65 includes:

[0130] The frequency calculation unit is used to calculate the first occurrence frequency, the second inverse document frequency, and the second occurrence frequency to obtain the frequency calculation result;

[0131] The weight determination unit is used to determine the first weight of the first word group based on the frequency calculation result;

[0132] The expression generation unit is used to obtain the feature vector expression of the corresponding first word group according to the first weight.

[0133] Optionally, the matching module 67 includes:

[0134] The maintenance description splitting unit is used to split the words in the maintenance solution description to form at least one descriptive word group consisting of a descriptive word and a connecting word, wherein the descriptive word is any word in the maintenance solution description, and the connecting word is formed by merging the descriptive word with another word that is continuous with the descriptive word in the maintenance solution description;

[0135] The vector transformation unit is used to calculate the third occurrence frequency and the third inverse document frequency of the connecting word in all maintenance manuals, and to perform vector transformation on the maintenance solution description based on the fourth occurrence frequency and the fourth inverse document frequency of the description word, as well as the third word and the third inverse document frequency of the connecting word, to obtain the target vector expression.

[0136] Optionally, the matching module 67 includes:

[0137] The similarity calculation unit is used to traverse the maintenance classification vector table, calculate the similarity between the target vector expression and all vectors in the maintenance classification vector table, and obtain the similarity result.

[0138] The maintenance classification matching unit is used to query the target vector in the maintenance classification vector table according to the similarity result, obtain the matching result, and determine the maintenance classification corresponding to the matching result.

[0139] Optionally, the classification and identification device for nuclear power equipment maintenance further includes:

[0140] The vector sorting module is used to sort the matching results in descending order of similarity based on the similarity results, so as to obtain the vector sorting results;

[0141] The sorting and filtering module is used to filter the vector sorting results according to a preset candidate range to obtain vector selection results;

[0142] The grading determination module is used to determine the maintenance grading corresponding to the matching result based on the vector selection result.

[0143] Optionally, the classification and identification device for nuclear power equipment maintenance further includes:

[0144] The field extraction module is used to obtain a preset table template and extract the title field from the table template;

[0145] The information classification module is used to classify the maintenance level according to the title field to obtain classification information;

[0146] The table filling module is used to fill the table template with category information to obtain the display table.

[0147] It should be noted that the information interaction and execution process between the above modules, units, and sub-units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0148] Figure 7 This is a schematic diagram of the structure of a computer device provided in Embodiment Seven of this application. Figure 7 As shown, the computer device of this embodiment includes: at least one processor ( Figure 7 The diagram shows only one of the following: a memory and a computer program stored in the memory and executable on at least one processor. When the processor executes the computer program, it implements the steps of any of the above-described classification identification methods for nuclear power equipment maintenance or the embodiments of the classification identification methods for nuclear power equipment maintenance.

[0149] This computer device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that... Figure 7 The examples of computer devices are merely examples and do not constitute a limitation on computer devices. Computer devices may include more or fewer components than shown in the illustration, or combinations of certain components, or different components, such as network interfaces, displays, and input devices.

[0150] The processor referred to can be a CPU, but it can also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0151] Memory includes readable storage media, internal memory, etc., wherein internal memory can be the RAM of a computer device, providing an environment for the operation of the operating system and computer-readable instructions stored in the readable storage media. The readable storage media can be the hard drive of a computer device, or in other embodiments, it can be an external storage device of the computer device, such as a plug-in hard drive, Smart Media Card (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory can include both internal storage units and external storage devices of the computer device. Memory is used to store the operating system, applications, bootloader, data, and other programs, such as program code for computer programs. Memory can also be used to temporarily store data that has been output or will be output.

[0152] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above device can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here. If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. A computer-readable medium can include at least: any entity or device capable of carrying computer program code, a recording medium, a computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0153] The implementation of all or part of the processes in the methods of the above embodiments can also be accomplished by a computer program product. When the computer program product is run on a computer device, it enables the computer device to execute the steps in the above method embodiments.

[0154] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0155] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0156] In the embodiments provided in this application, it should be understood that the disclosed apparatus / computer devices and methods can be implemented in other ways. For example, the apparatus / computer device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0157] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0158] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A graded identification method for nuclear power equipment maintenance, characterized in that, include: Obtain a real-time maintenance grading table and at least one latest version of a maintenance manual. Segment the text in each cell of the maintenance grading table to obtain N segments for the corresponding cell, where N is a positive integer. For any given cell, calculate the first occurrence frequency and the first inverse document frequency of each word in the cell across all maintenance manuals; From the cell text, filter out the word segments whose first inverse document frequency is zero and whose first occurrence frequency is greater than the first frequency threshold to obtain the filtered text corresponding to the cell; The word segments in the filtered text are split to form at least one first word group consisting of a first word segment and consecutive words, wherein the first word segment is any word segment in the filtered text, and the consecutive words segment is the first word segment combined with another word segment in the filtered text that is consecutive to the first word segment. Calculate the second occurrence frequency and the second inverse document frequency of the continuous word segment in all maintenance manuals. Based on the first occurrence frequency and the second inverse document frequency of the first word segment, and the second word segment and the second occurrence frequency of the continuous word segment, obtain the feature vector expression of the corresponding first word group. Traverse all cells of the maintenance classification table and the first phrase of each cell to obtain the feature vector expression of all cells. Write the feature vector expression of each cell into the corresponding cell to obtain the maintenance classification vector table. Obtain the repair solution description input by the user, perform vector transformation on the repair solution description according to all repair manuals to obtain the target vector expression, query the repair classification vector table according to the target vector expression to obtain the matching result, and determine the repair classification corresponding to the matching result.

2. The hierarchical identification method for nuclear power equipment maintenance according to claim 1, characterized in that, The step of splitting the word segments in the filtered text to form at least one first word group consisting of a first word segment and consecutive word segments includes: According to a preset word segmentation library, the filtered text is segmented in the word segmentation library to obtain the segmented text; Based on semantic logic, the split text is combined to form at least one first word group consisting of a first word segment and consecutive word segments.

3. The hierarchical identification method for nuclear power equipment maintenance according to claim 1, characterized in that, The step of obtaining the feature vector representation of the corresponding first word group based on the first occurrence frequency and second inverse document frequency of the first word segmentation, and the second word segmentation and second occurrence frequency of the continuous word segmentation, includes: The first occurrence frequency, the second inverse document frequency, and the second occurrence frequency are calculated to obtain the frequency calculation results; Based on the frequency calculation results, determine the first weight of the first word group; Based on the first weight, the feature vector representation of the corresponding first word group is obtained.

4. The hierarchical identification method for nuclear power equipment maintenance according to claim 1, characterized in that, The step of performing vector transformation on the repair plan description based on all repair manuals to obtain the target vector expression includes: The words in the repair plan description are split into at least one descriptive word group consisting of descriptive words and connecting words. The descriptive words are any words in the repair plan description, and the connecting words are the descriptive words and another word in the repair plan description that is continuous with the descriptive words. Calculate the third occurrence frequency and third inverse document frequency of the connective word in all maintenance manuals. Based on the fourth occurrence frequency and fourth inverse document frequency of the descriptive word, as well as the third word and third inverse document frequency of the connective word, perform vector transformation on the maintenance solution description to obtain the target vector expression.

5. The hierarchical identification method for nuclear power equipment maintenance according to claim 1, characterized in that, The step of querying the maintenance classification vector table based on the target vector expression to obtain matching results and determining the maintenance classification corresponding to the matching results includes: Traverse the maintenance classification vector table, calculate the similarity between the target vector expression and all vectors in the maintenance classification vector table, and obtain the similarity result; Based on the similarity results, the target vector is queried in the maintenance classification vector table to obtain matching results, and the maintenance classification corresponding to the matching results is determined.

6. The hierarchical identification method for nuclear power equipment maintenance according to claim 5, characterized in that, After querying the target vector in the maintenance grading vector table based on the similarity result to obtain the matching result, the method further includes: Based on the similarity results, the matching results are sorted in descending order of similarity to obtain the vector sorting results; Based on a preset candidate range, the vector sorting results are filtered to obtain the vector selection results; Based on the vector selection results, the maintenance level corresponding to the matching result is determined.

7. The hierarchical identification method for nuclear power equipment maintenance according to any one of claims 1 to 6, characterized in that, After querying the maintenance classification vector table based on the target vector expression to obtain a matching result and determining the maintenance classification corresponding to the matching result, the process further includes: Obtain a preset table template and extract the title field from the table template; Based on the title field, the maintenance classification information is categorized to obtain classification information; Fill the table template with the category information to obtain the display table.

8. A grading and identification device for nuclear power equipment maintenance, characterized in that, include: The word segmentation module is used to obtain a real-time maintenance classification table and at least one latest version of the maintenance manual, and to segment the cell text in each cell of the maintenance classification table to obtain N words for the corresponding cell, where N is an integer greater than zero. The frequency calculation module is used to calculate the first occurrence frequency and the first inverse document frequency of each word in the cell in all maintenance manuals for any given cell. The filtering module is used to filter the word segments from the cell text that have a first inverse document frequency of zero and a first occurrence frequency greater than a first frequency threshold, so as to obtain the filtered text corresponding to the cell. The word segmentation module is used to segment the words in the filtered text to form at least one first word group consisting of a first word segment and consecutive words, wherein the first word segment is any word in the filtered text, and the consecutive words segment is the first word segment combined with another word in the filtered text that is consecutive to the first word segment. The vector representation module is used to calculate the second occurrence frequency and the second inverse document frequency of the continuous word segment in all maintenance manuals, and to obtain the feature vector representation of the corresponding first word group based on the first occurrence frequency and the second inverse document frequency of the first word segment, as well as the second word segment and the second occurrence frequency of the continuous word segment. The vector table acquisition module is used to traverse all cells of the maintenance classification table and the first phrase of each cell to obtain the feature vector expression of all cells. The feature vector expression of each cell is written into the corresponding cell to obtain the maintenance classification vector table. The matching module is used to obtain the repair solution description input by the user, perform vector transformation on the repair solution description according to all repair manuals to obtain the target vector expression, query the repair classification vector table according to the target vector expression, obtain the matching result, and determine the repair classification corresponding to the matching result.

9. A computer device, characterized in that, The computer device includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the graded identification method for nuclear power equipment maintenance as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the graded identification method for nuclear power equipment maintenance as described in any one of claims 1 to 7.

Images