Power distribution equipment debugging method, electronic equipment, storage medium and program product

Abstract

The embodiment of the invention provides a debugging method of power distribution equipment, electronic equipment, a storage medium and a program product. The method comprises the following steps: acquiring an abnormal power distribution message of power distribution equipment; wherein the abnormal power distribution message is a power distribution message of the power distribution equipment extracted when the power distribution terminal of the power distribution equipment carries out analysis abnormity early warning; vectorizing the abnormal power distribution message to obtain an abnormal power distribution message vector; carrying out clustering processing on the abnormal power distribution message vector to obtain an abnormal cluster to which the abnormal power distribution message vector belongs; the abnormal cluster indicates a fault type; performing identification processing on the abnormal power distribution message vector and a fault type indicated by the abnormal cluster based on a large language model, and generating a debugging instruction set; wherein the debugging instruction set is used for debugging the power distribution equipment. The method is used for achieving the effect of improving debugging efficiency and accuracy.

Description

Technical Field

[0001] The present application relates to the field of power distribution technology, and in particular to a debugging method for power distribution equipment, electronic equipment, storage medium, and program product. Background Art

[0002] Distribution equipment refers to the various devices within a power system used to distribute electricity from the high-voltage transmission network to end users. These devices perform core functions such as voltage conversion, power distribution, and protection and control. However, if abnormalities occur during the operation of distribution equipment, they require debugging to ensure the normal operation of the power system.

[0003] In the prior art, after obtaining an abnormal power distribution message of a power distribution device, the power distribution device is usually debugged manually.

[0004] However, the manual approach can lead to low debugging efficiency and low accuracy. Summary of the Invention

[0005] The embodiments of the present application provide a debugging method, electronic equipment, storage medium and program product for power distribution equipment, so as to achieve the effect of improving debugging efficiency and accuracy.

[0006] In a first aspect, an embodiment of the present application provides a method for debugging a power distribution device, comprising:

[0007] Acquire an abnormal power distribution message of the power distribution equipment; wherein the abnormal power distribution message is a power distribution message of the power distribution equipment extracted when the power distribution terminal of the power distribution equipment performs abnormality warning analysis;

[0008] Performing vectorization processing on the abnormal power distribution message to obtain an abnormal power distribution message vector; and performing clustering processing on the abnormal power distribution message vector to obtain an abnormal cluster to which the abnormal power distribution message vector belongs; the abnormal cluster indicates a fault type;

[0009] Based on a large language model, the abnormal power distribution message vector and the fault type indicated by the abnormal cluster are identified and processed to generate a debugging instruction set; wherein, the debugging instruction set is used to debug the power distribution equipment.

[0010] In one possible implementation, the abnormal power distribution message indicates the type of power distribution equipment; the abnormal power distribution message is vectorized to obtain an abnormal power distribution message vector, including: obtaining a corresponding preset bag-of-words model according to the type of power distribution equipment indicated by the abnormal power distribution message; and vectorizing the abnormal power distribution message according to the corresponding preset bag-of-words model to obtain an abnormal power distribution message vector.

[0011] In one possible embodiment, the preset bag-of-words model indicates a set of terms corresponding to the type of power distribution equipment; the term set includes at least one term; the abnormal power distribution message is vectorized according to the corresponding preset bag-of-words model to obtain an abnormal power distribution message vector, including: performing field extraction processing on the abnormal power distribution message to obtain multiple message fields; matching processing is performed on each of the multiple message fields based on at least one term included in the term set to obtain a target term corresponding to the abnormal power distribution message and the frequency of each target term; the abnormal power distribution message vector is determined based on the target term and the frequency of each target term.

[0012] In one possible embodiment, the preset bag-of-words model indicates a set of terms corresponding to the type of distribution equipment; the set of terms includes at least one term; wherein the terms include numerical terms and / or category terms; the numerical terms include numerical types and numerical ranges; and category terms indicate instruction categories and / or status categories.

[0013] In one possible implementation, the abnormal power distribution message indicates the type of power distribution equipment; clustering is performed on the abnormal power distribution message vector to obtain the abnormal cluster to which the abnormal power distribution message vector belongs, including: obtaining a corresponding preset K-means clustering model according to the type of power distribution equipment indicated by the abnormal power distribution message; clustering is performed on the abnormal power distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs.

[0014] In one possible embodiment, the preset K-means clustering model indicates at least one cluster centroid; the cluster centroid corresponds to the abnormal cluster; the cluster centroid indicates the vector characteristics of the fault type corresponding to the abnormal cluster; the abnormal distribution message vector is clustered according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal distribution message vector belongs, including: based on the vector characteristics of the fault type, calculating the centroid offset distance between the abnormal distribution message vector and each cluster centroid of the at least one cluster centroid; determining the cluster centroid with the smallest centroid offset distance and / or the centroid offset distance less than a preset threshold as the cluster centroid corresponding to the abnormal distribution message vector, and determining the abnormal cluster corresponding to the determined cluster centroid as the abnormal cluster to which the abnormal distribution message vector belongs.

[0015] In one possible implementation, the abnormal power distribution message vector and the fault type indicated by the abnormal cluster are identified and processed based on a large language model to generate a debugging instruction set, including: generating an instruction generation prompt word according to the abnormal power distribution message vector and the fault type indicated by the abnormal cluster; the instruction generation prompt word is used to instruct the large language model to generate the debugging instruction set according to the abnormal power distribution message vector and the fault type indicated by the abnormal cluster; the instruction generation prompt word is input into the large language model, and the debugging instruction set is output.

[0016] In one possible embodiment, the instruction generation prompt word is also used to indicate the confidence level of the fault type indicated by the exception cluster generated by the large language model; the instruction generation prompt word is input into the large language model, and the debugging instruction set is output, including: inputting the instruction generation prompt word into the large language model, and if the confidence level of the fault type indicated by the generated exception cluster is greater than a preset threshold, outputting the debugging instruction set; otherwise, outputting a prompt message; wherein the prompt message includes the confidence level of the fault type indicated by the exception cluster.

[0017] In a second aspect, an embodiment of the present application provides a debugging device for power distribution equipment, comprising:

[0018] An acquisition module is used to acquire abnormal power distribution messages of the power distribution equipment; wherein the abnormal power distribution messages are power distribution messages of the power distribution equipment extracted when the power distribution terminal of the power distribution equipment performs abnormality warning analysis;

[0019] a processing module configured to perform vectorization processing on the abnormal power distribution message to obtain an abnormal power distribution message vector; and perform clustering processing on the abnormal power distribution message vector to obtain an abnormal cluster to which the abnormal power distribution message vector belongs; the abnormal cluster indicates a fault type;

[0020] A debugging module is used to identify and process the abnormal power distribution message vector and the fault type indicated by the abnormal cluster based on a large language model, and generate a debugging instruction set; wherein the debugging instruction set is used to debug the power distribution equipment.

[0021] In one possible implementation, the abnormal power distribution message indicates the type of power distribution equipment; the processing module is specifically used to obtain the corresponding preset word bag model according to the type of power distribution equipment indicated by the abnormal power distribution message; and the abnormal power distribution message is vectorized according to the corresponding preset word bag model to obtain the abnormal power distribution message vector.

[0022] In one possible embodiment, the preset bag-of-words model indicates a set of terms corresponding to the type of distribution equipment; the term set includes at least one term; the processing module is further specifically used to perform field extraction processing on the abnormal distribution message to obtain multiple message fields; based on at least one term included in the term set, each message field in the multiple message fields is matched to obtain the target term corresponding to the abnormal distribution message and the frequency of each target term; based on the target term and the frequency of each target term, the abnormal distribution message vector is determined.

[0023] In one possible embodiment, the preset bag-of-words model indicates a set of terms corresponding to the type of distribution equipment; the set of terms includes at least one term; wherein the terms include numerical terms and / or category terms; the numerical terms include numerical types and numerical ranges; and category terms indicate instruction categories and / or status categories.

[0024] In one possible implementation, the abnormal power distribution message indicates the type of power distribution equipment; the processing module is further specifically used to obtain a corresponding preset K-means clustering model based on the type of power distribution equipment indicated by the abnormal power distribution message; clustering is performed on the abnormal power distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs.

[0025] In one possible embodiment, the preset K-means clustering model indicates at least one cluster centroid; the cluster centroid corresponds to the abnormal cluster; the cluster centroid indicates the vector characteristics of the fault type corresponding to the abnormal cluster; the processing module is further specifically used to calculate the centroid offset distance between the abnormal distribution message vector and each cluster centroid of the at least one cluster centroid based on the vector characteristics of the fault type; the cluster centroid with the smallest centroid offset distance and / or the centroid offset distance less than a preset threshold is determined as the cluster centroid corresponding to the abnormal distribution message vector, and the abnormal cluster corresponding to the determined cluster centroid is determined as the abnormal cluster to which the abnormal distribution message vector belongs.

[0026] In one possible embodiment, the debugging module is specifically used to generate an instruction generation prompt word based on the abnormal power distribution message vector and the fault type indicated by the abnormal cluster; the instruction generation prompt word is used to instruct the large language model to generate the debugging instruction set based on the abnormal power distribution message vector and the fault type indicated by the abnormal cluster; the instruction generation prompt word is input into the large language model, and the debugging instruction set is output.

[0027] In one possible embodiment, the instruction generation prompt word is also used to indicate the confidence level of the fault type indicated by the exception cluster generated by the large language model; the debugging module is further specifically used to input the instruction generation prompt word into the large language model, and if the confidence level of the fault type indicated by the generated exception cluster is greater than a preset threshold, output the debugging instruction set; otherwise, output a prompt message; wherein the prompt message includes the confidence level of the fault type indicated by the exception cluster.

[0028] In a third aspect, an embodiment of the present application provides an electronic device, comprising: a memory, a processor;

[0029] The memory stores computer-executable instructions;

[0030] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementations of the first aspect.

[0031] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the first aspect above and / or various possible implementation methods of the first aspect.

[0032] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the above first aspect and / or various possible implementation methods of the first aspect.

[0033] The debugging method, electronic device, storage medium and program product of the distribution equipment provided in the embodiments of the present application obtain abnormal distribution messages of the distribution equipment, vectorize the abnormal distribution messages, and obtain abnormal distribution message vectors; and cluster the abnormal distribution message vectors to obtain the abnormal clusters to which the abnormal distribution message vectors belong, identify and process the fault types indicated by the abnormal distribution message vectors and the abnormal clusters based on a large language model, and generate a debugging instruction set, wherein the debugging instruction set is automatically generated by the large model, which can improve the debugging efficiency and accuracy, wherein, before the debugging instruction set is automatically generated by the large model, the abnormal distribution message vectors are identified and processed, which reduces the dimension of the data to be processed, which can further improve the debugging efficiency of the distribution equipment, wherein, before the debugging instruction set is automatically generated by the large model, the fault type corresponding to the abnormal cluster is identified and processed, which can further improve the debugging accuracy of the distribution equipment. In summary, the debugging method of the distribution equipment provided by the present application improves the debugging efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0035] Figure 1 A schematic diagram of a scenario for the debugging method of the power distribution equipment provided in this application;

[0036] Figure 2 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 1 ;

[0037] Figure 3 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 2 ;

[0038] Figure 4 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 3 ;

[0039] Figure 5 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 4 ;

[0040] Figure 6 A schematic diagram of the structure of the debugging device for the power distribution equipment provided in this application;

[0041] Figure 7 This is a schematic diagram of the structure of the electronic device provided in this application.

[0042] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0043] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0044] Figure 1 A schematic diagram of a debugging method for power distribution equipment provided in this application, such as Figure 1 As shown, the specific application scenario of this application includes an industrial computer 11 and at least one power distribution terminal 12;

[0045] The industrial computer 11, or industrial control computer, is a ruggedized personal computer designed specifically for industrial field environments. The power distribution terminal 12 is a general term for various remote monitoring and control units installed in the distribution network. It is primarily responsible for completing data acquisition, control, and communication functions, and is a key component of the distribution network automation system.

[0046] The industrial computer 11 is connected to each of the at least one power distribution terminal 12 to implement the method in the embodiment shown below;

[0047] Specifically, the application scenario of the present application also includes at least one power distribution device 13;

[0048] The power distribution terminal 12 is connected to at least one of the at least one power distribution equipment 13 and is used to collect message data from the connected power distribution equipment 13, transmit and analyze the collected message data, so as to assist operation and maintenance personnel in completing fault diagnosis, optimization and debugging of the power distribution equipment, and execute debugging or control instructions;

[0049] Distribution equipment 13 refers to various devices and systems used in a power system to receive, distribute, and manage electrical energy. Optionally, distribution equipment 13 includes, but is not limited to, converters, transformers, high-voltage distribution equipment, low-voltage distribution equipment, circuit breakers, fuses, transformers, cables, busbars, and other equipment.

[0050] In view of the above scenario, in the prior art, after obtaining the abnormal power distribution message of the power distribution equipment, the power distribution equipment is usually debugged manually. However, the manual method leads to low debugging efficiency and low accuracy.

[0051] The debugging method for distribution equipment provided in the present application obtains abnormal distribution messages of the distribution equipment, vectorizes the abnormal distribution messages, and obtains abnormal distribution message vectors; and clusters the abnormal distribution message vectors to obtain the abnormal cluster to which the abnormal distribution message vectors belong, identifies and processes the fault types indicated by the abnormal distribution message vectors and the abnormal clusters based on a large language model, and generates a debugging instruction set, wherein the debugging instruction set is automatically generated by the large model, which can improve the debugging efficiency and accuracy, wherein, before the debugging instruction set is automatically generated by the large model, the abnormal distribution message vectors are identified and processed, which reduces the dimension of the data to be processed, which can further improve the debugging efficiency of the distribution equipment, wherein, before the debugging instruction set is automatically generated by the large model, the fault type corresponding to the abnormal cluster is identified and processed, which can further improve the debugging accuracy of the distribution equipment. In summary, the debugging method for distribution equipment provided in the present application improves the debugging efficiency and accuracy.

[0052] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0053] Figure 2 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 1 , the execution subject is an industrial computer or server, which can generally be implemented through software, hardware, or a combination of software and hardware, such as Figure 2 As shown, this embodiment describes in detail a debugging method for power distribution equipment, which includes:

[0054] Step S201: Acquire abnormal power distribution messages of power distribution equipment.

[0055] Specifically, the abnormal power distribution message of the power distribution equipment can be obtained.

[0056] Among them, the abnormal power distribution message is the power distribution message of the power distribution equipment extracted when the power distribution terminal of the power distribution equipment analyzes the abnormality warning.

[0057] Specifically, based on Figure 1 In the description of the application scenario shown, the distribution terminal collects distribution messages from the distribution equipment and performs real-time analysis on them. If the real-time message analysis result is a message abnormality, for example, the voltage exceeds the threshold, a real-time warning of the abnormal situation will be issued, that is, an analysis abnormality warning will be generated.

[0058] Specifically, the real-time message analysis results of the distribution terminal can be monitored. If an analysis abnormality warning is monitored, the distribution message of the distribution equipment corresponding to the analysis abnormality warning, that is, the abnormal distribution message, can be extracted from all distribution messages collected by the distribution terminal.

[0059] Among them, the present application does not limit the process of extracting the distribution messages of the distribution equipment corresponding to the analysis abnormal warning from all the distribution messages collected by the distribution terminal. Optionally, the timestamp of the analysis abnormal warning can be determined first; then, the target extraction time range can be determined based on the timestamp of the analysis abnormal warning and the preset extraction time range threshold; then, the distribution messages within the target extraction time range are extracted from all the distribution messages collected by the distribution terminal, and determined as the distribution messages of the distribution equipment corresponding to the analysis abnormal warning.

[0060] Optionally, the preset extraction time range threshold can be a preset time period before the timestamp of analyzing the abnormality warning to a preset time period after the timestamp of analyzing the abnormality warning, wherein the two preset time periods can be the same value, such as 10 seconds, or different values, for example, the preset time period before the timestamp of analyzing the abnormality warning is 20 seconds, and the preset time period after the timestamp of analyzing the abnormality warning is 10 seconds.

[0061] Step S202: perform vectorization processing on the abnormal power distribution message to obtain an abnormal power distribution message vector; and perform clustering processing on the abnormal power distribution message vector to obtain an abnormal cluster to which the abnormal power distribution message vector belongs.

[0062] Specifically, due to the large number of power distribution messages and the complexity of the data, after obtaining abnormal power distribution messages, the abnormal power distribution messages can be vectorized to obtain abnormal power distribution message vectors to reduce complexity and improve data quality. Optionally, the abnormal power distribution messages can be vectorized based on a vectorization method to obtain abnormal power distribution message vectors. Optionally, vectorization methods include but are not limited to: one-hot encoding, word vector model method, bag-of-words model method, etc.

[0063] Specifically, after obtaining the abnormal power distribution message vector, the abnormal power distribution message vector can be clustered to obtain the abnormal cluster to which the abnormal power distribution message vector belongs. The abnormal cluster indicates the fault type. Optionally, the abnormal power distribution message vector can be clustered based on a clustering algorithm to obtain the abnormal cluster to which the abnormal power distribution message vector belongs. Optionally, the clustering algorithm includes but is not limited to: K-Means clustering algorithm, DBSCAN density-based clustering algorithm, hierarchical clustering algorithm, and other algorithms.

[0064] Step S203: Identify and process the fault type indicated by the abnormal power distribution message vector and the abnormal cluster based on the large language model, and generate a debugging instruction set.

[0065] Specifically, based on the large language model, the abnormal power distribution message vector obtained in step S202 and the fault type indicated by the abnormal cluster can be identified and processed to generate a debugging instruction set.

[0066] Large Language Models (LLMs) refer to deep learning models with hundreds of millions or even hundreds of billions of parameters. The following is a detailed explanation of the concept of large models:

[0067] Specifically, the main features of large models include: Large parameter scale: Large models usually contain billions or even hundreds of billions of parameters, which enables them to capture complex language regularities and patterns. For example, the parameter scale of GPT-3 has reached 175 billion. Large data demand: In order to train these models, massive amounts of data are usually required, covering multi-source information such as Internet text, books, codes, etc., and the scale can reach TB level. This data is used to enable the model to learn a wide range of language knowledge and patterns. Strong task generalization ability: By learning general knowledge through pre-training, large models can be adapted to a variety of diverse tasks such as translation, question answering, and creation, showing strong generalization ability.

[0068] Specifically, the core principles of large models are as follows: Architectural basis: Large models generally adopt the Transformer architecture, the core of which is the self-attention mechanism (Self-Attention), which can dynamically measure the importance of each word in the input sequence and capture long-distance dependencies. Training method: Large models usually use unsupervised learning to learn language patterns from massive texts, and introduce supervised data to teach the model to understand human instructions. In addition, through methods such as human feedback reinforcement learning (RLHF) or direct preference optimization (DPO), the output of the model can be made more in line with human values. Generation logic: The large model generates content through token-by-token prediction. The text is divided into the smallest semantic unit (Token), and the model predicts the probability distribution of the next Token based on the context, and controls the generation diversity through the temperature parameter (Temperature).

[0069] Specifically, large models have a wide range of applications in natural language processing, including but not limited to language understanding, information extraction, machine translation, question-answering systems, and text generation. They can also be applied to computer vision, speech recognition, recommendation systems, and the financial industry, providing accurate analysis and prediction results.

[0070] In summary, large models, with their large parameter scale, massive data requirements, strong task generalization capabilities, and wide range of application areas, are leading a new round of revolution in the field of artificial intelligence.

[0071] The debugging instruction set is used to debug the power distribution equipment. The debugging instruction set is a collection of instructions for debugging the power distribution equipment. The debugging instruction set includes at least one debugging instruction, which is used to debug the power distribution equipment to address or resolve the power distribution equipment failures described above.

[0072] Specifically, the present application identifies and processes the fault types indicated by the abnormal power distribution message vector and the abnormal cluster based on the large language model, and does not limit the process of generating a debugging instruction set. Optionally, an instruction generation prompt word can be generated first according to the fault type indicated by the abnormal power distribution message vector and the abnormal cluster; then the instruction generation prompt word is input into the large language model, and the debugging instruction set is output.

[0073] The debugging method for distribution equipment provided in an embodiment of the present application obtains abnormal distribution messages of the distribution equipment, vectorizes the abnormal distribution messages, and obtains abnormal distribution message vectors; and clusters the abnormal distribution message vectors to obtain the abnormal cluster to which the abnormal distribution message vectors belong, identifies and processes the fault types indicated by the abnormal distribution message vectors and the abnormal clusters based on a large language model, and generates a debugging instruction set, wherein the debugging instruction set is automatically generated by the large model, which can improve the debugging efficiency and accuracy, wherein, before the debugging instruction set is automatically generated by the large model, the abnormal distribution message vectors are identified and processed, which reduces the dimension of the data to be processed, which can further improve the debugging efficiency of the distribution equipment, wherein, before the debugging instruction set is automatically generated by the large model, the fault type corresponding to the abnormal cluster is identified and processed, which can further improve the debugging accuracy of the distribution equipment. In summary, the debugging method for distribution equipment provided in the present application improves the debugging efficiency and accuracy.

[0074] Figure 3 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 2 ,like Figure 3 As shown, this embodiment Figure 2 Based on the embodiment, vectorization processing of abnormal power distribution messages to obtain abnormal power distribution message vectors is described in detail. The method includes:

[0075] Step S301: Obtain a corresponding preset bag-of-words model according to the type of power distribution equipment indicated by the abnormal power distribution message.

[0076] Specifically, according to the type of power distribution equipment indicated by the abnormal power distribution message, a corresponding preset bag-of-words model can be obtained.

[0077] The Bag of Words (BoW) model is a widely used text representation method in natural language processing and information retrieval. Its core idea is to simplify text into a "bag" containing only the words and their frequencies, while ignoring the order, grammar, and semantic relationships of the words.

[0078] Among them, during the debugging process of power distribution equipment, compared with the process of message vectorization processing based on one-hot encoding and word vector model method, the message vectorization processing based on the bag-of-words model has the following advantages: simplified processing flow, intuitive reflection of key information and high computational efficiency.

[0079] Specifically, in terms of simplifying the processing flow, the bag-of-words model treats text as a simple collection of words, ignoring the order and grammatical structure between words, and only focusing on the frequency of words appearing in the text. This processing method simplifies the text vectorization process. Especially in the commissioning of power distribution equipment, messages often contain a large amount of professional terms and key information. The bag-of-words model can quickly extract the frequency characteristics of this information, facilitating subsequent analysis. One-hot encoding is also an effective text vectorization method, but it requires creating a separate binary vector for each word, resulting in a high vector dimension. Especially when the vocabulary is large, the computational complexity and storage requirements will increase significantly. Although the word vector model can capture the semantic relationship between words, it requires a large amount of corpus for training and has a high computational complexity. In the message processing of power distribution equipment commissioning, if there is a lack of sufficient training data or computing resources, the application of the word vector model may be limited.

[0080] Specifically, the bag-of-words model, by counting the frequency of word occurrences in messages, can intuitively reflect key information in messages. In power distribution equipment debugging, this key information is often closely related to equipment status, fault type, and other factors. The bag-of-words model's processing method helps quickly identify this information. While one-hot encoding can indicate the presence or absence of a word, it cannot directly reflect the frequency characteristics of a word. Therefore, it may not be as intuitive as the bag-of-words model in capturing key information in messages. While word vector models can capture semantic relationships between words, these relationships may not be the primary focus in message processing for power distribution equipment debugging. In contrast, the bag-of-words model's method of reflecting key information by counting word frequencies is more direct and effective.

[0081] Specifically, in terms of computational efficiency, the bag-of-words model, because it ignores the order and grammatical structure between words, offers high computational efficiency when processing messages. This is especially true when processing large message datasets, where the bag-of-words model can quickly complete vectorization, providing timely support for subsequent analysis. When processing large vocabularies, one-hot encoding significantly increases computational complexity and storage requirements due to the high dimensionality of the vectors, potentially impacting processing efficiency. Word vector models require calculating semantic relationships between words, which often involves complex matrix operations and deep learning algorithms. Therefore, they may not be as computationally efficient as the bag-of-words model.

[0082] In summary, using the bag-of-words model for message vectorization has the advantages of simplifying the processing flow, intuitively reflecting key information, and achieving high computational efficiency during the power distribution equipment debugging process. These advantages make the bag-of-words model an effective method for processing power distribution equipment debugging messages.

[0083] Specifically, based on the description of the abnormal power distribution message in step S201, the abnormal power distribution message is extracted when the power distribution terminal analyzes messages of various types of connected power distribution devices. Therefore, the abnormal power distribution message corresponds to the type of power distribution device.

[0084] Specifically, if the abnormal power distribution message is vectorized based on the bag-of-words model, a preset bag-of-words model corresponding to the type of power distribution equipment indicated by the abnormal power distribution message can be obtained to improve the accuracy of vectorization of the abnormal power distribution message indicating the type of power distribution equipment based on the bag-of-words model.

[0085] Step S302: vectorize the abnormal power distribution message according to the corresponding preset bag-of-words model to obtain an abnormal power distribution message vector.

[0086] Specifically, after obtaining the corresponding preset bag-of-words model according to the type of power distribution equipment indicated by the abnormal power distribution message, the abnormal power distribution message can be vectorized according to the obtained preset bag-of-words model to obtain the abnormal power distribution message vector.

[0087] Optionally, the preset bag-of-words model indicates a set of terms corresponding to the type of the power distribution equipment, wherein the set of terms includes at least one term.

[0088] The terms include numerical terms and / or category terms. The numerical terms include numerical types and numerical ranges.

[0089] Specifically, the term is a preset field referenced during field extraction during vectorization of abnormal power distribution messages. The term set is a set consisting of terms corresponding to the types of power distribution equipment indicated by the preset bag-of-words model.

[0090] Among them, when the types of power distribution equipment are different, the corresponding terms and term sets are also different.

[0091] Among them, the numerical term is a term that represents the numerical range of a numerical type. The present application does not limit the numerical type. Optionally, it can be one of the numerical types such as voltage, current, temperature, etc. The present application does not limit the numerical range. Optionally, multiple numerical ranges can be set for the same numerical type, for example, "voltage_1-5V", "temperature>80℃"; for example, "voltage_1-5V", "voltage_5-10V".

[0092] The category term indicates the instruction category and / or status category, for example, "communication status code_0xEE"; for example, "command A_0x01".

[0093] Among them, in the process of vectorizing the abnormal distribution message according to the corresponding preset bag-of-words model to obtain the abnormal distribution message vector, the diversity of the target term is improved by setting numerical terms and type terms in the term set, thereby improving the accuracy of determining the abnormal distribution message vector and further improving the accuracy of the debugging method of the distribution equipment.

[0094] Optionally, the preset bag-of-words model indicates a set of terms corresponding to the type of the power distribution equipment, wherein the set of terms includes at least one term.

[0095] The abnormal power distribution message is vectorized according to the corresponding preset bag-of-words model to obtain the abnormal power distribution message vector, including:

[0096] Field extraction processing is performed on the abnormal power distribution message to obtain multiple message fields.

[0097] A matching process is performed on each of the multiple message fields based on at least one term included in the term set to obtain a target term corresponding to the abnormal power distribution message and the frequency of each target term.

[0098] According to the target terms and the frequencies of the target terms, the abnormal power distribution message vector is determined.

[0099] Among them, the present application does not limit the process of performing field extraction and processing on the abnormal power distribution message to obtain multiple message fields. Optionally, the communication protocol corresponding to the abnormal power distribution message can be determined first; then the abnormal power distribution message can be parsed and processed according to the communication protocol to obtain the parsed abnormal power distribution message; then, according to the specifications indicated by the communication protocol, the message fields can be extracted from the parsed abnormal power distribution message.

[0100] Specifically, in the process of matching each of the multiple message fields based on at least one term included in the term set to obtain the target term corresponding to the abnormal power distribution message and the frequency of each target term, if there is a message field that is identical to a term included in the term set, then the message field is determined to be the target term corresponding to the abnormal power distribution message, and the frequency of the target term corresponding to the target term is increased by one until all message fields are matched, wherein the initial frequency of the target term is zero. Specifically, if there is a message field that is different from all terms included in the term set, then it is determined that the field is not the target term corresponding to the abnormal power distribution message.

[0101] Specifically, after traversing all message fields, a sparse vector can be constructed based on all the target terms and the frequency of each target term, for example: [voltage_5-10V: 3 times, temperature>80℃: 6 times, communication status code_0xEE: 2 times, ...].

[0102] Among them, in the process of vectorizing the abnormal distribution message according to the corresponding preset word bag model to obtain the abnormal distribution message vector, the set of terms corresponding to the type of distribution equipment indicated by the preset word bag model can accurately determine the target term and the frequency of the target term, thereby improving the accuracy of the abnormal distribution message vector and further improving the accuracy of the debugging method of the distribution equipment.

[0103] The embodiment of the present application provides a process for vectorizing abnormal power distribution messages to obtain abnormal power distribution message vectors, by obtaining a corresponding preset bag-of-words model according to the type of power distribution equipment indicated by the abnormal power distribution message, and vectorizing the abnormal power distribution message according to the corresponding preset bag-of-words model to obtain the abnormal power distribution message vector, wherein, based on the characteristics of the bag-of-words model described above, vectorizing the abnormal power distribution message based on the bag-of-words model can improve the efficiency of vectorization processing, wherein, in the process of obtaining the bag-of-words model, obtaining the bag-of-words model corresponding to the type of power distribution equipment indicated by the abnormal power distribution message can improve the accuracy of vectorization processing, based on the above description, the process for vectorizing the abnormal power distribution message to obtain the abnormal power distribution message vector provided by the present embodiment can improve the efficiency and accuracy of vectorization processing, and further improve the debugging efficiency and accuracy of power distribution equipment.

[0104] Figure 4 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 3 ,like Figure 4 As shown, this embodiment Figure 2 or Figure 3 Based on the embodiment, clustering is performed on the abnormal power distribution message vector to obtain the abnormal cluster to which the abnormal power distribution message vector belongs. The method includes:

[0105] Step S401: According to the type of power distribution equipment indicated by the abnormal power distribution message, a corresponding preset K-means clustering model is obtained.

[0106] Specifically, the K-means clustering model is an unsupervised clustering algorithm that divides a given sample set into K clusters according to the distance between the samples, so that the points within the cluster are connected as closely as possible, while the distance between clusters is as large as possible.

[0107] Among them, during the debugging process of power distribution equipment, compared with the process of clustering message vectors using the density-based clustering algorithm DBSCAN and the hierarchical clustering algorithm, the clustering process of message vectors based on the K-means clustering model has the following advantages: the algorithm principle is simple and intuitive, the processing efficiency of large data sets is high, and the interpretability is strong.

[0108] Specifically, the K-Means algorithm, with its simple and intuitive principle, divides data points into K clusters through iterative calculations. Each cluster has a center (centroid), and data points are assigned to the closest cluster based on their distance to the cluster center. This simple principle makes the K-Means algorithm easy to understand and implement, making it particularly easy to quickly grasp and apply to real-world problems, particularly when dealing with abnormal message vectors during power distribution equipment debugging.

[0109] Specifically, the K-Means algorithm is highly efficient when processing large datasets. This is because it gradually optimizes clustering results by iteratively updating the positions of cluster centers. Each iteration involves only calculating the distance from each data point to the cluster center and reallocating the data point to the nearest cluster. This iterative process maintains good scalability even on large datasets, making the K-Means algorithm advantageous when processing large numbers of abnormal message vectors.

[0110] Specifically, the K-Means algorithm's clustering results are highly interpretable. Each cluster has a clear center point that can be used to explain and describe its characteristics. This helps understand the characteristics of different fault types during power distribution equipment commissioning, leading to more accurate fault identification. Furthermore, by visualizing the K-Means algorithm's clustering results, the distribution of different clusters can be intuitively seen, facilitating further analysis and optimization of clustering results.

[0111] In summary, the K-Means algorithm-based clustering of abnormal power distribution message vectors has the advantages of simple and intuitive algorithmic principles, high efficiency in processing large data sets, and strong interpretability. These advantages make the K-Means algorithm widely applicable in fault identification and analysis during power distribution equipment commissioning.

[0112] Specifically, based on the description of the abnormal power distribution message in step S201, the abnormal power distribution message is extracted when the power distribution terminal analyzes messages of various types of connected power distribution devices. Therefore, the abnormal power distribution message corresponds to the type of power distribution device.

[0113] Based on the advantages of the K-Means algorithm described above, it is also necessary to note the limitations of the K-Means algorithm. For example, it is sensitive to the selection of the initial cluster center, and different initial cluster centers may lead to different clustering results; the K value needs to be set manually, and the selection of this value is usually difficult; it is sensitive to noise and outliers, which may cause the center of the cluster to shift or produce undesirable clusters; and it is not suitable for non-convex clusters, etc. Therefore, in practical applications, it is necessary to select a suitable clustering algorithm based on specific scenarios and needs, and optimize and adjust it in combination with domain knowledge and expert experience. Specifically, if the abnormal distribution message vector is clustered based on the K-means clustering model to obtain the abnormal cluster to which the abnormal distribution message vector belongs, then the preset K-means clustering model corresponding to the type of distribution equipment indicated by the abnormal distribution message can be obtained to improve the accuracy of clustering the abnormal distribution message vector indicating the type of distribution equipment based on the K-means clustering model.

[0114] Step S402: cluster the abnormal power distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs.

[0115] Specifically, after obtaining the corresponding preset K-means clustering model according to the type of distribution equipment indicated by the abnormal distribution message, the abnormal distribution message vector can be clustered according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal distribution message vector belongs.

[0116] Optionally, the preset K-means clustering model indicates at least one cluster centroid. The cluster centroid corresponds to the abnormal cluster. The cluster centroid indicates a vector feature of the fault type corresponding to the abnormal cluster.

[0117] Specifically, based on the description in step S401, the preset K-means clustering model is obtained based on the type of power distribution equipment indicated by the abnormal power distribution message. Therefore, the fault type indicated by each cluster centroid in the at least one cluster centroid indicated by the preset K-means clustering model corresponds to the type of power distribution equipment indicated by the abnormal power distribution message. For example, if the type of power distribution equipment is a converter, the abnormal clusters indicated by each cluster centroid in the at least one cluster centroid indicated by the preset K-means clustering model include: a communication fault (k=1), a heat dissipation system overheating fault (k=2), a DC bus overvoltage fault (k=3), an insulated gate bipolar transistor module (IGBT) overload fault (k=4), a control instruction conflict fault (k=5), a local short circuit fault (k=6), and the like. For example, if the abnormal cluster is a communication failure (k=1), the vector features of the corresponding fault type can be the high voltage fluctuation and communication interruption terms; if the abnormal cluster is a cooling system overheating failure (k=2), the vector features of the corresponding fault type can be temperature exceeding the limit and control instruction conflict.

[0118] Optionally, clustering is performed on the abnormal power distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs, including:

[0119] Based on the vector characteristics of the fault type, a centroid offset distance between the abnormal power distribution message vector and each cluster centroid in at least one cluster centroid is calculated.

[0120] The cluster centroid with the smallest centroid offset distance and / or a centroid offset distance less than a preset threshold is determined as the cluster centroid corresponding to the abnormal power distribution message vector, and the abnormal cluster corresponding to the determined cluster centroid is determined as the abnormal cluster to which the abnormal power distribution message vector belongs.

[0121] The present application does not limit the centroid offset distance, and it can optionally be a Euclidean offset distance or a cosine offset distance.

[0122] Among them, in the process of clustering the abnormal distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal distribution message vector belongs, according to at least one cluster centroid indicated by the preset K-means clustering model, the abnormal cluster to which the abnormal distribution message vector belongs can be accurately determined, thereby further improving the accuracy of the debugging method of the distribution equipment.

[0123] The embodiment of the present application provides a process for clustering abnormal power distribution message vectors to obtain the abnormal cluster to which the abnormal power distribution message vector belongs. The process obtains the corresponding preset K-means clustering model according to the type of power distribution equipment indicated by the abnormal power distribution message, and clusters the abnormal power distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs. Based on the characteristics of the K-means clustering model described above, clustering the abnormal power distribution message vector based on the K-means clustering model can improve the efficiency of clustering. In the process of obtaining the K-means clustering model, obtaining the K-means clustering model corresponding to the type of power distribution equipment indicated by the abnormal power distribution message can improve the accuracy of vectorization. In summary, the process of clustering the abnormal power distribution message vectors provided by the present embodiment can improve the efficiency and accuracy of clustering, and further improve the debugging efficiency and accuracy of the power distribution equipment.

[0124] Figure 5 Schematic diagram of the process of debugging the power distribution equipment provided in this application Figure 4 ,like Figure 5 As shown, this embodiment Figure 2 or Figure 3 or Figure 4 Based on the embodiment, the fault type indicated by the abnormal power distribution message vector and the abnormal cluster is identified and processed based on the large language model, and a debugging instruction set is generated. The method includes:

[0125] Step S501: Generate an instruction generation prompt word according to the abnormal power distribution message vector and the fault type indicated by the abnormal cluster.

[0126] The instruction generation prompt word is used to instruct the large language model to generate a debugging instruction set according to the abnormal power distribution message vector and the fault type indicated by the abnormal cluster.

[0127] Specifically, the present application does not limit the form of the instruction generation prompt word. Any word that can be used to instruct the large language model to generate a debugging instruction set based on the abnormal power distribution message vector and the fault type indicated by the abnormal cluster can be used as the instruction generation prompt word provided by the present application, for example, "Please generate a debugging instruction set based on the abnormal power distribution message vector... and the fault type indicated by the abnormal cluster..., where the generation requirements of the debugging instruction set are..."

[0128] Step S502: Input the instruction generation prompt words into the large language model and output a debugging instruction set.

[0129] Specifically, the instruction generation prompt words generated in step S501 are input into the large language model, and a debugging instruction set is output.

[0130] Input the instruction generation prompt words into the large language model and output a set of debugging instructions, including:

[0131] The instruction generation prompt is input into the large language model. If the confidence level of the fault type indicated by the generated exception cluster exceeds a preset threshold, the debug instruction set is output. Otherwise, a prompt message is output. The prompt message includes the confidence level of the fault type indicated by the exception cluster.

[0132] The command generation prompt also indicates the confidence level of the fault type indicated by the anomaly cluster generated by the large language model. For example, "Please determine the confidence level of the fault type indicated by the anomaly cluster based on the abnormal power distribution message vector... If the confidence level is greater than the threshold..., then generate a debugging command set based on the abnormal power distribution message vector... and the fault type indicated by the anomaly cluster..., where the generation requirement for the debugging command set is... Otherwise, output the prompt message: The confidence level of the fault type indicated by the anomaly cluster is... ..."

[0133] Among them, by modifying the way of generating prompt words of instructions, the large voice model is instructed to perform confidence calculation on the abnormal cluster to which the abnormal distribution message vector determined in step S402 belongs before generating the debugging instruction set, thereby improving the generation efficiency and accuracy of the debugging instruction set, and further improving the efficiency and accuracy of power equipment debugging.

[0134] The embodiment of the present application provides a method for identifying and processing the fault types indicated by abnormal power distribution message vectors and abnormal clusters based on a large language model, generating a debugging instruction set, generating instruction generation prompt words based on the fault types indicated by the abnormal power distribution message vectors and abnormal clusters, inputting the instruction generation prompt words into the large language model, and outputting a debugging instruction set. Specifically, by generating instruction generation prompt words, the generation process and generation range of the debugging instruction set are limited, so that the generated debugging instruction set is more accurate, further improving the accuracy of power equipment debugging.

[0135] Figure 6 The schematic diagram of the structure of the debugging device for the power distribution equipment provided in this application is as follows: Figure 6 As shown, the debugging device 60 of the power distribution equipment provided in this embodiment includes:

[0136] The acquisition module 601 is used to acquire abnormal power distribution messages of the power distribution equipment; wherein the abnormal power distribution messages are the power distribution messages of the power distribution equipment extracted when the power distribution terminal of the power distribution equipment performs abnormality warning analysis;

[0137] Processing module 602 is used to perform vector processing on the abnormal power distribution message to obtain an abnormal power distribution message vector; and perform clustering processing on the abnormal power distribution message vector to obtain an abnormal cluster to which the abnormal power distribution message vector belongs; the abnormal cluster indicates the fault type;

[0138] The debugging module 603 is used to identify and process the fault type indicated by the abnormal power distribution message vector and the abnormal cluster based on the large language model, and generate a debugging instruction set; wherein the debugging instruction set is used to debug the power distribution equipment.

[0139] In a possible embodiment, the abnormal power distribution message indicates the type of power distribution equipment; the processing module 602 is specifically used to obtain the corresponding preset word bag model according to the type of power distribution equipment indicated by the abnormal power distribution message; the abnormal power distribution message is vectorized according to the corresponding preset word bag model to obtain the abnormal power distribution message vector.

[0140] In a possible embodiment, a preset bag-of-words model indicates a set of terms corresponding to the type of power distribution equipment; the term set includes at least one term; the processing module 602 is further specifically used to perform field extraction processing on the abnormal power distribution message to obtain multiple message fields; based on at least one term included in the term set, each message field in the multiple message fields is matched to obtain the target term corresponding to the abnormal power distribution message and the frequency of each target term; based on the target term and the frequency of each target term, the abnormal power distribution message vector is determined.

[0141] In a possible embodiment, a preset bag-of-words model indicates a set of terms corresponding to the type of power distribution equipment; the set of terms includes at least one term; wherein the terms include numerical terms and / or category terms; the numerical terms include numerical types and numerical ranges; the category terms indicate instruction categories and / or status categories.

[0142] In one possible embodiment, the abnormal power distribution message indicates the type of power distribution equipment; the processing module 602 is further specifically used to obtain the corresponding preset K-means clustering model according to the type of power distribution equipment indicated by the abnormal power distribution message; clustering the abnormal power distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs.

[0143] In a possible embodiment, a preset K-means clustering model indicates at least one cluster centroid; the cluster centroid corresponds to an abnormal cluster; the cluster centroid indicates a vector feature of a fault type corresponding to the abnormal cluster; the processing module 602 is further specifically used to calculate the centroid offset distance between the abnormal power distribution message vector and each cluster centroid in at least one cluster centroid based on the vector feature of the fault type; the cluster centroid with the smallest centroid offset distance and / or the centroid offset distance less than a preset threshold is determined as the cluster centroid corresponding to the abnormal power distribution message vector, and the abnormal cluster corresponding to the determined cluster centroid is determined as the abnormal cluster to which the abnormal power distribution message vector belongs.

[0144] In a possible embodiment, the debugging module 603 is specifically used to generate an instruction generation prompt word based on the fault type indicated by the abnormal power distribution message vector and the abnormal cluster; the instruction generation prompt word is used to instruct the large language model to generate a debugging instruction set based on the fault type indicated by the abnormal power distribution message vector and the abnormal cluster; the instruction generation prompt word is input into the large language model, and the debugging instruction set is output.

[0145] In one possible embodiment, the instruction generation prompt word is also used to indicate the confidence level of the fault type indicated by the exception cluster generated by the large language model; the debugging module 603 is further specifically used to input the instruction generation prompt word into the large language model, and if the confidence level of the fault type indicated by the generated exception cluster is greater than a preset threshold, a debugging instruction set is output; otherwise, a prompt message is output; wherein the prompt message includes the confidence level of the fault type indicated by the exception cluster.

[0146] The debugging device for power distribution equipment provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effects are similar, and are not described in detail in this embodiment.

[0147] Figure 7 This is a schematic diagram of the structure of the electronic device provided in this application. Figure 7 As shown, the electronic device 70 provided in this embodiment includes: at least one processor 701 and a memory 702. Optionally, the electronic device 70 further includes a communication component 703. The processor 701, the memory 702 and the communication component 703 are connected via a bus 704.

[0148] During the specific implementation process, at least one processor 701 executes the computer-executable instructions stored in the memory 702, so that the at least one processor 701 performs the above method.

[0149] The specific implementation process of the processor 701 can be found in the above method embodiment. Its implementation principle and technical effects are similar and will not be repeated here in this embodiment.

[0150] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASICs), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly executed by a hardware processor or by a combination of hardware and software modules within the processor.

[0151] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage.

[0152] A bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be categorized as address buses, data buses, and control buses. For ease of illustration, the buses in the drawings of this application are not limited to just one bus or just one type of bus.

[0153] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0154] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.

[0155] The readable storage medium may be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium may be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0156] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be an integral part of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0157] The division of units is merely a logical functional division; actual implementations may employ alternative divisions, such as combining or integrating multiple units or components into another system, or omitting or disabling certain features. Furthermore, any direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection between devices or units, either through an interface, electrical, mechanical, or other means.

[0158] Units described as separate components may or may not be physically separate, and 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 these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0159] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0160] If a function 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, the technical solution of the present invention, or the portion that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the method of the present invention. The aforementioned storage medium includes various media that can store program code, such as USB flash drives, mobile hard drives, read-only memories (ROMs), random access memories (RAMs), magnetic disks, or optical disks.

[0161] Those skilled in the art will appreciate that all or part of the steps in the above-described method embodiments can be implemented using hardware associated with program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0162] Finally, it should be noted that those skilled in the art will readily identify other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art not disclosed herein. The present invention is not limited to the precise structure described above and illustrated in the accompanying drawings, and various modifications and variations may be made without departing from the scope thereof. The scope of the present invention is limited solely by the appended claims.

Claims

1. A method for debugging a power distribution device, characterized in that: Applied to industrial computers, including: Acquire an abnormal power distribution message of the power distribution equipment; wherein the abnormal power distribution message is a power distribution message of the power distribution equipment extracted when the power distribution terminal of the power distribution equipment performs abnormality warning analysis; Performing vectorization processing on the abnormal power distribution message to obtain an abnormal power distribution message vector; and performing clustering processing on the abnormal power distribution message vector to obtain an abnormal cluster to which the abnormal power distribution message vector belongs; the abnormal cluster indicates a fault type; Based on a large language model, the abnormal power distribution message vector and the fault type indicated by the abnormal cluster are identified and processed to generate a debugging instruction set; wherein, the debugging instruction set is used to debug the power distribution equipment.

2. The method according to claim 1, characterized in that The abnormal power distribution message indicates the type of power distribution equipment; Vectoring the abnormal power distribution message to obtain an abnormal power distribution message vector includes: According to the type of power distribution equipment indicated by the abnormal power distribution message, obtaining a corresponding preset bag-of-words model; The abnormal power distribution message is vectorized according to the corresponding preset bag-of-words model to obtain an abnormal power distribution message vector.

3. The method according to claim 2, characterized in that The preset bag-of-words model indicates a set of terms corresponding to the type of power distribution equipment; the set of terms includes at least one term; The abnormal power distribution message is vectorized according to the corresponding preset bag-of-words model to obtain an abnormal power distribution message vector, including: Performing field extraction processing on the abnormal power distribution message to obtain multiple message fields; Performing matching processing on each of the plurality of message fields based on at least one term included in the term set to obtain a target term corresponding to the abnormal power distribution message and a frequency of each target term; The abnormal power distribution message vector is determined according to the target term and the frequency of each target term.

4. The method according to claim 2, characterized in that The preset bag-of-words model indicates a set of terms corresponding to the type of power distribution equipment; the set of terms includes at least one term; Among them, the terms include numerical terms and / or category terms; the numerical terms include numerical types and numerical ranges; and category terms indicate instruction categories and / or status categories.

5. The method according to claim 1, wherein The abnormal power distribution message indicates the type of power distribution equipment; Clustering the abnormal power distribution message vector to obtain the abnormal cluster to which the abnormal power distribution message vector belongs includes: According to the type of power distribution equipment indicated by the abnormal power distribution message, a corresponding preset K-means clustering model is obtained; The abnormal power distribution message vector is clustered according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs.

6. The method according to claim 5, characterized in that The preset K-means clustering model indicates at least one cluster centroid; the cluster centroid corresponds to the abnormal cluster; the cluster centroid indicates a vector feature of the fault type corresponding to the abnormal cluster; Clustering the abnormal power distribution message vector according to the corresponding preset K-means clustering model to obtain the abnormal cluster to which the abnormal power distribution message vector belongs, including: Calculating a centroid offset distance between the abnormal power distribution message vector and each cluster centroid of the at least one cluster centroid based on the vector feature of the fault type; The cluster centroid with the smallest centroid offset distance and / or the centroid offset distance less than a preset threshold is determined as the cluster centroid corresponding to the abnormal power distribution message vector, and the abnormal cluster corresponding to the determined cluster centroid is determined as the abnormal cluster to which the abnormal power distribution message vector belongs.

7. The method according to any one of claims 1 to 6, characterized in that Identifying and processing the abnormal power distribution message vector and the fault type indicated by the abnormal cluster based on a large language model, and generating a debugging instruction set, including: Generate an instruction generation prompt word according to the abnormal power distribution message vector and the fault type indicated by the abnormal cluster; the instruction generation prompt word is used to instruct the large language model to generate the debugging instruction set according to the abnormal power distribution message vector and the fault type indicated by the abnormal cluster; The instruction generation prompt word is input into the large language model, and the debugging instruction set is output.

8. The method according to claim 7, characterized in that The instruction generates a prompt word, and is also used to indicate the confidence level of the fault type indicated by the abnormal cluster generated by the large language model; Inputting the instruction generation prompt word into the large language model and outputting the debugging instruction set includes: The instruction generation prompt word is input into the large language model. If the confidence of the fault type indicated by the generated exception cluster is greater than a preset threshold, the debugging instruction set is output; otherwise, a prompt message is output; wherein the prompt message includes the confidence of the fault type indicated by the exception cluster.

9. A debugging device for power distribution equipment, characterized in that: include: An acquisition module is used to acquire abnormal power distribution messages of the power distribution equipment; wherein the abnormal power distribution messages are power distribution messages of the power distribution equipment extracted when the power distribution terminal of the power distribution equipment performs abnormality warning analysis; a processing module, configured to perform vector processing on the abnormal power distribution message to obtain an abnormal power distribution message vector; and perform clustering processing on the abnormal power distribution message vector to obtain an abnormal cluster to which the abnormal power distribution message vector belongs; the abnormal cluster indicates a fault type; A debugging module is used to identify and process the abnormal power distribution message vector and the fault type indicated by the abnormal cluster based on a large language model, and generate a debugging instruction set; wherein the debugging instruction set is used to debug the power distribution equipment.

10. An electronic device, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 8 when executed by a processor.

12. A computer program product, characterized in that The invention comprises a computer program, which implements the method according to any one of claims 1 to 8 when executed by a processor.

Images