A low-voltage operation and distribution work order intelligent scheduling method, device, equipment and medium

Abstract

This invention discloses an intelligent scheduling method, device, equipment, and medium for low-voltage operation and maintenance work orders. The method includes: processing multi-source operation and maintenance business data based on the Guangming Big Data Model to construct a corresponding business knowledge base; obtaining work orders to be scheduled; acquiring search results using a dual-channel retrieval method based on the business knowledge base; enhancing the search results and inputting them into the Guangming Big Data Model to obtain scheduling decision information; constructing a multi-factor constrained scheduling model based on the scheduling decision information; and executing the scheduling model through a scheduling intelligent agent constructed based on the Guangming Big Data Model to schedule the work orders to be scheduled. This achieves efficient and accurate intelligent scheduling of low-voltage operation and maintenance work orders, improving scheduling rationality and the utilization rate of operation and maintenance resources.

Description

Technical Field

[0001] This invention relates to the field of low-voltage operation and maintenance technology, and in particular to a method, device, equipment and medium for intelligent scheduling of low-voltage operation and maintenance work orders. Background Technology

[0002] Low-voltage operation and maintenance work orders are the core carrier for power companies to carry out low-voltage power distribution and operation and maintenance services. The rationality of their scheduling directly affects operation and maintenance efficiency, service quality and resource utilization, and is a key link in the efficient development of low-voltage operation and maintenance business.

[0003] Currently, low-voltage operation and maintenance work order scheduling mostly adopts manual scheduling or simple rule-based scheduling schemes. Manual scheduling mainly relies on the experience of operation and maintenance personnel, and manually allocates operators and work time based on basic information such as work order priority and work location. Simple rule-based scheduling mechanically sorts and allocates work orders through preset fixed rules. Some schemes will introduce basic data statistics functions to assist in scheduling decisions.

[0004] The aforementioned existing technical solutions have obvious drawbacks: manual scheduling is inefficient, easily affected by differences in personnel experience, and the scheduling results are highly subjective, making it difficult to achieve multi-factor collaborative optimization; simple rule-based scheduling lacks in-depth utilization of multi-source operation and maintenance business data, cannot accurately match work order requirements with operation and maintenance resources, lacks scientific and rational decision-making, and cannot form a closed loop of data processing and scheduling execution, resulting in a disconnect between scheduling results and actual operation and maintenance scenarios, making it difficult to meet the large-scale and refined operation and maintenance needs of low-voltage operation and maintenance business. Summary of the Invention

[0005] This invention provides a method, device, equipment, and medium for intelligent scheduling of low-voltage operation and maintenance work orders, so as to achieve efficient and accurate intelligent scheduling of low-voltage operation and maintenance work orders, improve the rationality of scheduling and the utilization rate of operation and maintenance resources, and solve the problems of low efficiency and unscientific decision-making in existing scheduling schemes.

[0006] According to one aspect of the present invention, a method for intelligent scheduling of low-voltage operation and maintenance work orders is provided, comprising: Based on the Guangming Big Data Model, multi-source marketing and distribution business data are processed to build a corresponding business knowledge base; Obtain the work orders to be scheduled, retrieve the search results using a dual-channel retrieval method based on the business knowledge base, enhance the search results and input them into the Guangming Big Data Model to obtain scheduling decision information; A multi-factor constrained scheduling model is constructed based on the scheduling decision information. The scheduling model is executed by a scheduling agent constructed based on the Guangming Big Data Model to schedule the work orders to be scheduled.

[0007] According to another aspect of the present invention, a low-voltage operational work order intelligent scheduling device is provided, comprising: The knowledge base construction module is used to process multi-source operation and distribution business data based on the Guangming Big Data Model and build the corresponding business knowledge base. The retrieval enhancement module is used to obtain work orders to be scheduled, obtain retrieval results based on the business knowledge base using a dual-channel retrieval method, enhance the retrieval results and input them into the Guangming Big Data Model to obtain scheduling decision information; The scheduling execution module is used to construct a multi-factor constrained scheduling model based on the scheduling decision information, and execute the scheduling model through a scheduling agent constructed based on the Guangming big data model to schedule the work orders to be scheduled.

[0008] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and memory that is communicatively connected to at least one processor; The memory stores a computer program that can be executed by at least one processor, which is then executed by the at least one processor to enable the at least one processor to execute the low-voltage distribution work order intelligent scheduling method of any embodiment of the present invention.

[0009] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the intelligent scheduling method for low-voltage operation and distribution work orders according to any embodiment of the present invention.

[0010] The technical solution of this invention constructs a business knowledge base by processing multi-source operation and maintenance business data based on the Guangming Big Data Model, obtains scheduling decision information by combining dual-channel retrieval with the Guangming Big Data Model, and then executes the multi-factor constraint scheduling model to schedule the work orders to be scheduled through a scheduling intelligent agent. This solves the problems of low efficiency, unscientific decision-making, and inability to form a closed loop in existing scheduling schemes, and realizes accurate and efficient scheduling of low-voltage operation and maintenance work orders, improves the utilization rate of operation and maintenance resources and service quality, and adapts to the refined operation and maintenance needs of low-voltage operation and maintenance business.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 A flowchart illustrating an intelligent scheduling method for low-voltage operation and maintenance work orders provided in an embodiment of the present invention; Figure 2 This is an architecture diagram of a low-voltage distribution work order intelligent scheduling system provided in an embodiment of the present invention; Figure 3 Flowchart for the implementation plan of multi-source heterogeneous data fusion and knowledge transformation; Figure 4 Flowchart of the implementation plan for enhancing knowledge retrieval for work order scheduling; Figure 5 Flowchart of the solution for constructing a multi-factor work order scheduling intelligent agent; Figure 6 Design a schematic diagram for the implementation of an intelligent agent; Figure 7 This is a schematic diagram of the structure of a low-voltage distribution work order intelligent scheduling device provided in an embodiment of the present invention; Figure 8 A schematic diagram of the structure of an electronic device for implementing the intelligent scheduling method for low-voltage operation and maintenance work orders in this embodiment of the invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] Figure 1This is a flowchart illustrating an intelligent scheduling method for low-voltage operation and maintenance work orders provided in an embodiment of the present invention. This embodiment is applicable to intelligent scheduling scenarios for various maintenance work orders in the low-voltage operation and maintenance field. The method can be executed by an intelligent scheduling device for low-voltage operation and maintenance work orders, which can be implemented in hardware and / or software and can be configured in an electronic device. Figure 1 As shown, the method specifically includes the following steps: S110. Based on the Guangming Big Data Model, process multi-source operation and distribution business data and construct a corresponding business knowledge base.

[0017] Multi-source operation and maintenance business data comprises various related data generated during the low-voltage operation and maintenance business. This data can cover maintenance equipment data, historical work order data, operator data, and other work order scheduling-related data. The Guangming Big Data Model is an artificial intelligence model with powerful semantic understanding and knowledge extraction capabilities. It can perform in-depth processing on various operation and maintenance business data to extract valuable information. The business knowledge base is a structured storage medium for storing operation and maintenance business-related knowledge, providing knowledge support for subsequent work order scheduling.

[0018] Specifically, various types of multi-source marketing and distribution business data can be collected, the collected data can be organized and adapted, and then the processed multi-source marketing and distribution business data can be input into the Guangming Big Data Model. The Guangming Big Data Model can then extract knowledge and perform semantic transformation on the data, transforming unstructured and semi-structured data into structured knowledge.

[0019] Furthermore, a corresponding business knowledge base is built based on structured knowledge to complete the storage and organization of knowledge, providing a knowledge foundation for scheduling retrieval and effectively solving the problems of lack of knowledge support and unscientific decision-making in existing scheduling schemes.

[0020] S120. Obtain the work orders to be scheduled, and use a dual-channel retrieval method based on the business knowledge base to obtain the retrieval results. Enhance the retrieval results and input them into the Guangming Big Data Model to obtain scheduling decision information.

[0021] The work orders to be scheduled can be low-voltage operation and maintenance work orders that require job allocation and time arrangement. These work orders can include basic information such as work order type, job requirements, and job location. The dual-channel retrieval method combines two different retrieval paths, enabling comprehensive and accurate retrieval of relevant knowledge in the business knowledge base. Scheduling decision information serves as the basis for guiding work order scheduling decisions, supporting the subsequent construction and execution of the scheduling model.

[0022] Specifically, all work orders awaiting scheduling are retrieved from the operations and maintenance system. Basic information on these work orders is then compiled. Based on this basic information, a dual-channel retrieval method is used to search the business knowledge base, obtaining search results relevant to the work orders awaiting scheduling. These results are then enhanced through reordering and filtering to remove irrelevant and redundant information, improving the relevance and accuracy of the search results. Finally, the enhanced search results are input into the Guangming Big Data Model. Through the model's reasoning and analysis capabilities, scheduling decision information that meets the needs of the work orders awaiting scheduling is generated.

[0023] S130. Construct a multi-factor constrained scheduling model based on the scheduling decision information, and execute the scheduling model through a scheduling agent constructed based on the Guangming big data model to schedule the work orders to be scheduled.

[0024] Among them, the multi-factor constraint scheduling model is a model constructed by considering various constraints that affect the rationality of scheduling. This model can achieve multi-dimensional optimization of scheduling results. The scheduling agent is an intelligent unit with autonomous decision-making and execution capabilities built based on the Guangming big model. The scheduling agent can parse the scheduling model and execute scheduling operations.

[0025] Specifically, based on the obtained scheduling decision information and combined with the actual needs of low-voltage operation and maintenance work order scheduling, various constraints affecting scheduling are identified, a multi-factor constraint scheduling model is constructed, the optimization goals and execution logic of scheduling are clarified, and then a scheduling intelligent agent built based on the Guangming big data model is used to analyze the constraints and execution logic of the scheduling model. Combined with the specific situation of the work orders to be scheduled, the scheduling model is executed to complete the allocation of operators, scheduling of work time, and planning of work paths for the work orders to be scheduled, thereby realizing intelligent scheduling of the work orders to be scheduled, improving scheduling efficiency and rationality, and forming a closed loop from data processing to scheduling execution.

[0026] In some possible implementations, the process of processing multi-source operation and distribution business data based on the Guangming Big Data Model to construct a corresponding business knowledge base includes: unified access and standardization of the multi-source operation and distribution business data; knowledge extraction and semantic transformation of the processed multi-source operation and distribution business data through the Guangming Big Data Model to form structured knowledge; and construction and storage of the corresponding business knowledge base based on the structured knowledge.

[0027] Among these, unified access is a centralized method for acquiring multi-source operational and distribution business data, which can shield the access differences between different data sources. Normalization is the process of formatting data, improving data consistency and usability. Knowledge extraction is the process of extracting effective business information from data, uncovering key content within the data. Semantic transformation is the step of converting information into a unified semantic expression, facilitating model understanding and subsequent use. Structured knowledge is knowledge organized according to a fixed organizational form, improving the efficiency of knowledge retrieval.

[0028] In this embodiment of the invention, unified access processing can be performed on multi-source operation and distribution business data, followed by data cleaning, format normalization and other standardization operations. Then, information extraction and semantic normalization are performed on the processed data through the Guangming Big Data Model to form well-organized structured knowledge. Finally, the business knowledge base is constructed and stored based on this structured knowledge.

[0029] In some possible implementations, the construction and storage of the corresponding business knowledge base based on the structured knowledge includes: performing metadata scanning of the service catalog interface through the application framework; combining the system interface documentation with authentication methods, interface examples, and parameter requirements; and transforming the fused information into knowledge content that the Guangming Big Data Model can understand in order to construct and store the business knowledge base.

[0030] The application framework is the software framework that supports the operation of the business system, providing the basic environment for interface scanning and invocation. Metadata scanning is the automatic collection of interface information, obtaining basic configuration information of the interface. System interface documentation is a document recording the usage specifications of the interface, providing a reference for understanding the interface. Authentication methods are the verification methods required to access the interface, ensuring the security of interface calls.

[0031] Specifically, the application framework can automatically scan relevant information of service catalog interfaces, and then combine it with system interface documentation to integrate authentication methods, interface examples, parameter requirements, and other content. The integrated information is then converted into a knowledge form that the Guangming Big Data Model can recognize, completing the construction and storage of the business knowledge base. This enables automated sorting and knowledge-based expression of interface information, improving the automation and accuracy of knowledge base construction.

[0032] In some possible implementations, obtaining search results based on the business knowledge base using a dual-channel retrieval method includes: performing a first-channel retrieval based on keyword matching; performing a second-channel retrieval based on semantic vector matching; and determining the search results based on the results of the first-channel retrieval and the second-channel retrieval.

[0033] Keyword matching is a retrieval method based on precise matching of text content, which can quickly locate information containing specified content. Semantic vector matching is a retrieval method based on semantic similarity, which can uncover related information with similar meanings. The first-path retrieval and the second-path retrieval are two parallel retrieval paths, which can complete knowledge retrieval from different dimensions.

[0034] Specifically, the business knowledge base is searched using both keyword matching and semantic vector matching. The results from both methods are then combined to determine the final search results. This approach balances the accuracy and comprehensiveness of the search, avoids information omissions caused by relying on a single search method, and improves the completeness of the search results.

[0035] In some possible implementations, enhancing the search results and inputting them into the Guangming Big Data Model to obtain scheduling decision information includes: reordering and filtering the search results based on the context information of the work order to be scheduled; inputting the processed search results as context into the Guangming Big Data Model; and performing reasoning analysis through the Guangming Big Data Model to obtain the corresponding scheduling decision information.

[0036] Among these, contextual information refers to the scenario information related to the work order to be scheduled, reflecting the actual scheduling environment of the work order. Re-ranking is the operation of rearranging search results according to relevance, which can improve the priority of effective information. Filtering is the step of filtering search results, removing redundant or irrelevant content. Reasoning analysis is the process by which the Guangming Big Data Model makes logical judgments based on information, generating reasonable decision content.

[0037] Specifically, the context information of the work orders to be scheduled is used to reorder and filter the search results. The filtered results are then used as reference information and input into the Guangming Big Data Model. The Guangming Big Data Model then performs logical judgment and analysis to obtain the corresponding scheduling decision information. This can improve the quality of the input information and enhance the rationality and relevance of the scheduling decision information.

[0038] In some possible implementations, the construction of a multi-factor constraint scheduling model based on the scheduling decision information includes: determining at least one constraint factor among work order attributes, job location, personnel capabilities, and job time window; and constructing the corresponding multi-factor constraint scheduling model based on the constraint factor and the scheduling decision information.

[0039] The work order attributes are the unique characteristics of a work order, reflecting its type and urgency. The work location is the specific location of the work, influencing maintenance paths and resource allocation. Personnel capabilities are the skills and qualifications of the personnel performing the work, determining the appropriate personnel for the work order. The work time window is the permitted time period for executing the work, regulating when the work should be performed.

[0040] Specifically, at least one type of constraint factor can be identified from work order attributes, work location, personnel capabilities, and work time window. By combining scheduling decision information with the constraint factors, a multi-factor constraint scheduling model can be constructed, which can make the scheduling model fit the actual business scenario and improve the rationality and executability of the scheduling results.

[0041] In some possible implementations, the step of executing the scheduling model through a scheduling agent built on the Guangming big model to schedule the work orders to be scheduled includes: parsing the constraints of the scheduling model through the scheduling agent; sorting, allocating time, and planning paths for the work orders to be scheduled based on the parsing results, and obtaining the scheduling results.

[0042] Among these, constraints are the conditions that limit the scheduling rules in the scheduling model, ensuring that the scheduling results meet business requirements. Sorting is the operation of arranging work orders in the order of execution, determining their processing priority. Time allocation is the step of assigning task time to work orders, clarifying their execution periods. Path planning is the process of arranging task routes, improving the efficiency of maintenance operations.

[0043] Specifically, by analyzing the constraints in the scheduling model through a scheduling agent, the execution order, time arrangement, and path planning of the work orders to be scheduled are arranged according to the analysis results, resulting in standardized scheduling results. This enables the automated execution of the scheduling process and improves the efficiency and standardization of work order scheduling.

[0044] The technical solution of this invention constructs a business knowledge base by processing multi-source operation and maintenance business data based on the Guangming Big Data Model, obtains scheduling decision information by combining dual-channel retrieval with the Guangming Big Data Model, and then executes the multi-factor constraint scheduling model to schedule the work orders to be scheduled through a scheduling intelligent agent. This solves the problems of low efficiency, unscientific decision-making, and inability to form a closed loop in existing scheduling schemes, and realizes accurate and efficient scheduling of low-voltage operation and maintenance work orders, improves the utilization rate of operation and maintenance resources and service quality, and adapts to the refined operation and maintenance needs of low-voltage operation and maintenance business.

[0045] Figure 2This is an architectural diagram of a low-voltage operational and distribution work order intelligent scheduling system provided in an embodiment of the present invention. This embodiment is a preferred embodiment of the above embodiments, and its specific implementation can be found in the technical solution of this embodiment. Technical terms that are the same as or corresponding to those in the above embodiments will not be repeated here.

[0046] 3. Detailed technical solution description: 3.1. Definitions Work order scheduling: In power operation and maintenance scenarios, it is the process of rationally arranging the order of operations and resource allocation based on multiple factors such as work order content, personnel, equipment, and environmental conditions, with the goal of improving work efficiency and shortening response time.

[0047] Guangming Power's Large Model: A large language model trained specifically for the business characteristics of the power industry. It possesses semantic understanding, reasoning, knowledge question answering, and natural language generation capabilities, and can be applied to business scenarios such as load forecasting, power grid operation planning, and work order scheduling.

[0048] Multi-source heterogeneous data: Data collections from different sources and in different formats, such as structured (database records), semi-structured (Excel, logs), unstructured (text, PDF, images), and external real-time data (meteorology, GIS).

[0049] Retrieval-Augmented Generation (RAG): Retrieve relevant information from the knowledge base first, and then combine it with a large model to generate results, thereby reducing model illusion and improving professionalism.

[0050] Intelligent Agent: An autonomous computing unit based on a large model, possessing the capabilities of "perception, reasoning, decision-making, and execution," and capable of collaboratively completing complex tasks.

[0051] Agent Orchestration: Organizing and scheduling multiple agents to work together according to a process.

[0052] Service catalog: A structured list that provides a unified description, classification and management of various services provided by the system or platform. It typically includes information such as service name, function description, invocation method, input and output parameters and usage constraints, and is used to support the unified discovery, invocation and governance of services.

[0053] Prompt engineering: A series of methods and techniques for designing, organizing, and optimizing input prompts to guide generative artificial intelligence systems, such as large language models, to produce high-quality, controllable output results according to expected goals.

[0054] 3.2. Technical Solution This invention proposes an intelligent scheduling method for low-voltage operation and maintenance work orders based on the Guangming big data model. It proposes solutions for multi-source heterogeneous data fusion and knowledge transformation, knowledge retrieval enhancement for low-voltage operation and maintenance work order scheduling, and construction of multi-factor work order scheduling intelligent agents.

[0055] 3.2.1. Multi-source heterogeneous data fusion and knowledge transformation To address the challenges of fragmented data sources, diverse data types, and inconsistent semantics in low-voltage operational work order scheduling scenarios, which hinders effective understanding and utilization by large models, this invention proposes a multi-source heterogeneous data fusion and knowledge transformation method. This method utilizes a unified data access and semantic processing approach to transform data from different systems and formats into structured knowledge that can be accessed by large models and intelligent agents, providing fundamental support for subsequent knowledge retrieval enhancement and intelligent decision-making in work order scheduling.

[0056] This method is implemented using the following technical approach: 1. Unified access and standardized processing of multi-source heterogeneous data. A unified data access mechanism is established to collect structured data from business systems such as work order systems, equipment management systems, and customer systems, as well as semi-structured and unstructured data from document systems, report files, and API return results. Through data cleaning, deduplication, format conversion, and semantic standardization, differences in field naming, data granularity, and expression methods among different data sources are eliminated, forming a unified data representation.

[0057] 2. Knowledge Extraction and Semantic Transformation Based on Large Models. Utilizing the semantic understanding capabilities of Guangming Power's large model, entity extraction, attribute extraction, and relationship identification are performed on the standardized data. The work order information, personnel information, equipment information, and their relationships in the original data are transformed into structured knowledge units. Through vectorization representation or knowledge relationship modeling, the semantic expression of knowledge content is realized, improving subsequent retrieval and reasoning capabilities.

[0058] 3. Service Capability Knowledge-Based Modeling Method. Abstract models are created for the interfaces provided by business systems such as the work order system, resource management system, map service, and weather service, constructing a unified service catalog. By providing structured descriptions of interface functions, input / output parameters, calling constraints, and authentication methods, the service capabilities are expressed in a knowledge-based manner, enabling the large model to understand and correctly invoke relevant services during the scheduling process.

[0059] 4. Knowledge Dynamic Update and Consistency Maintenance Mechanism. Establish a data change monitoring and triggering mechanism. When the content of the data source changes, incremental updates are performed on the affected knowledge content according to predetermined knowledge extraction and transformation rules. Through knowledge version management and update record mechanisms, the timeliness and consistency of knowledge content in the knowledge base are ensured.

[0060] Through the above technical approach, this invention realizes the automatic transformation of multi-source heterogeneous data related to low-voltage operation and maintenance work order scheduling into a unified knowledge representation, solves the problem that business data is difficult to be effectively utilized by large models, and provides a reliable knowledge foundation for subsequent scheduling decisions and multi-agent collaboration based on retrieval enhancement generation.

[0061] 3.2.2. Enhanced Knowledge Retrieval for Work Order Scheduling To address the issue that scheduling decisions in low-voltage operational work order scheduling scenarios rely simultaneously on business rules, historical experience, and real-time environmental information, and that general-purpose large models are prone to generating unreliable results and producing inconsistencies when lacking business constraints, this invention proposes a knowledge retrieval enhancement method for work order scheduling. This method employs a fusion of knowledge representation optimization and multi-strategy retrieval, providing highly relevant knowledge support to the large model before it generates scheduling solutions, thereby improving the accuracy and reliability of scheduling decisions.

[0062] This method is implemented using the following technical approach: 1. Multi-strategy retrieval mechanism based on content features. Different retrieval methods and weights are set according to the characteristics of the knowledge content to achieve integrated ranking of multi-strategy retrieval results. For example: full-text search is used for keyword-sensitive scenarios such as technical documents, professional terminology, and entity naming; vector retrieval is used for content described in natural language, balancing comprehensiveness and accuracy of recall; user behavior feedback (hit rate, bounce rate, correction records, etc.) is introduced to optimize retrieval ranking, and the recall quality is iteratively improved through the RAG feedback learning mechanism; personalized and scenario-based knowledge retrieval is achieved based on contextual information such as user-input work order type, time, geographical location, and historical records.

[0063] 2. Enhanced search results mechanism based on scheduling context. During the search process, scheduling context information such as work order type, geographical location, time window, and personnel information is incorporated to filter and reorder search results, ensuring that the retrieved knowledge content remains consistent with the current scheduling scenario and improving the targeting of knowledge retrieval.

[0064] 3. Retrieval-enhanced scheduling decision support. The merged and ranked retrieval results are input into the Guangming Power large model as contextual information. The retrieval-enhanced generation method guides the large model to reason based on the retrieved business knowledge and rules during the scheduling decision process, avoiding the generation of scheduling results that deviate from business constraints.

[0065] Through the above technical approach, this invention implements a "retrieve first, generate later" knowledge enhancement mechanism in the work order scheduling decision-making process, which effectively reduces the randomness and illusion risk of the generated results of large models, improves the applicability of scheduling results under business rules, historical experience and actual constraints, and provides reliable knowledge support for the construction of subsequent multi-factor work order scheduling intelligent agents.

[0066] 3.2.3. Construction of a Multi-Factor Work Order Scheduling Intelligent Agent This invention proposes a method for constructing a low-voltage operation and maintenance work order scheduling intelligent agent. It uses the Guangming Power large model as the core reasoning engine, integrates technical capabilities such as knowledge base and service catalog, and introduces mechanisms such as prompting engineering, semantic understanding, and intelligent agent orchestration to construct a multi-factor work order scheduling intelligent agent with closed-loop capabilities of understanding-reasoning-decision-execution.

[0067] 1. Scheduling Modeling and Solution under Multi-Factor Constraints. This study analyzes scheduling tasks in various operational scenarios, including emergency repair orders, fault handling orders, and new installation application orders, identifying key scheduling factors. A multi-factor model influencing scheduling is constructed, covering elements such as geographical location, order type and priority, work time window, weather conditions, site conditions, operator skill profiles, and equipment configuration. Scheduling objectives are defined, including prioritizing urgency, shortest completion time, maximizing resource utilization, ensuring work balance, and minimizing service response. A heuristic scheduling method guided by a large model is researched, combined with RAG enhancement, to optimize order orchestration under complex constraints.

[0068] 2. Construction of a work order scheduling intelligent agent based on the Guangming Big Data Model. A scheduling agent with capabilities such as semantic understanding, task planning, and service invocation is constructed to autonomously acquire scheduling knowledge, analyze work order content, and formulate scheduling strategies. Prompt word templates are designed to guide the Guangming Big Data Model in understanding scheduling intent, invoking knowledge, and generating scheduling plans.

[0069] 4. Advantages of this invention: This invention has the following advantages: 1. Achieve efficient integration and dynamic updates of multi-source heterogeneous data. Through unified access, extraction, and transformation methods, structured processing of multi-source heterogeneous data such as work orders, equipment, users, geographical locations, and weather is achieved, enabling cross-system data integration; combined with change awareness and incremental update mechanisms, timely updates to the knowledge base are ensured.

[0070] 2. Improve the accuracy and professionalism of scheduling results. Based on optimized knowledge representation methods and a hybrid retrieval mechanism of "full-text search + vector search", combined with RAG technology, the large model illusion problem is effectively reduced, ensuring that the scheduling results are more in line with the logic of power business.

[0071] 3. Support dynamic optimization scheduling under multiple constraints. Construct a scheduling modeling method that integrates factors such as geographical location, personnel skills, work order type, weather, and work time window. Utilize large model inference and tool calling capabilities to achieve dynamic priority adjustment and multi-objective optimization, thereby improving the scientific nature and flexibility of scheduling.

[0072] 4. Forms an integrated closed loop of knowledge, model, and service with good scalability. Through a service catalog mechanism, it connects to various business system interfaces, supports structured service calls by intelligent agents, and forms a closed-loop application of "understanding—reasoning—decision-execution." This is not only applicable to work order scheduling but can also be extended to scenarios such as fault diagnosis and operation monitoring.

[0073] 5. Detailed implementation methods and accompanying drawings: To better illustrate the present invention, the specific embodiments of the present invention will be described in detail below in conjunction with the implementation route.

[0074] 5.1. Overall Architecture The intelligent scheduling method for low-voltage operation and maintenance work orders proposed in this invention, based on the Guangming Big Data Model, consists of three parts: a data access layer, a knowledge construction and retrieval layer, and an intelligent decision-making layer, forming a complete technical route of "data fusion—knowledge enhancement—intelligent scheduling." For details, please refer to... Figure 2 .

[0075] The data access layer is responsible for the unified access of multi-source heterogeneous data, including multi-source data access processing and API interface adaptation.

[0076] The multi-source data access, processing, acquisition, intervention, and scheduling involve structured and unstructured data from multiple systems. By leveraging the capabilities of the Guangming Big Data Model with few or even zero samples, data cleaning and semantic standardization can be performed without relying on large-scale data annotation.

[0077] API interface adaptation enriches the semantic content of interface data scanned from multiple systems, combined with interface description documents, and uses it as a corpus for knowledge base construction.

[0078] The knowledge construction and retrieval layer includes knowledge base construction and retrieval, as well as the dynamic updating of knowledge content.

[0079] Knowledge extraction utilizes the Guangming Big Data Model to extract information such as work order information and emergency repair information, obtaining entity information, relationship information, and event information.

[0080] Knowledge modeling utilizes the Guangming Big Data Model to analyze work order scheduling factors, enabling ontology recognition, multi-system knowledge alignment, knowledge generation, rule generation, and reasoning.

[0081] Service catalog construction involves scanning service catalog interface metadata through the application framework, and combining it with system interface documentation to integrate authentication methods, interface examples, parameter requirements, and other content to build a knowledge base that the large model can understand.

[0082] Knowledge is dynamically updated by detecting changes in data across various systems through scheduled tasks or message notifications, and then synchronizing the updated content to the knowledge base.

[0083] Enhanced search capabilities allow for different search methods based on the knowledge base content. Full-text search is used for keyword-sensitive scenarios such as searching technical documents, standards, and technical terms; vector search is used for natural language, similar questions, and abstract questions. Alternatively, a hybrid search combining both methods can be used, with different result priority settings for each search method.

[0084] RAG enhancement is used to provide retrieval information to a large model, and then provide material for the generated content, thereby enhancing the richness and accuracy of the generated content.

[0085] User feedback has been optimized, allowing users to annotate and modify answers, providing a basis for optimizing subsequent answers.

[0086] The intelligent decision-making layer utilizes the reasoning capabilities of the Guangming big model, combined with prompt word engineering and Agent components, to construct multiple intelligent agents with clearly defined roles and collaborative work, forming a closed-loop mechanism of "perception-reasoning-decision-execution" to achieve intelligent scheduling of low-voltage operation and distribution work orders.

[0087] The data verification agent enables automatic acquisition of work order information and verification of information such as work order information, employee information, and scheduling strategies.

[0088] The factor assessment agent, based on retrieval-enhanced contextual input, quantifies the five major factors through large-scale model inference. It evaluates the scores of each factor according to work order information, scheduling strategies, and reference information retrieved from the knowledge base.

[0089] The scheduling decision agent references external data, such as map route information and weather information, and comprehensively considers factors such as work order duration, work order type, and employee preferences to make intelligent scheduling decisions based on the Guangming big data model.

[0090] 5.2. Implementation Plan 5.2.1. Multi-source heterogeneous data fusion and knowledge transformation Figure 3 A flowchart for implementing multi-source heterogeneous data fusion and knowledge transformation.

[0091] Step S1: Multi-source data acquisition. The system acquires raw data from power business systems, external data sources, document systems, etc., through various modules in the data access layer. Examples of data structures for different systems: Key information in work order data: work order type, work order address, username, and user level. Example of employee proficiency information: Employee XX completed 100 long-term work orders and 150 short-term work orders in July. Example of employee habits: Employee XX usually eats lunch out; Employee XX usually eats lunch at the office. Sensitive customer example information: XX electricity user level is 1; XX electricity user level is 2; Weather information: Province, city, and district address information, latitude and longitude information, temperature information, weather type, and rainfall level. Step S2: Data Preprocessing. Based on the Guangming big data model and combined with the prompt word engineering, the collected data is cleaned, deduplicated, and standardized in format. An example is shown below: #Character Commands You are the key information extractor / generator, and you need to extract and process information according to the specified requirements. #Processing Guidelines 1. Field extraction priority: orderAddress / deviceName>orderContent>orderTypeShortName (Core verb) - If orderAddress does not exist, extract the value of deviceName, in the order of deviceName > orderContent > orderTypeShortName. - You only need to extract the contents of the three fields orderAddress, orderContent, and orderTypeShortName and generate key information within 15 characters. The generated information should be placed in the content field. If orderAddress does not exist, extract the contents of the deviceName field.

[0092] Input example: { "orderId":"9987074", "orderTypeShortName":"rush repair", "orderAddress":"Traffic light at the intersection of Guoxiao Road and Zhengyi Road, Yangpu District, xx City", "sensitiveConsumer":null, "importantConsumerLevel":null, "orderContent": "

One Household Without Power

[0093] 3. Technical terminology: Retain professional terminology ("tower tilt" is preferred over "equipment tilt") #Data to be processed: {{#context#}} Output requirements: 1. Preserve the original JSON structure 2. Only the key content extracted from the orderAddress, orderContent, and orderTypeShortName fields is concatenated and placed in the content field of the same object. 3. Chinese characters ≤ 15 4. Validate the number of characters for line breaks before output. 5. It is not necessary to convert the keys of the key-value pairs in the fields to Chinese; they should be kept as is. The output object only needs an additional `content` field to hold the extracted key information. 6. For pole / tower tilt issues, only extract the pole / tower tilt. In similar cases, please flexibly remove duplicate content. 7. The output only shows the result and does not include the thought process. Please generate a JSON object that conforms to the following specifications: the generated JSON object should remove the backslashes "json", and the JSON data should not contain escape characters such as / n. Step S3: Knowledge Extraction. Based on the Guangming Big Data Model and combined with prompt word engineering, structured knowledge is extracted from the preprocessed data. An example is shown below: def generate_single_entry(text:str)->Dict: prompt=f""" Character settings You are a knowledge extraction assistant, responsible for extracting structured knowledge from text. Please strictly follow the output requirements and do not add any extra explanations.

[0094] Mission Objectives The input text is analyzed to extract entities, attributes, relationships, events, and other content, and output in a unified JSON format to ensure that it is structured and parsable.

[0095] Extraction requirements 1. Entity extraction: Identify key entities in the text (such as names, locations, organizations, equipment, time, etc.).

[0096] 2. Attribute extraction: Supplement the entity with relevant attributes (such as quantity, parameters, status).

[0097] 3. Relation extraction: Identify semantic relationships between entities (such as "belongs to", "connects", "acts on").

[0098] 4. Event Extraction: Extract events related to time and behavior (such as "equipment failure", "task assignment", "accident occurrence").

[0099] 5. Output requirements: The result must be in JSON format, with fixed fields; do not omit any.

[0100] Output format (example) json { "entities": [ {"id": "E1", "name": "Transformer", "type": "Equipment"}, {"id": "E2", "name": "September 12, 2023", "type": "Time"} ], "attributes": [ {"entity_id": "E1", "key": "Capacity", "value": "100kVA"} , "relations": {"subject": "E1", "relation": "Failed", "object": "E2"} , "events": {"id": "EV1", "trigger": "Power outage", "time": "September 12, 2023", "participants": ["E1"]} } Text to be extracted: {text} Ensure that all generated content is directly related to the given text, is in a valid JSON format, and the content is high-quality, accurate, and detailed.

[0101] try: response = client.chat( model="Guangming 32b-chat", messages=[{"role": "user", "content": prompt}], # temperature=0.7,# Increase the temperature to improve diversity # max_tokens=4098 ) logger.info(f"API response: {response['message']['content']}") # Omit some response parsing code except Exception as e: logger.error(f"Error occurred while generating entries: {str(e)}") raise def generate_dataset(folder_path: str, entries_per_file: int = 2) ->List[Dict]: dataset = [] ​for filename in os.listdir(folder_path): if filename.endswith(".txt"): file_path = os.path.join(folder_path, filename) logger.info(f"Processing file: {filename}") text = read_file(file_path) for j in range(entries_per_file): logger.info(f" Generate the {j+1} / {entries_per_file}th entry") entry = generate_single_entry(text) if entry and all(key in entry for key in ['instruction', 'input', 'output', 'text']): dataset.append(entry) logger.info(f"Successfully generated 1 complete entry") else: logger.warning(f"Skip incomplete entries") time.sleep(2) # Adds a 2-second delay between requests. return dataset Step S4: Knowledge Storage. By calling the Dify knowledge base API, the processed knowledge is stored in the knowledge base, typically using a Weaviate vector database to establish an efficient index structure.

[0102] Create Knowledge Base API: curl --location --request POST 'http: / / ip:port / v1 / datasets' \ --header 'Authorization: Bearer {api_key}' \ --header 'Content-Type: application / json' \ --data-raw '{"name": "dataset_name", "permission": "all_team_members"}' Creating documents from text: curl--location--request POST http: / / ip:port / v1 / datasets / {dataset_id} / document / create-by-text' \ --header 'Authorization: Bearer {api_key}' \ --header 'Content-Type: application / json' \ --data-raw '{"name": "text","text": "text content","indexing_technique": "high_quality","process_rule": {"mode": "automatic"}} Step S5: Dynamic Updates. Monitor data source changes via scheduled tasks or message notifications, generate knowledge shards as per the previous steps, and call the Dify knowledge base API to achieve incremental updates and version management of knowledge.

[0103] Update document paragraphs: curl --location --request POST 'http: / / ip:port / v1 / datasets / {dataset_id} / documents / {document_id} / segments / {segment_id}' \ --header 'Authorization: Bearer {api_key}' \ --header 'Content-Type: application / json'\ --data-raw '{"segment": {"content": "1","answer": "1", "keywords": ["a"], "enabled": false}}' 5.2.2. Enhanced Knowledge Retrieval for Work Order Scheduling Figure 4 Flowchart of an implementation plan to enhance knowledge retrieval for work order scheduling The retrieval system adopts a dual-channel architecture, selecting either a full-text search channel, a vector search channel, or a combination of both depending on the scenario. It also incorporates a user feedback optimization mechanism to improve search performance.

[0104] Full-text search channel: Suitable for scenarios sensitive to keywords, such as technical documents, standards, and product specifications, where precise matching of a term, model, code, ID, formula, etc. is required; for large amounts of text requiring precise positioning, such as a knowledge base with millions of documents, full-text search can directly hit relevant document fragments; for professional terminology and named entity retrieval, such as searching for information related to "Shangdazha Yintai District" in a knowledge base, keyword search is more practical.

[0105] Vector retrieval channel: Suitable for semantically similar but keyword-different situations, such as when user questions and document wording are inconsistent. For example, the question is "How to reduce power grid losses?", while the document says "reduce line losses" or "improve power transmission efficiency". Natural language question answering scenarios, such as when users are used to asking questions in everyday language, for example, "How long does it take for an electric vehicle charging station to fully charge?", while the document says "Full charge time is about 6 hours". Search result fusion: The search results from the two channels are merged and sorted according to the set weight ratio, and then the Reranker re-ranking model is used to sort them according to their relevance to the user's question.

[0106] User feedback and optimization: In the scheduling results output dialog box, users can annotate the answers with likes or dislikes, and can also click the edit button to modify the content. When answering similar questions in the future, the saved user-annotated data is loaded and provided to the large model for reference, achieving the goal of continuous optimization of subsequent question-and-answer sessions.

[0107] 5.2.3. Construction of a Multi-Factor Work Order Scheduling Intelligent Agent The multi-factor work order scheduling intelligent agent is built based on the Guangming Power large model. By splitting different functional modules, it forms three functional units: data verification agent, factor evaluation agent, and scheduling decision agent. Through integration with the knowledge base and the work order scheduling-oriented retrieval enhancement mechanism, it comprehensively scores five core factors: distance, urgency, customer, proficiency, and habit, and executes work order sorting based on these factors to achieve intelligent work order scheduling under multi-factor constraints.

[0108] Figure 5 Flowchart for constructing a multi-factor work order scheduling intelligent agent.

[0109] Figure 6 The schematic diagram for the intelligent agent design is shown. Its core implementation process is as follows: input work order data → data verification and parsing → influencing factor calculation → weather risk assessment → comprehensive score ranking → time schedule generation → output scheduling results.

[0110] (a) Implementation of Data Verification Agent The data verification agent is used to acquire and verify basic data before work order scheduling. Its main functions include work order information parsing, personnel information acquisition, and scheduling strategy parameter verification.

[0111] In the specific implementation process, the data verification agent first obtains the work order data to be scheduled from the work order system and performs integrity verification on key information such as work order address, work order type, priority, and work time window; at the same time, it obtains the current location, skill information and available time period of the operators from the personnel management system; and performs consistency and legality checks on the strategy parameters (such as weight configuration, work time constraints, etc.) that need to be used in the scheduling process.

[0112] When missing or abnormal data is detected, the data verification agent supplements or corrects the data by calling default rules or historical reference data stored in the knowledge base, and then passes the verified structured data to the factor evaluation agent as input for subsequent calculations. Compared to work order scheduling methods that only use a single retrieval method, this invention improves the recall accuracy of scheduling-related knowledge through a fusion ranking mechanism of full-text search and vector search.

[0113] (II) Implementation of Factor Evaluation Agent The factor evaluation agent is used to quantitatively evaluate various factors affecting work order scheduling and is the core computational module of the scheduling agent. In this embodiment, the factor evaluation agent evaluates five categories of factors: distance factor, urgency factor, customer factor, proficiency factor, and habit factor. During the evaluation process, the factor evaluation agent invokes knowledge retrieval enhancement methods for work order scheduling to obtain relevant knowledge support based on the work order type and evaluation requirements. For rule-based knowledge such as work order priority rules and customer level determination criteria, full-text search is used to retrieve the corresponding rules from the knowledge base; For knowledge content with strong semantic relevance, such as historical work order processing experience and skill matching cases, vector retrieval is used to obtain similar case information. The search results are fused using preset weights and used as context input into the Guangming Power big data model to guide factor score calculation.

[0114] For example, when calculating the proficiency factor, the factor evaluation agent retrieves historical work order processing records similar to the current work order type from the knowledge base, and combines them with the technician's historical processing volume and success rate to comprehensively evaluate the skill matching degree; when calculating the distance factor, it obtains the path planning results by calling the map service interface, and calculates the distance score by combining it with the path evaluation rules stored in the knowledge base.

[0115] Finally, the factor evaluation agent outputs the factor scores and comprehensive score recommendations for each work order, which serve as input for the scheduling decision agent.

[0116] (III) Implementation of Scheduling Decision Agent The scheduling decision agent is used to generate the final work order execution order and time schedule based on the factor evaluation results.

[0117] In the specific implementation process, the scheduling decision agent first determines the weight configuration of each influencing factor based on the currently adopted scheduling strategy (comprehensive scheduling, time priority, or urgency priority); then, combining the factor score results output by the factor evaluation agent, it calculates the comprehensive score of each work order and sorts the work orders according to the score results. The strategy is stored in the system in the form of configuration parameters and is automatically read and applied by the calculation module during the scheduling calculation process.

[0118] When generating a specific schedule, the scheduling decision agent calls the work time rules, lunch break rules, and weather impact rules stored in the knowledge base, and combines them with real-time weather data to adjust the order or postpone work orders that are not suitable for execution under the current conditions. At the same time, based on the travel time information returned by the map service, it accurately calculates the connection time between work orders and generates a complete work order execution schedule.

[0119] Ultimately, the scheduling decision agent outputs scheduling results that include the work order execution order, start and end times, route planning information, and scheduling performance indicators, and supports manual review and adjustment.

[0120] Figure 7 This is a schematic diagram of a low-voltage distribution work order intelligent scheduling device provided in an embodiment of the present invention. Figure 7 As shown, the device includes: The knowledge base construction module 710 is used to process multi-source operation and distribution business data based on the Guangming Big Data Model and build the corresponding business knowledge base. The retrieval enhancement module 720 is used to obtain work orders to be scheduled, obtain retrieval results based on the business knowledge base using a dual-channel retrieval method, enhance the retrieval results and input them into the Guangming Big Data Model to obtain scheduling decision information. The scheduling execution module 730 is used to construct a multi-factor constrained scheduling model based on the scheduling decision information, and execute the scheduling model through a scheduling intelligent agent constructed based on the Guangming big data model to schedule the work orders to be scheduled.

[0121] In some possible implementations, the knowledge base construction module 710 includes: The data processing submodule is used to perform unified access and standardized processing of the multi-source operation and distribution business data; The knowledge transformation submodule is used to extract knowledge and semantically transform the processed multi-source operation and distribution business data through the Guangming Big Data Model to form structured knowledge. The knowledge base construction submodule is used to construct and store the corresponding business knowledge base based on the structured knowledge.

[0122] In some possible implementations, the knowledge base construction submodule includes: The interface scanning unit is used to perform metadata scanning of service catalog interfaces through the application framework; The information fusion unit is used to combine system interface documentation with authentication methods, interface examples, and parameter requirements. The knowledge adaptation unit is used to transform the fused information into knowledge content that the Guangming Big Model can understand, so as to build and store the business knowledge base.

[0123] In some possible implementations, the retrieval enhancement module 720 includes: The first retrieval submodule is used for first-path retrieval based on keyword matching; The second retrieval submodule is used for second-path retrieval based on semantic vector matching; The retrieval result determination submodule is used to determine the retrieval result based on the results of the first path retrieval and the second path retrieval.

[0124] In some possible implementations, the retrieval enhancement module 720 includes: The result optimization submodule is used to reorder and filter the search results based on the context information of the work order to be scheduled. The context input submodule is used to input the processed search results as context into the Guangming Big Model; The reasoning and decision-making submodule is used to perform reasoning analysis through the Guangming big model to obtain the corresponding scheduling decision information.

[0125] In some possible implementations, the scheduling execution module 730 includes: The constraint determination submodule is used to determine at least one constraint factor among work order attributes, work location, personnel capabilities, and work time window. The model building submodule is used to build the corresponding multi-factor constrained scheduling model based on the constraints and the scheduling decision information.

[0126] In some possible implementations, the scheduling execution module 730 includes: The constraint parsing submodule is used to parse the constraints of the scheduling model through the scheduling agent; The scheduling execution submodule is used to sort, allocate time, and plan paths for the work orders to be scheduled based on the parsing results, so as to obtain the scheduling results.

[0127] The low-voltage operation and maintenance work order intelligent scheduling device provided in the embodiments of the present invention can execute the low-voltage operation and maintenance work order intelligent scheduling method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0128] Figure 8 This is a schematic diagram of the structure of an electronic device for implementing the intelligent scheduling method for low-voltage distribution work orders according to embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0129] like Figure 8 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded into the RAM 13 from storage unit 18. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

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

[0131] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of 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 suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the intelligent scheduling method for low-voltage distribution work orders.

[0132] In some embodiments, the low-voltage operation and maintenance work order intelligent scheduling method can be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the low-voltage operation and maintenance work order intelligent scheduling method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to execute the low-voltage operation and maintenance work order intelligent scheduling method by any other suitable means (e.g., by means of firmware).

[0133] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include: implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0134] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0135] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0136] To provide interaction with a user, the systems and techniques described herein 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 pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; 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 sound input, voice input, or tactile input).

[0137] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0138] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0139] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0140] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for intelligent scheduling of low-voltage operation and maintenance work orders, characterized in that, include: Based on the Guangming Big Data Model, multi-source marketing and distribution business data are processed to build a corresponding business knowledge base; Obtain the work orders to be scheduled, retrieve the search results using a dual-channel retrieval method based on the business knowledge base, enhance the search results and input them into the Guangming Big Data Model to obtain scheduling decision information; A multi-factor constrained scheduling model is constructed based on the scheduling decision information. The scheduling model is executed by a scheduling agent constructed based on the Guangming Big Data Model to schedule the work orders to be scheduled.

2. The method according to claim 1, characterized in that, The process of processing multi-source marketing and distribution business data based on the Guangming Big Data Model to construct a corresponding business knowledge base includes: The multi-source operation and distribution business data are uniformly accessed and standardized. The Guangming Big Data Model is used to extract knowledge and perform semantic transformation on the processed multi-source marketing and distribution business data to form structured knowledge. The corresponding business knowledge base is constructed and stored based on the structured knowledge.

3. The method according to claim 2, characterized in that, The construction and storage of the corresponding business knowledge base based on the structured knowledge includes: Metadata scanning of the service catalog interface is performed through the application framework; Combine the system interface documentation with information on authentication methods, interface examples, and parameter requirements; The merged information is transformed into knowledge content that the Guangming Big Data model can understand, in order to build and store the business knowledge base.

4. The method according to claim 1, characterized in that, The process of obtaining search results using a dual-channel retrieval method based on the business knowledge base includes: First-path retrieval based on keyword matching; Second-path retrieval based on semantic vector matching; The search results are determined based on the results of the first and second search methods.

5. The method according to claim 1, characterized in that, The process of enhancing the search results and inputting them into the Guangming Big Data Model to obtain scheduling decision information includes: The search results are reordered and filtered based on the context information of the work order to be scheduled; The processed search results are used as context input into the Guangming Big Data Model; By performing reasoning analysis using the aforementioned Guangming Big Data Model, the corresponding scheduling decision information is obtained.

6. The method according to claim 1, characterized in that, The construction of a multi-factor constrained scheduling model based on the scheduling decision information includes: Determine at least one constraint factor among work order attributes, work location, personnel capabilities, and work time window; Based on the constraints and the scheduling decision information, a scheduling model with corresponding multi-factor constraints is constructed.

7. The method according to claim 1, characterized in that, The process of executing the scheduling model through a scheduling agent built based on the Guangming Big Data model to schedule the work orders to be scheduled includes: The scheduling agent parses the constraints of the scheduling model. Based on the analysis results, the work orders to be scheduled are sorted, time-allocated, and path-planned to obtain the scheduling results.

8. A low-voltage operational work order intelligent scheduling device, characterized in that, include: The knowledge base construction module is used to process multi-source operation and distribution business data based on the Guangming Big Data Model and build the corresponding business knowledge base. The retrieval enhancement module is used to obtain work orders to be scheduled, obtain retrieval results based on the business knowledge base using a dual-channel retrieval method, enhance the retrieval results and input them into the Guangming Big Data Model to obtain scheduling decision information; The scheduling execution module is used to construct a multi-factor constrained scheduling model based on the scheduling decision information, and execute the scheduling model through a scheduling agent constructed based on the Guangming big data model to schedule the work orders to be scheduled.

9. An electronic device, characterized in that, The electronic device includes: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores a computer program that can be executed by the at least one processor, which is then executed by the at least one processor to enable the at least one processor to perform the intelligent scheduling method for low-voltage operation and distribution work orders as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the intelligent scheduling method for low-voltage operation and distribution work orders as described in any one of claims 1-7.

Images