An operation ticket generation method, apparatus, device, and medium

Abstract

Embodiments of the present application disclose a kind of operation ticket generation method, device, equipment and medium.Therein, the method includes: obtaining target maintenance order, target maintenance order includes maintenance task description;Parameter identification result is obtained by target parameter identification to target maintenance order, determine candidate query object according to parameter identification result;Wherein, target parameter includes target task type and target task key element;According to the power knowledge graph constructed in advance, determine the target query object associated with the candidate query object;Wherein, the power knowledge graph includes power equipment entity, the attribute information of power equipment entity and the association relationship between power equipment entity;According to target task type and target query object, generate target operation ticket.This technical solution can automatically generate high-accuracy and high-standard operation ticket based on power knowledge graph, improve the generation efficiency of operation ticket, and reduce the generation cost of operation ticket.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power dispatching, and in particular to an operation ticket generation method, device, equipment and medium. BACKGROUND

[0002] An operation ticket refers to a written basis for electrical operation in a power system. Dispatchers need to perform operations according to the procedures and steps specified in the operation ticket to ensure stable operation of the power system and normal work of the equipment, and to reduce the occurrence of accidents and failures.

[0003] In related solutions, an artificial ticketing method is used, in which experienced power system dispatchers manually prepare operation tickets. However, the artificial ticketing method has the following problems: 1. In the artificial ticketing process, there is a risk of writing errors due to human negligence, fatigue or misunderstanding, resulting in inaccurate or incomplete operation tickets. 2. Artificial ticketing is easily influenced by personal preferences and experience, resulting in a lack of standardization in the written operation tickets, which brings difficulties to dispatcher training, shift handover and review. 3. Artificial ticketing requires a lot of time and effort to collect information and write operation steps, and the human and time costs are high. SUMMARY

[0004] The present application provides an operation ticket generation method, device, equipment and medium, which can automatically generate operation tickets with high accuracy and high standardization based on a power knowledge graph, improving the generation efficiency of operation tickets and reducing the generation cost of operation tickets.

[0005] According to an aspect of the present application, an operation ticket generation method is provided, which comprises:

[0006] Obtaining a target maintenance order, which includes a maintenance task description;

[0007] Performing target parameter identification on the target maintenance order to obtain a parameter identification result, and determining a candidate query object according to the parameter identification result; wherein the target parameters include a target task type and a target task key element;

[0008] Determining a target query object associated with the candidate query object according to a pre-constructed power knowledge graph; wherein the power knowledge graph includes power equipment entities, attribute information of the power equipment entities and association relationships between the power equipment entities;

[0009] Generating a target operation ticket according to the target task type and the target query object.

[0010] According to another aspect of the present application, an operation ticket generation device is provided, which comprises:

[0011] The target maintenance order acquisition module is configured to acquire a target maintenance order, wherein the target maintenance order comprises a maintenance task description;

[0012] The candidate query object determination module is configured to perform target parameter identification on the target maintenance order to obtain a parameter identification result, and determine a candidate query object according to the parameter identification result, wherein the target parameter comprises a target task type and a target task key element.

[0013] The target query object determination module is configured to determine a target query object associated with the candidate query object according to a pre-constructed power knowledge graph, wherein the power knowledge graph comprises power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities.

[0014] The target operation ticket generation module is configured to generate a target operation ticket according to the target task type and the target query object.

[0015] According to another aspect of the present application, an electronic device is provided, which comprises:

[0016] at least one processor; and

[0017] a memory connected to the at least one processor in communication; wherein

[0018] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the operation ticket generation method according to any one of the embodiments of the present application.

[0019] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to perform the operation ticket generation method according to any one of the embodiments of the present application when executed.

[0020] The technical scheme of the embodiments of the present application acquires a target maintenance order, wherein the target maintenance order comprises a maintenance task description; performs target parameter identification on the target maintenance order to obtain a parameter identification result, and determines a candidate query object according to the parameter identification result, wherein the target parameter comprises a target task type and a target task key element; determines a target query object associated with the candidate query object according to a pre-constructed power knowledge graph, wherein the power knowledge graph comprises power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities; and generates a target operation ticket according to the target task type and the target query object. The technical scheme can automatically generate an operation ticket with high accuracy and high standardization based on the power knowledge graph, improves the generation efficiency of the operation ticket, and reduces the generation cost of the operation ticket.

[0021] It should be understood that nothing in this section is intended to limit the scope of the embodiments of the present application nor are they intended to represent key or essential features of the embodiments of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative effort.

[0023] Figure 1 is a flow chart of an operation ticket generation method according to an embodiment of the present application;

[0024] Figure 2 is a flow chart of an operation ticket generation method according to an embodiment of the present application;

[0025] Figure 3 is a flow chart of an operation ticket generation method according to an embodiment of the present application;

[0026] Figure 4 is a structural schematic diagram of an operation ticket generation device according to an embodiment of the present application;

[0027] Figure 5 is a structural schematic diagram of an electronic device for implementing an operation ticket generation method according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the technical personnel in the art better understand the present application, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should be within the scope of protection of the present application.

[0029] It is to be understood that the terms "first", "second", "target", etc. in the description and claims of the application and the above drawings are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0030] Embodiment one

[0031] Figure 1 A flowchart of an operation ticket generation method provided for embodiment one of the application, the embodiment can be applicable to automatically and accurately generate standardized operation tickets, the method can be executed by an operation ticket generation device, the operation ticket generation device can be realized in the form of hardware and / or software, and the operation ticket generation device can be configured in an electronic device with data processing capability. As shown in the figure, the method comprises: Figure 1

[0032] S110, obtaining a target maintenance order, the target maintenance order including maintenance task description.

[0033] In the embodiment, first, the target maintenance order is obtained, and the target maintenance order includes maintenance task description. Specifically, the target maintenance order can be in text form, and the maintenance task description can include task type, device name, operation requirement and time, etc. Among them, the task type can include power equipment fault diagnosis, power load prediction, power market price prediction, power system state monitoring, etc.

[0034] S120, target parameter identification is performed on the target maintenance order to obtain parameter identification result, and candidate query objects are determined according to the parameter identification result.

[0035] ​The target parameters include a target task type and a target task key element. In this embodiment, after obtaining the target maintenance order, the target maintenance order can be subjected to target parameter identification to obtain a parameter identification result. The target parameters include the target task type and the target task key element. The task key element can include a device name, an operation type, a time limit, and the like. For example, the target maintenance order can be input into a pre-trained intent identification model. The target maintenance order can be subjected to target parameter identification by using the intent identification model. The parameter identification result can be determined according to an output result of the intent identification model. The intent identification model is trained based on the candidate maintenance order and the parameters (the task type and the task key element) manually labeled for the candidate maintenance order.

[0036] After obtaining the parameter identification result, the candidate query object can be determined according to the parameter identification result. Optionally, the candidate query object is determined according to the parameter identification result, including: determining a target query template according to the target task type, the query template being used to describe the task key element related to the task type; and screening the target task key element according to the target query template, and determining the candidate query object according to a result of the screening.

[0037] In this embodiment, the corresponding query templates for different task types can be constructed in advance by using the knowledge and experience of domain experts, and a mapping relationship between the task types and the query templates can be established, so as to ensure the accuracy and completeness of the query templates. For example, for a fault maintenance task, the query template can include a device name, a fault type, a maintenance method, and the like. After obtaining the target task type, the target query template can be determined according to the mapping relationship between the task types and the query templates. The target task key element can be screened according to the task key element described in the target query template, so as to select the task key element required by the target query template from the target task key element. The selected target task key element after the screening can be used as the candidate query object. Further, the candidate query object can be filled into the target query template, so as to perform a subsequent power knowledge graph query process according to the filled target query template.

[0038] By using the above-mentioned arrangement, the power knowledge graph is queried based on the query templates under different task types, so as to improve the accuracy and completeness of the graph query.

[0039] S130, determining a target query object associated with the candidate query object according to the pre-constructed power knowledge graph.

[0040] The power knowledge graph includes power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities. In this embodiment, the power knowledge graph needs to be constructed in advance. Optionally, the construction process of the power knowledge graph includes: obtaining target knowledge, the target knowledge including power equipment knowledge, technical operation knowledge, and field expert knowledge; determining a target triple according to the target knowledge, the triple including a first entity, a second entity, and an association relationship between the first entity and the second entity; and constructing the power knowledge graph according to the target triple.

[0041] When constructing the power knowledge graph, first, target knowledge needs to be obtained, the target knowledge including power equipment knowledge, technical operation knowledge, and field expert knowledge. The power equipment knowledge can include basic attributes of equipment, power grid topology structure, and real-time states of equipment, and the like. Specifically, the basic attributes of equipment can be obtained by information extraction from power grid equipment account information, and specifically include unique identifiers of equipment, equipment types, equipment parameters, equipment states, location information, and the like. Related data of a power grid connection diagram can be extracted from an OMS (outage management system), and specifically include connection relationships of various equipment, topology information between equipment, directions and connection points of lines, and the like. Real-time state data of equipment can be obtained from the OMS, and specifically include running states of equipment, fault information, switch states, line load conditions, and the like.

[0042] The technical operation knowledge can include safety measures, technical measures, and organizational measures related to technical specifications and operation requirements. It should be noted that the technical operation knowledge is usually stored in the form of loose in the technical specifications and operation requirements documents, and therefore, the related technologies are needed to process the documents to obtain the technical operation knowledge. For example, the professional terms and keywords in the documents can be identified by using the term recognition technology. These professional terms and keywords usually correspond to the knowledge of safety measures, technical measures, and organizational measures of power grid dispatching. The term recognition can be based on a dictionary, machine learning, or deep learning method. The key entities in the documents, such as device names, operation steps, parameter values, and times, can be identified by using the named entity recognition technology. The entity recognition can be based on a rule, machine learning, or deep learning method, and the entities in the documents are identified according to the trained model or rule matching. The relationship extraction technology is used to extract the relationship between entities from the text. The relationship extraction can use a rule-based, pattern matching, machine learning, or deep learning method. Some rules or patterns can be defined to capture the semantic relationship between entities in the documents, or a supervised learning method can be used to train a model to predict the relationship between entities. In addition, to avoid the situation that the same entity is identified as different entities due to different expressions, the extracted entities can be aligned with the existing knowledge base or domain ontology by using the entity linking technology, to ensure that the extracted entities are consistent with the entities in the existing knowledge base, thereby improving the accuracy of entity extraction. For example, the similarity between the extracted entities and the existing knowledge base or domain ontology can be calculated, and if the similarity is greater than a preset threshold, the two entities are determined to be the same entity, otherwise they are determined to be different entities.

[0043] The domain expert knowledge can be obtained by face-to-face interviews or communications with domain experts. During the interview, relevant questions can be asked to guide the experts to describe their experiences, best practices, solutions, etc. In addition, the experts can be asked to share relevant documents, reports, or other materials to further obtain domain expert knowledge. The knowledge collected from the experts is then sorted and summarized to extract key concepts, rules, and principles, etc. For example, natural language processing technology and text mining technology can be used to identify and extract important entities, relationships, and attributes. The use of domain expert knowledge can increase the richness and accuracy of the power knowledge graph, and provide more comprehensive and authoritative power dispatching knowledge. At the same time, close cooperation with domain experts can also ensure the updating and timely adjustment of knowledge to adapt to the changing needs of the power dispatching field.

[0044] After obtaining the target knowledge, further reasoning and mining can be performed on the target knowledge, and the target triplets can be determined according to the target knowledge. The triplets include a first entity, a second entity, and an association relationship between the first entity and the second entity. Specifically, the target knowledge is first abstracted and formally described using ontology modeling technology. Ontology is a structured model for defining and representing domain concepts and relationships, and forms a shared semantic model by defining classes, properties, relationships, and constraints. Common ontology modeling languages include OWL (Web Ontology Language) and RDF (Resource Description Framework). Then, the target knowledge is formally represented using knowledge representation languages such as RDF and OWL, and logical reasoning and inference are performed using reasoning techniques. Reasoning techniques can derive new conclusions from existing knowledge to discover implicit relationships and patterns in knowledge. Further, data mining and machine learning techniques can be used to analyze and mine the obtained target knowledge to discover patterns, trends, and rules from large amounts of data, which helps to understand and utilize knowledge. Common data mining and machine learning algorithms include clustering, classification, association rule mining, etc. In addition, for knowledge expressed in natural language, natural language processing techniques such as morphological analysis, syntactic analysis, semantic parsing, etc. can be used for parsing, understanding, and extracting, and natural language text can be converted into structured form that can be processed by computers.

[0045] The target knowledge obtained from each of the above knowledge sources can be integrated and fused to construct a structured power knowledge graph. The power knowledge graph represents power equipment entities, attribute information of power equipment entities, and association relationships between power equipment entities in the form of a graph, providing a rich semantic association and query capability, and enabling more comprehensive description and utilization of knowledge. The power knowledge graph can be stored in the form of triplets (Subject-Predicate-Object). Each triplet consists of a subject (Subject), predicate (Predicate), and object (Object), representing entities, relationships, and attributes. Common triplet storage systems include RDF databases (such as Virtuoso, Blazegraph) and triplet storage modules in graph databases. Graph databases are databases specifically designed for storing and processing graph data, based on graph structures, using nodes (entities) and edges (relationships) to represent data. Common graph databases include Neo4j, Apache Jena, JanusGraph, etc. These graph databases can provide efficient storage and query functions, supporting complex graph query operations.

[0046] Exemplarily, taking the basic attributes of the device as an example, the device can be taken as the Subject, various attributes can be taken as the Predicate, and the attribute values can be taken as the Object, so as to determine a target triple. Taking the power grid topology as an example, the device can be taken as the Subject, the connection relationship can be taken as the Predicate, and the connected device or line can be taken as the Object, so as to determine a target triple. Taking the real-time state of the device as an example, the device can be taken as the Subject, the state information can be taken as the Predicate, and the state value can be taken as the Object, so as to determine a target triple. After the target triple is determined, the power knowledge graph can be constructed according to the target triple.

[0047] According to the scheme, the power knowledge graph is constructed based on the power equipment knowledge, the technical operation knowledge and the field expert knowledge, which provides a rich and reliable knowledge source for the construction of the knowledge graph, and improves the richness, the professionalism and the accuracy of the power knowledge graph.

[0048] Based on the pre-constructed power knowledge graph, the target query object associated with the candidate query object can be determined. Optionally, the target query object associated with the candidate query object is determined according to the pre-constructed power knowledge graph, including: querying the target entity associated with the candidate query object and the target association relationship related to the target entity from the pre-constructed power knowledge graph; and determining the target query object associated with the candidate query object according to the target entity and the target association relationship.

[0049] Exemplarily, assuming that the candidate query object is device A, the target entity associated with device A is first queried from the pre-constructed power knowledge graph, assuming that the target entity is device B and device C, and the association relationship related to device B and device C is then respectively queried from the power knowledge graph as the target association relationship of device B and device C respectively. Thus, device B and the target association relationship thereof and device C and the target association relationship thereof can be taken as the target query object associated with device A.

[0050] It should be noted that the knowledge graph query operation can include node query, edge query, path query, etc. Among them, the node query is used to find entity nodes that meet certain conditions, the edge query is used to find relationship edges that meet certain conditions, and the path query is used to find the path between two entities. These query operations can be combined with predicates, attributes and constraint conditions for more accurate queries. In addition, the knowledge graph query operation can also use various graph algorithms to discover hidden relationships and patterns. For example, the shortest path algorithm can be used to find the shortest path between two entities, the community detection algorithm can be used to discover groups of entities with tight connections, and the PageRank algorithm can be used to determine the importance of entities. In order to more intuitively understand and explore the knowledge graph, visualization tools can be used to display the query results in a graphical form. In this way, the characteristics of entities and relationships can be represented through the layout, color, size and other visual attributes of nodes and edges, which helps to quickly understand and analyze the query results.

[0051] Through such a setting, the target query object associated with the candidate query object can be quickly, comprehensively and accurately queried based on the power knowledge graph.

[0052] In S140, a target operation ticket is generated according to the target task type and the target query object.

[0053] After determining the target query object associated with the candidate query object, a target operation ticket can be generated according to the target task type and the target query object. Optionally, generating the target operation ticket according to the target task type and the target query object includes: inputting the target task type and the target query object into a pre-trained large language model; wherein the large language model is trained based on the task type of the maintenance order, the entities and the associated relationships related to the task type in the power knowledge graph, and the operation ticket corresponding to the maintenance order generated by the human; and determining the target operation ticket according to the output result of the large language model.

[0054] In this embodiment, the target operation ticket can be generated based on a pre-trained large language model. The large language model (LLM) is a class of deep learning-based models designed to understand and generate natural language text. LLM learns the statistical rules and semantic representations of language by pre-training on a large amount of text data, and can be used for various natural language processing tasks such as text classification, named entity recognition, and text generation. The advantages of LLM are as follows:

[0055] 1、LLM can learn rich language knowledge and representation ability by pre-training on large-scale text data. This enables the model to capture the complex structure and semantic relationship of text, and thus performs well on various tasks.2、LLM can effectively utilize context information to understand the meaning of words, phrases and sentences in different contexts. By learning the context relationship in large-scale corpus, the model can more accurately resolve ambiguity and understand the semantics of the text.3、LLM has certain zero-shot learning ability, that is, it can process text tasks in new domains based on previously learned language rules and knowledge without explicit training samples. This makes LLM have certain transfer learning ability and can be applied in various fields and tasks.4、LLM has achieved remarkable success in text generation tasks. Through the learning of existing text, LLM can generate text with coherence and semantic reasonableness, including natural language dialogue, document summarization, story generation, etc.

[0056] Therefore, the application of LLM in the field of electric power is of great significance. Its generation ability can provide intelligent ticketing and operation guidance, helping to improve operation safety and efficiency; its understanding ability has the possibility of fault diagnosis and prediction, improving the reliability of the power system; supporting energy management and optimization, achieving efficient use of energy; providing safety monitoring and risk assessment to ensure the safe operation of the power system. The application of LLM can promote the intelligentization and optimization of the power industry, improve the operation efficiency and sustainable development level of the power system. However, there is no precedent for applying LLM in the field of electric power, and this scheme first applies LLM to the intelligent ticketing technology in the field of electric power.

[0057] Unlike general fields, the electric power field has extremely high requirements for safety and stability. Although LLM has made great achievements in general chat fields, it may create nonsense and words may not convey the intended meaning. This situation is not harmful in general chat fields, but it will be fatal when applied to the electric power field. Therefore, how to train a safe and reliable electric power LLM has become an urgent problem. This scheme introduces an electric power knowledge graph to enhance LLM, which can enrich domain knowledge, improve semantic understanding, make up for data gaps, enhance reasoning ability, and support specific domain professional applications. This fusion can improve the accuracy, generalization ability and application effect of the model, making it more suitable for tasks and demands in the field of electric power.

[0058] By integrating the domain knowledge in the power knowledge graph with the LLM, a more comprehensive and in-depth domain knowledge can be provided to the model, making it more accurate and comprehensive in understanding and processing texts in specific domains. At the same time, LLMs are usually pre-trained in an unsupervised or semi-supervised manner when processing natural language, and their representation capabilities are mainly based on large-scale text data. However, this representation may be limited in understanding the semantics and concepts of the power domain. The power knowledge graph provides a semantic network based on expert knowledge, which can help LLM better understand and represent domain-specific concepts and relationships. In addition, in the power domain, there is very limited data available for training LLMs, or the data quality is low. The power knowledge graph can provide additional data support to fill in the gaps or missing data, helping the model to better generalize and reason. By utilizing the structure and associations in the power knowledge graph, the model can supplement the lack of data by reasoning and referencing information in the power knowledge graph. Enhancing LLM with power knowledge graph can make the model more specialized and targeted, improving the performance of the model in power domain related tasks, and helping the power industry to improve efficiency, reduce errors, and better cope with complex power operation and management requirements.

[0059] In the process of large language model training, first of all, the data set needs to be determined, which can include positive samples, negative samples and original pre-training data (including the task type of the maintenance order and the operation ticket corresponding to the artificially generated maintenance order). Specifically, according to the entities and relationships in the power knowledge graph, the triplets with representative and important can be selected as positive samples. Consider selecting entities and relationships that cover different aspects of the power domain to ensure the diversity and coverage of the training data set. For tasks where the input and output are continuous numbers, natural language templates need to be generated first, and then the input and output are filled in the templates to make the input and output have the form of natural language. In order to increase the diversity of data and the robustness of training, negative samples can be generated in a random way. By breaking the association of true triplets in the power knowledge graph, negative samples that are not related to positive samples can be generated. In this way, the number of negative samples and positive samples can be balanced to avoid training bias towards any category. The generated positive samples and negative samples are combined with the original pre-training corpus to form a data set, ensuring that the sample distribution in the data set is uniform and not biased towards a specific domain or relationship.

[0060] An appropriate loss function needs to be selected for model training. In addition to the cross-entropy loss function, other loss functions suitable for the task can also be considered, such as the binary classification loss function, the cross-entropy loss function in sequence labeling tasks, etc. Select the appropriate loss function according to the task requirements and characteristics of the data set. Conventional optimization algorithms such as stochastic gradient descent (SGD) or adaptive optimization algorithms (such as Adam) can be used to optimize the parameters of the model. Hyperparameter tuning can also be performed on the model, where hyperparameters can include learning rate, batch size, and iteration number. Finally, select the appropriate evaluation metrics to evaluate the model. The evaluation metrics can include accuracy, recall, F1 score, etc., which can be used for classification tasks; evaluation metrics can also include precision matching rate, fuzzy matching rate, etc., which can be used for question and answer tasks.

[0061] After the large language model is trained, the target task type and target query object can be input into the large language model, and the model output result is the target operation ticket. The target operation ticket contains detailed operation steps, device information, safety precautions, etc. The large language model can automatically generate the specific content of the operation ticket according to the requirements of the task and related information, thereby ensuring the completeness and accuracy of the operation ticket.

[0062] It should be noted that the final generated target operation ticket text can have different forms, such as structured format or human-readable text form, which can be set according to actual needs. The structured operation ticket text adopts a certain data format, which can facilitate data exchange, storage and processing. This format can be a standard data format such as XML or JSON, or a custom structured format to adapt to specific operation ticket systems and application requirements. The human-readable text form of the operation ticket is presented in natural language form, which is easy to read and understand, and can be directly read and executed by the operator. The generated text can be formatted according to the pre-defined template structure and contain detailed operation steps, device information, safety precautions, etc.

[0063] The present scheme combines the power knowledge graph with the large language model, which can automatically and efficiently generate operation tickets, improving the standardization, comprehensiveness and accuracy of the operation tickets.

[0064] The technical solution of this invention involves: acquiring a target maintenance order, which includes a description of the maintenance task; identifying target parameters in the target maintenance order to obtain parameter identification results; determining candidate query objects based on the parameter identification results; wherein the target parameters include the target task type and key elements of the target task; determining target query objects associated with the candidate query objects based on a pre-constructed power knowledge graph; wherein the power knowledge graph includes power equipment entities, attribute information of power equipment entities, and relationships between power equipment entities; and generating a target operation ticket based on the target task type and the target query object. This technical solution can automatically generate high-accuracy and highly standardized operation tickets based on the power knowledge graph, improving the efficiency of operation ticket generation and reducing the cost of operation ticket generation.

[0065] In this embodiment, optionally, after obtaining the target maintenance order, the method further includes: performing data preprocessing on the target maintenance order, including data cleaning and word segmentation; correspondingly, performing target parameter identification on the target maintenance order to obtain parameter identification results, including: performing target parameter identification on the target maintenance order after data preprocessing to obtain parameter identification results.

[0066] In this embodiment, to enhance data accuracy and avoid the adverse effects of irrelevant data, data preprocessing can be performed on the target maintenance order after acquisition. Data preprocessing includes data cleaning and word segmentation. Data cleaning removes noise, special characters, and irrelevant information, ensuring text accuracy and consistency. Word segmentation breaks the text into words or phrases for subsequent task understanding and entity recognition. This converts the text into a computer-processable form. After data preprocessing of the target maintenance order, target parameter recognition can be performed to obtain parameter recognition results, thereby improving the accuracy of parameter recognition.

[0067] This solution, through such a setup, can, to some extent, avoid the adverse effects of irrelevant data by using data preprocessing, thereby enhancing data accuracy.

[0068] Example 2

[0069] Figure 2 This is a flowchart of an operation ticket generation method provided in Embodiment 2 of the present invention. This embodiment is based on the above embodiment and optimized. Specifically, the optimization is as follows: after generating the target operation ticket according to the target task type and the target query object, it further includes: determining the target supplementary information of the target task type, the target supplementary information including safety precautions; and supplementing the content of the target operation ticket according to the target supplementary information.

[0070] like Figure 2 As shown, the method in this embodiment specifically includes the following steps:

[0071] S210, obtain a target maintenance order, the target maintenance order including a maintenance task description.

[0072] S220, perform target parameter identification on the target maintenance order to obtain a parameter identification result, and determine a candidate query object according to the parameter identification result.

[0073] The target parameters include a target task type and a target task key element.

[0074] S230, determine a target query object associated with the candidate query object according to a pre-constructed power knowledge graph.

[0075] The power knowledge graph includes power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities.

[0076] S240, generate a target operation order according to the target task type and the target query object.

[0077] The specific implementation of S210-S240 can refer to the detailed description in S110-S140, which will not be repeated here.

[0078] S250, determine target supplementary information of the target task type, the target supplementary information including safety precautions.

[0079] In this embodiment, to further enhance the completeness of the target operation order, content supplement can be performed on the target operation order. For example, content can be filled in through an operation order template. Specifically, an operation order template is defined in advance for different task types, the operation order template including supplementary information

[0080] (such as safety precautions), and the safety precautions and other information required for a specific maintenance task can be filled in according to the operation order template. In addition, supplementary information can also be extracted from domain expert knowledge, and professional knowledge and experience in the power field are used to analyze relevant documents and specifications to generate safety precautions and other information suitable for a specific maintenance task as supplementary information. These information can be extracted and summarized through natural language processing technology and knowledge extraction methods.

[0081] S260, supplement the content of the target operation order according to the target supplementary information.

[0082] In this embodiment, after obtaining the target supplementary information, the content of the target operation ticket can be supplemented according to the target supplementary information. By supplementing the content of the operation ticket, the detail and standardization of the operation ticket can be ensured. The provision of the operation steps can help the operator to perform the task in the correct order, and the description of the safety precautions can remind the operator to pay attention to the potential risks and safety requirements. Therefore, the operation ticket will become a comprehensive and clear guidance document, which helps the operator to complete the maintenance task efficiently and safely.

[0083] The technical scheme of the embodiment of the present application, after generating the target operation ticket according to the target task type and the target query object, determines the target supplementary information of the target task type, and the target supplementary information includes safety precautions; and supplements the content of the target operation ticket according to the target supplementary information. Based on the automatic generation of the operation ticket with high accuracy and high standardization based on the power knowledge graph, the generation efficiency of the operation ticket is improved, and the generation cost of the operation ticket is reduced. On the basis of being able to supplement the content of the operation ticket, the detail and standardization of the operation ticket are ensured.

[0084] Embodiment three

[0085] Figure 3 The flowchart of the operation ticket generation method provided by the third embodiment of the present application is based on the optimization of the above-mentioned first embodiment. The specific optimization is: after generating the target operation ticket according to the target task type and the target query object, it further includes: auditing the target operation ticket, and adjusting the target operation ticket according to the auditing result.

[0086] As Figure 3 shown, the method of the present embodiment specifically includes the following steps:

[0087] S310, obtaining a target maintenance order, the target maintenance order including a maintenance task description.

[0088] S320, performing target parameter identification on the target maintenance order to obtain a parameter identification result, and determining a candidate query object according to the parameter identification result.

[0089] Among them, the target parameters include the target task type and the target task key elements.

[0090] S330, determining a target query object associated with the candidate query object according to a pre-constructed power knowledge graph.

[0091] Among them, the power knowledge graph includes power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities.

[0092] S340, generating a target operation ticket according to the target task type and the target query object.

[0093] The specific implementation of S310-S340 can refer to the detailed description in S110-S140, which will not be repeated here.

[0094] S350, auditing the target operation ticket, and adjusting the target operation ticket according to the auditing result.

[0095] In this embodiment, in order to further improve the accuracy of the target operation ticket, after the target operation ticket is generated, the target operation ticket can be audited, and the target operation ticket is adjusted according to the auditing result. Specifically, the content, format and safety of the operation ticket text can be audited by professional personnel to ensure its accuracy and compliance. Once the operation ticket passes the audit, it can be confirmed and put into the actual operation link. Otherwise, the operation ticket needs to be further adjusted.

[0096] The technical scheme of the embodiment of the application, after generating the target operation ticket according to the target task type and the target query object, audits the target operation ticket, and adjusts the target operation ticket according to the auditing result. This technical scheme, on the basis of automatically generating an operation ticket with high accuracy and high standardization based on the power knowledge graph, improves the generation efficiency of the operation ticket and reduces the generation cost of the operation ticket, and also ensures the accuracy and compliance of the operation ticket through the manual auditing link.

[0097] Embodiment four

[0098] Figure 4 A structural schematic diagram of an operation ticket generation device provided by the fourth embodiment of the application is provided. The device can execute the operation ticket generation method provided by any embodiment of the application, and has the corresponding function modules and beneficial effects of the execution method. As shown in the figure, the device comprises: Figure 4

[0099] The target maintenance order acquisition module 410 is configured to acquire a target maintenance order, wherein the target maintenance order comprises a maintenance task description.

[0100] The candidate query object determination module 420 is configured to perform target parameter identification on the target maintenance order to obtain a parameter identification result, and determine a candidate query object according to the parameter identification result. The target parameters include a target task type and a target task key element.

[0101] The target query object determination module 430 is configured to determine a target query object associated with the candidate query object according to a pre-constructed power knowledge graph. The power knowledge graph comprises power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities.

[0102] ​The target operation ticket generation module 440 is configured to generate a target operation ticket according to the target task type and the target query object.

[0103] Optionally, the candidate query object determination module 420 is specifically configured to:

[0104] determine a target query template according to the target task type, the query template being used to describe task key elements related to the task type;

[0105] perform screening on the target task key elements according to the target query template, and determine a candidate query object according to a result of the screening.

[0106] Optionally, the target query object determination module 430 is specifically configured to:

[0107] query a target entity associated with the candidate query object and a target association relationship related to the target entity from a pre-constructed power knowledge graph;

[0108] determine a target query object associated with the candidate query object according to the target entity and the target association relationship.

[0109] Optionally, the apparatus further includes a power knowledge graph construction module, which is specifically configured to:

[0110] obtain target knowledge, the target knowledge including power equipment knowledge, technical operation knowledge and domain expert knowledge;

[0111] determine a target triple according to the target knowledge, the triple including a first entity, a second entity and an association relationship between the first entity and the second entity;

[0112] construct a power knowledge graph according to the target triple.

[0113] Optionally, the apparatus further includes a data preprocessing module, which is specifically configured to:

[0114] perform data preprocessing on the target maintenance order after the target maintenance order is obtained, the data preprocessing including a data cleaning operation and a word segmentation operation;

[0115] Correspondingly, the candidate query object determination module 420 is further configured to:

[0116] perform target parameter recognition on the target maintenance order after the data preprocessing to obtain a parameter recognition result.

[0117] Optionally, the target operation ticket generation module 440 is specifically configured to:

[0118] input the target task type and the target query object to a pre-trained large language model; wherein the large language model is trained based on a task type of an inspection sheet, entities and associated relations related to the task type in the power knowledge graph, and an operation ticket corresponding to the artificially generated inspection sheet;

[0119] determine a target operation ticket according to an output result of the large language model.

[0120] Optionally, the apparatus further comprises a content supplementing module, specifically configured to:

[0121] after generating the target operation ticket according to the target task type and the target query object, determine target supplement information of the target task type, wherein the target supplement information comprises safety precautions;

[0122] supplement the target operation ticket according to the target supplement information.

[0123] Optionally, the apparatus further comprises an auditing and adjusting module, specifically configured to:

[0124] after generating the target operation ticket according to the target task type and the target query object, audit the target operation ticket, and adjust the target operation ticket according to an auditing result.

[0125] The operation ticket generation apparatus provided in the embodiments of the present application can execute the operation ticket generation method provided in any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of the execution method.

[0126] Embodiment five

[0127] Figure 5 A structural schematic diagram of an electronic device 10 that can be used to implement embodiments of the present application is shown. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, tablets, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices (e.g., headsets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be examples only, and are not intended to limit the implementations of the present application described and / or claimed in this document.

[0128] As shown in Figure 5 The electronic device 10 includes at least one processor 11, and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, and the like, which are in communication with the at least one processor 11.

[0129] The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes in accordance with the computer programs stored in the read-only memory (ROM) 12 or loaded from the storage unit 18 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0130] Various components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16 such as a keyboard, a mouse, and the like; an output unit 17 such as various types of displays, a speaker, and the like; a storage unit 18 such as a magnetic disk, an optical disk, and the like; and a communication unit 19 such as a network card, a modem, a wireless communication transceiver, and the like. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0131] The processor 11 can be various general and / or special purpose processing components having processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, and the like. The processor 11 performs various methods and processes described above, such as the operation ticket generation method.

[0132] The processor 11 can be various general and / or special purpose processing components having processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, and the like. The processor 11 performs various methods and processes described above, such as the operation ticket generation method.

[0133] In some embodiments, the operation ticket generation method can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the operation ticket generation method described above can be performed. Alternatively, in other embodiments, the processor 11 can be configured to perform the operation ticket generation method by any other appropriate means, such as by means of firmware.

[0134] The various embodiments of the systems and techniques described above can be implemented in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a load programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0135] Computer programs used to implement the processes of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer program, when executed, can cause instructions defined in the flow charts and / or block diagrams to be implemented. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a standalone software package and partially on a remote machine or entirely on a remote machine or server.

[0136] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store computer programs for use by or in connection with an instruction execution system, apparatus, or device. Computer-readable storage media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0137] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0138] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0139] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. A server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0140] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present disclosure. For example, the steps recited in the present disclosure can be executed in parallel, executed in sequence, or executed in a different order, as long as the desired results of the present disclosure are achieved, and the present disclosure is not limited herein.

[0141] The specific embodiments described above are not intended to be limiting, and persons skilled in the art will appreciate that various modifications, combinations, sub-combinations and alternatives can be made to the specific embodiments without departing from the spirit and principles of the disclosure. Accordingly, the disclosure is not limited to the specific embodiments described above, but only by the scope of the appended claims.

Claims

1. An operation ticket generation method characterized by comprising: The method comprises: obtaining a target maintenance order, wherein the target maintenance order comprises a maintenance task description; performing target parameter identification on the target maintenance order to obtain a parameter identification result, and determining a candidate query object according to the parameter identification result; wherein the target parameters comprise a target task type and a target task key element; determining a target query object associated with the candidate query object according to a pre-constructed power knowledge graph; wherein the power knowledge graph comprises power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities; generating a target operation ticket according to the target task type and the target query object; the construction process of the power knowledge graph comprises: obtaining target knowledge, wherein the target knowledge comprises power equipment knowledge, technical operation knowledge, and domain expert knowledge; determining a target triple according to the target knowledge, wherein the triple comprises a first entity, a second entity, and an association relationship between the first entity and the second entity; constructing a power knowledge graph according to the target triple; generating a target operation ticket according to the target task type and the target query object comprises: inputting the target task type and the target query object into a pre-trained large language model; wherein the large language model is trained based on the task type of the maintenance order, the entities and association relationships related to the task type in the power knowledge graph, and the operation ticket corresponding to the maintenance order generated by humans; determining a target operation ticket according to the output result of the large language model; after generating a target operation ticket according to the target task type and the target query object, the method further comprises: determining target supplementary information of the target task type, wherein the target supplementary information comprises safety precautions; supplementing the content of the target operation ticket according to the target supplementary information.

2. The method of claim 1, wherein, determining a candidate query object according to the parameter identification result comprises: determining a target query template according to the target task type, wherein the query template is used to describe task key elements related to the task type; screening the target task key elements according to the target query template, and determining a candidate query object according to the screened result.

3. The method of claim 2, wherein, determining a target query object associated with the candidate query object according to the pre-constructed power knowledge graph comprises: querying a target entity associated with the candidate query object and a target association relationship related to the target entity from the pre-constructed power knowledge graph; determining a target query object associated with the candidate query object according to the target entity and the target association relationship.

4. The method of claim 1, wherein, after obtaining a target maintenance order, the method further comprises: performing data preprocessing on the target maintenance order, wherein the data preprocessing comprises data cleaning and word segmentation; correspondingly, performing target parameter identification on the target maintenance order to obtain a parameter identification result comprises: performing target parameter identification on the target maintenance order after data preprocessing to obtain a parameter identification result.

5. The method according to claim 1 or 4, characterized in that, after generating a target operation ticket according to the target task type and the target query object, the method further comprises: The target operation ticket is audited, and the target operation ticket is adjusted according to an auditing result.

6. An operation ticket generating apparatus characterized by comprising: The device comprises: A target maintenance order acquisition module is configured to acquire a target maintenance order, wherein the target maintenance order comprises a maintenance task description; A candidate query object determination module is configured to perform target parameter identification on the target maintenance order to obtain a parameter identification result, and determine a candidate query object according to the parameter identification result; wherein the target parameter comprises a target task type and a target task key element; A target query object determination module is configured to determine a target query object associated with the candidate query object according to a pre-constructed power knowledge graph; wherein the power knowledge graph comprises power equipment entities, attribute information of the power equipment entities, and association relationships between the power equipment entities; A target operation ticket generation module is configured to generate a target operation ticket according to the target task type and the target query object; The device further comprises a power knowledge graph construction module, which is specifically configured to: Acquire target knowledge, wherein the target knowledge comprises power equipment knowledge, technical operation knowledge, and domain expert knowledge; Determine target triples according to the target knowledge, wherein the triples comprise a first entity, a second entity, and an association relationship between the first entity and the second entity; Construct a power knowledge graph according to the target triples; The target operation ticket generation module is specifically configured to: Input the target task type and the target query object into a pre-trained large language model; wherein the large language model is trained based on a task type of a maintenance order, entities and association relationships related to the task type in the power knowledge graph, and an operation ticket corresponding to a manually generated maintenance order; Determine a target operation ticket according to an output result of the large language model; The device further comprises a content supplement module, which is specifically configured to: After generating a target operation ticket according to the target task type and the target query object, determine target supplementary information of the target task type, wherein the target supplementary information comprises safety precautions; Supplement the content of the target operation ticket according to the target supplementary information.

7. The apparatus of claim 6, wherein, The candidate query object determination module is specifically configured to: Determine a target query template according to the target task type, wherein the query template is used to describe a task key element related to the task type; Filter the target task key element according to the target query template, and determine a candidate query object according to a filtered result.

8. The apparatus of claim 7, wherein, The target query object determination module is specifically configured to: Query a target entity associated with the candidate query object and a target association relationship related to the target entity from a pre-constructed power knowledge graph; Determine a target query object associated with the candidate query object according to the target entity and the target association relationship.

9. The apparatus of claim 6, wherein, The device further comprises a data preprocessing module, which is specifically configured to: After acquiring a target maintenance order, perform data preprocessing on the target maintenance order, wherein the data preprocessing comprises a data cleaning operation and a word segmentation operation; Correspondingly, the candidate query object determination module is further configured to: The target parameter identification result is obtained by performing target parameter identification on the target maintenance order after data preprocessing.

10. The apparatus of claim 6, wherein, The device further comprises an audit adjustment module, specifically configured to: After generating the target operation order according to the target task type and the target query object, the target operation order is audited, and the target operation order is adjusted according to the audit result.

11. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected with the at least one processor in communication; wherein The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the operation order generation method of any one of claims 1-5.

12. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing the processor to execute the operation order generation method of any one of claims 1-5.

Images