Intelligent low-voltage industry expansion new installation, increment and decrement method and related equipment

Abstract

The application relates to an intelligent low-voltage business expansion, new installation, capacity increase and reduction method and related equipment, which comprises the following steps: auditing power consumption application data and power consumption demand information, generating a business work order after the audit is passed; obtaining power grid data based on the power consumption address of a customer, determining an access scheme based on the power consumption demand information and the power grid data, and generating an investigation result based on the access scheme; generating a metering material demand list according to the investigation result and the power consumption demand information, matching inventory materials according to the metering material demand list to generate a material scheduling plan, generating a construction plan according to the access scheme, the material scheduling plan, real-time operation data of the power grid and construction team resource data, collecting construction data and performing real-time verification, auditing a construction completion report, and performing a power-on operation after the audit is passed; collecting whole-process data for intelligent verification, generating an electronic file and storing the electronic file after the verification is passed. The application realizes intelligent whole-process processing of low-voltage business expansion, new installation, capacity increase and reduction, and improves the business processing efficiency.

Description

Technical Field

[0001] This application relates to the field of power systems and their automation technology, and in particular to an intelligent method and related equipment for expanding, increasing or decreasing capacity in low-voltage power systems. Background Technology

[0002] Low-voltage business expansion and capacity increase / decrease is a core foundation of electricity marketing services. It refers to providing low-voltage electricity customers (including residential customers and non-residential customers such as small shops, restaurants, and processing plants) with a full-chain service from electricity application to final power connection, as well as subsequent power capacity adjustment. It is an important link in the interaction between the power grid company and customers, and is directly related to the customer's electricity experience and the safe and stable operation of the power grid.

[0003] Currently, the processing flow of this business mainly follows a model that combines traditional manual operation with some system assistance, and can be divided into four stages: (1) During the business acceptance stage, customers submit electricity application materials through offline business halls or online APP. Staff manually check the completeness and compliance of the materials one by one. After the materials are approved, they are entered into the marketing system to generate a work order.

[0004] (2) During the on-site survey stage, the survey personnel need to go to the site to verify the power supply conditions, rely on manual experience to judge the feasibility of grid connection, determine the location of the connection point, and manually enter the survey results into the system.

[0005] (3) During the meter installation and power connection stage, staff manually check the material requirements and coordinate the scheduling. After formulating the construction plan, the construction team is arranged to work on site and the construction information is sent back after completion.

[0006] (4) During the work order archiving stage, staff manually collect, organize, and verify the data throughout the process, and complete paper scanning or electronic entry to achieve work order closure.

[0007] Some power companies have tried to introduce information systems to assist in business processing, such as realizing work order flow through marketing systems and obtaining some power grid data through metering automation systems. However, core processes still rely heavily on manual intervention, achieving only "online flow" and not reaching the level of "intelligent decision-making".

[0008] The relevant technologies have at least the following shortcomings: First, data review relies on manual labor, which is inefficient, prone to errors, and prone to backlog during peak periods; second, on-site surveys require personnel to travel back and forth, which is costly and time-consuming, and the judgment of access points relies on personal experience, resulting in poor accuracy; third, the data interoperability of multiple systems such as marketing, materials, and power grids is poor, requiring manual synchronization, resulting in poor coordination between various links and a long overall process cycle; fourth, data archiving requires manual sorting, verification, and entry, which is tedious, prone to omissions, and has high management costs. Summary of the Invention

[0009] This application addresses the aforementioned shortcomings or deficiencies by providing an intelligent method and related equipment for low-voltage business expansion, installation, capacity increase, and reduction. This application enables intelligent processing of the entire low-voltage business expansion, installation, capacity increase, and reduction process, improving efficiency and accuracy.

[0010] This application provides an intelligent method for expanding, installing, increasing, or reducing capacity in low-voltage electrical applications, based on a first aspect. The method includes: receiving electricity application materials and electricity demand information sent by a client; conducting multi-dimensional review of the application materials and demand information; generating a work order upon successful review; responding to the work order by obtaining grid data based on the customer's electricity address; determining an access scheme based on the electricity demand information and grid data; generating a survey result based on the access scheme; generating a metering material demand list based on the survey result and electricity demand information; matching inventory materials according to the metering material demand list to generate a material scheduling plan; generating a construction plan based on the access scheme, material scheduling plan, real-time grid operation data, and construction team resource data; collecting and verifying construction data in real-time during construction based on the construction plan; reviewing the construction completion report upon receipt; and performing a power-on operation upon successful review; and upon completion of the power-on operation, collecting all process data for intelligent verification; generating and storing an electronic file upon successful verification.

[0011] According to a second aspect, this application provides an intelligent low-voltage industrial expansion and capacity reduction device, which includes: The intelligent review module is used to receive electricity application materials and electricity demand information sent by the client, conduct multi-dimensional review of the electricity application materials and electricity demand information, and generate a business work order after the review is approved. The access scheme and survey result generation module is used to respond to business work orders, obtain power grid data based on the customer's electricity address, determine the access scheme based on electricity demand information and power grid data, and generate survey results based on the access scheme. The construction management module is used to generate a metering material demand list based on the survey results and electricity demand information, match the inventory materials based on the metering material demand list to generate a material scheduling plan, generate a construction plan based on the access scheme, material scheduling plan, real-time power grid operation data and construction team resource data, collect construction data and perform real-time verification during the construction process based on the construction plan, and review the construction completion report upon receiving it. After the review is approved, the power-on operation is executed. The intelligent archiving module is used to collect all process data for intelligent verification in response to the completion of the power-on operation. After the verification is successful, electronic archives are generated and stored.

[0012] According to a third aspect, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the intelligent low-voltage industrial expansion and capacity reduction methods described in the above embodiments.

[0013] According to a fourth aspect, this application provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executed, implements any of the intelligent low-voltage industrial expansion and capacity reduction methods described in the above embodiments.

[0014] This application utilizes fully automated processing of intelligent review, intelligent exploration, intelligent construction, and intelligent archiving to improve business processing efficiency and accuracy, reduce exploration costs and business cycles, and enhance multi-system collaboration and business transparency.

[0015] First, the electricity application materials and electricity demand information undergo multi-dimensional review, and a work order is generated upon approval. Previously, the technology relied on manual verification of the completeness and compliance of the materials, which was prone to omissions and misjudgments, leading to severe backlogs during peak periods. This new application system replaces manual verification with multi-dimensional intelligent review, automatically completing triple checks on completeness, compliance, and authenticity. This eliminates the subjective errors and fatigue effects of manual review, significantly improving review efficiency and accuracy, and avoiding rework in subsequent processes due to review issues.

[0016] Secondly, the grid connection scheme is determined based on grid data and electricity demand information, and survey results are generated. Related technologies require survey personnel to verify power supply conditions on-site and judge access points based on experience, which is costly, time-consuming, and the accuracy of access point selection is difficult to guarantee. This application uses intelligent algorithms to analyze grid topology and real-time load data, automatically determines the grid connection scheme, and generates survey results. It transforms the traditional survey decision-making process, which relies on manual experience and repeated on-site visits, into a data-driven, remote, intelligent judgment, reducing on-site survey costs and time, and improving the accuracy of access point selection.

[0017] Furthermore, based on the survey results, a material demand list is generated, a material scheduling plan is generated by matching inventory, and a construction plan is generated. During construction, construction data is verified in real time, and power is supplied only after the construction report is reviewed. In related technologies, data from multiple systems such as marketing, materials, and power grids are fragmented, requiring manual synchronization and coordination. Material scheduling and construction plans rely on manual formulation, making it difficult to guarantee construction quality. This application, through automated collaboration, connects the entire data chain from planning to materials, construction, and acceptance. Intelligent algorithms automatically generate material lists, scheduling plans, and construction plans, and verify data in real time and automatically review construction reports during construction. This achieves seamless integration and closed-loop quality control between materials and construction stages, shortening the business cycle and ensuring construction quality.

[0018] Finally, all process data is collected and intelligently verified to generate and store electronic archives. Previously, related technologies required manual data organization, verification, and entry, which was tedious and prone to omissions. This application replaces manual organization with intelligent verification, automatically performing integrity and consistency checks to generate structured electronic archives. This eliminates the tedious operations and error risks of manual archiving, reduces management costs, and improves data retrieval efficiency. Attached Figure Description

[0019] Figure 1 This is a flowchart illustrating an intelligent low-voltage industrial expansion and capacity reduction method according to one or more embodiments of this application; Figure 2 This describes the overall business processing flow in one or more embodiments of this application; Figure 3 This is a diagram illustrating the architecture of the intelligent data review model in one or more embodiments of this application. Figure 4 This is an architecture diagram of the intelligent access point judgment model in one or more embodiments of this application; Figure 5 This is a schematic diagram of the structure of an intelligent low-voltage industrial expansion and capacity reduction device in one or more embodiments of this application; Figure 6 This is a schematic diagram of the internal structure of a computer device according to one or more embodiments of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings. It should be understood that the described embodiments are merely some embodiments of this application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0021] In the following description, when referring to the accompanying drawings, the same numbers in different drawings denote the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0022] In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances. Furthermore, in the description of this application, unless otherwise stated, "multiple" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0023] To address the shortcomings or defects of related technologies, this application provides an intelligent method for expanding, adding, or reducing the capacity of low-voltage installations. This method improves business processing efficiency and accuracy through fully automated processing of intelligent review, intelligent surveying, intelligent construction, and intelligent archiving. It also reduces surveying costs and business cycles, and enhances multi-system collaboration and business transparency.

[0024] In some exemplary embodiments of this application, such as Figure 1 As shown, the method includes steps S110-S140. This method can be applied to intelligent low-voltage industrial expansion, installation, capacity increase / decrease systems (hereinafter referred to as the system, which can be a standalone server or a server cluster). The following section combines... Figure 2 The overall business process flow is shown, with each step explained in detail.

[0025] S110: Receives electricity application materials and electricity demand information sent by the client, conducts multi-dimensional review of the electricity application materials and electricity demand information, and generates a business work order after the review is approved.

[0026] Electricity application materials refer to the supporting documents submitted by the customer, including photos of their ID card, property ownership certificate or business license, list of electrical equipment, and photos of the electricity address.

[0027] Electricity demand information refers to the business parameters filled in by the customer, including electricity type (residential or non-residential), installed capacity, expected power connection time, scheduled service date, purpose of connection and / or electricity category, etc.

[0028] Multi-dimensional review refers to the intelligent verification of the completeness, compliance, and authenticity of electricity application materials and electricity demand information.

[0029] A work order is a standardized work order that includes customer information, application information, and review results, and is used to drive subsequent processes.

[0030] In this step, the system receives electricity application materials and electricity demand information uploaded by the customer through the client. The application materials include both image and text documents. The system performs OCR (Optical Character Recognition) on the image materials to extract text information. Then, the extracted text information, along with the text materials and electricity demand information, is input into a pre-trained semantic understanding model for semantic encoding and keyword extraction. The image materials are input into an image recognition model to extract image features. The extracted text and image features are then fused and combined with pre-defined rule bases for completeness, compliance, and authenticity for a comprehensive judgment. If the review is successful, a business work order containing customer information and electricity demand is generated; if the review fails, the item is marked as unqualified and supplementary requirements are pushed to the client.

[0031] Compared to the manual verification of data using related technologies, this step replaces manual judgment with multi-dimensional intelligent review, automatically completing triple verification of completeness, compliance, and authenticity. This eliminates the subjective errors and fatigue effects of manual review, significantly improving review efficiency and accuracy, effectively handling the massive number of applications during peak business periods, and avoiding rework in subsequent processes due to review issues.

[0032] S120: In response to a business work order, obtain power grid data based on the customer's electricity address, determine the access scheme based on electricity demand information and power grid data, and generate survey results based on the access scheme.

[0033] Power grid data is multi-source power grid data, which includes power grid topology data (including line routes, transformer locations and capacities, distribution of existing access points, etc.), real-time load data (including transformer load rate, line current, etc.), and visualized image data, among other multi-source power grid information.

[0034] An access scheme refers to a technical solution that includes data such as the optimal access point location, line laying method, and metering point location.

[0035] The survey results refer to the final access plan and related verification records after on-site or remote surveys by staff.

[0036] In this step, the system responds to business work orders by retrieving grid topology data from the grid management platform based on the customer's electricity address, real-time load data from the metering automation system, and visualized image data from the grid visualization platform. Based on electricity demand information and grid data, candidate access points are screened using a graph neural network, the feasibility of each candidate access point is evaluated using a gradient boosting tree model, and the optimal access point is determined from the feasible candidate access points using the analytic hierarchy process. An access scheme is generated, including the access point location, line laying method, and metering point location. Based on the access scheme, survey results are generated for subsequent construction execution.

[0037] Compared to related technologies that require on-site verification by surveyors and judgment of access points based on experience, this step uses intelligent algorithms to analyze the power grid topology and real-time load data, automatically determine the access scheme and generate survey results. This transforms the traditional survey decision-making process, which relies on manual experience and on-site visits, into a data-driven remote intelligent judgment, significantly reducing on-site survey costs and time, improving the accuracy of access point selection, and avoiding construction rework and safety hazards caused by unreasonable access point selection.

[0038] S130: Generate a metering material demand list based on the survey results and electricity demand information; match inventory materials according to the metering material demand list to generate a material scheduling plan; generate a construction plan based on the access scheme, material scheduling plan, real-time power grid operation data, and construction team resource data; collect construction data and perform real-time verification during the construction process based on the construction plan; review the construction completion report upon receiving it; and perform power-on operation after the review is passed.

[0039] The metering material requirement list is a detailed list of materials automatically generated based on the access plan, including information such as the model, specifications, and quantity of metering instruments, cables, and safety protection materials.

[0040] The material scheduling plan is an execution plan for material preparation, transportation, and handover generated based on inventory matching results.

[0041] A construction plan is an execution scheme that includes the construction period, allocation of construction personnel, standardized construction procedures, and technical requirements.

[0042] Construction data refers to real-time data collected during the construction process, such as the installation location of meters and the wiring status.

[0043] A construction completion report is a final report submitted after construction is completed, which includes construction photos, test data (line insulation resistance, meter errors, etc.).

[0044] In this step, the system matches meter models, cable specifications, and safety protection materials based on the survey results and electricity demand information, generating a metering material demand list. It then matches the inventory of materials to the demand list; if inventory is sufficient, a material issuance order is generated; otherwise, a procurement process is triggered, generating a material scheduling plan. Based on the access scheme, material scheduling plan, real-time power grid operation data, and construction team resource data, the system determines the construction time period, matches the construction team and personnel responsibilities, generates a standardized construction process, and forms a construction plan. During construction, the system collects and verifies construction data in real time, pushing out warning information when deviations are detected. After construction is completed, the system receives a construction completion report, automatically reviews the construction photos and test data, and executes the power-on operation after approval.

[0045] Compared to the fragmented data across multiple systems such as marketing, materials, and power grids in related technologies, which require manual synchronization, this step achieves seamless data integration across the entire process from planning to materials, construction, and acceptance through automated collaboration. Intelligent algorithms automatically generate material lists, scheduling plans, and construction plans, and verify data in real time and automatically review construction reports during construction. This enables seamless integration and closed-loop quality control between materials and construction, significantly shortening the business cycle, ensuring construction quality, and reducing rework rates.

[0046] S140: In response to the completion of the power-on operation, collect all process data for intelligent verification, and generate and store electronic files after successful verification.

[0047] The complete set of data includes electricity application materials, electricity demand information, multi-dimensional review results of the above electricity application materials and electricity demand information, access plans, survey reports, survey results, material scheduling plans, construction plans, construction completion reports, metering operation data, and power supply contracts generated throughout the entire business process.

[0048] Intelligent verification refers to the integrity verification (such as determining whether the data is complete) and consistency verification (comparing key information at each stage, such as customer name, electricity address, and installed capacity, to ensure consistency) of the data throughout the entire process.

[0049] Electronic archives refer to structured electronic documents that include classification catalogs and search tags.

[0050] In this step, upon completion of the power-on operation, the system automatically collects data from the entire business process. The collected data is then categorized by business stage, standardized in format, and organized. Based on a pre-set list of archived data, the system performs a completeness check on all data, marking missing items. It also compares key information from each stage (customer name, electricity address, applied capacity) for consistency checks, marking inconsistencies. After successful verification, an electronic archive containing a categorized directory and search tags is generated, encrypted and stored using blockchain, and supports multi-dimensional searching by customer name, electricity address, work order number, and other criteria.

[0051] Compared to related technologies that require manual data processing, verification, and entry throughout the entire process, this step replaces manual processing with intelligent verification, automatically completing integrity and consistency checks, generating structured electronic archives, eliminating the tedious operation and error risk of manual archiving, reducing management costs, improving data retrieval efficiency, and providing convenience for subsequent queries and audits.

[0052] In some embodiments, multi-dimensional review of electricity application materials and electricity demand information includes: performing text recognition on image materials in the electricity application materials to obtain text information within the images; inputting the text information within the images, the text materials in the electricity application materials, and the electricity demand information into a semantic understanding model for semantic encoding and keyword extraction to obtain text features; inputting the image materials in the electricity application materials into an image recognition model to extract image features; fusing text features and image features, and performing integrity review, compliance review, and authenticity review based on the fused features and a preset integrity rule base, compliance rule base, and authenticity rule base.

[0053] Semantic understanding models are deep learning models used to understand and extract semantic information from text. For example, the BERT (Bidirectional Encoder Representations from Transformers) model can be used to encode semantic information and extract key information from textual materials.

[0054] Image recognition models refer to deep learning models used to analyze and extract image features. For example, CNN (Convolutional Neural Networks) models can be used to extract features from image data and can be combined with OCR technology to convert text in images into a processable text format.

[0055] The integrity rule base, compliance rule base, and authenticity rule base refer to pre-defined sets of rules used for auditing and judgment. The integrity rule base defines the list of required documents for different customer types (residents or non-residents); the compliance rule base includes policy requirements such as power industry regulations and the Guangdong Power Grid Low-Voltage Business Expansion Management Standards; and the authenticity rule base includes comparison standards for judging the authenticity of address information, photos, etc.

[0056] The operation process of this embodiment is as follows: After receiving the electricity application materials and electricity demand information uploaded by customers, the system can call the intelligent data review model to conduct multi-dimensional review of these data.

[0057] The input data for the intelligent document verification model includes all application documents submitted by the customer (electricity application documents and electricity demand information). Furthermore, the input data can also include related auxiliary data for all application documents. Related auxiliary data can include the power grid topology map from the Southern Power Grid Smart View Platform, satellite maps (for authenticity verification), supplementary customer information synchronized from the government service platform (for compliance verification), preset document lists, and rule bases, etc.

[0058] The output data of the intelligent document review model is structured review results, which specifically include the following data: (1) Overall audit conclusion: Audit passed / Audit failed; (2) Itemized audit results, namely: a. Completeness audit results, which are used to identify missing data items, such as "non-residential customers are missing land use rights certificates" or "the list of electrical equipment does not indicate the capacity of each item"; b. Compliance audit results, which are used to identify non-compliant items and the basis for violations, such as "the ID card has expired, which does not comply with Article X of the 'Administrative Measures for Electricity User Application for Connection'" and "the total capacity of the list of electrical equipment exceeds the standard, which does not comply with the low-voltage power supply standard". c. Authenticity verification results, which are the results of judging the authenticity of the electricity address, such as "the electricity address photo matches the satellite map and is authentic and valid", "the photo shows signs of tampering and its authenticity is questionable", etc. (3) Auxiliary output, namely, suggestions for supplementing information. When there are projects that have not passed the review, specify the types and requirements of supplementary information for the projects that have not passed the review.

[0059] In some examples, the architecture of the intelligent document review model can be found in [reference needed]. Figure 3 As shown below, in conjunction with Figure 3 The model is explained.

[0060] Regarding the model architecture of the intelligent document review model, it adopts a hybrid architecture of BERT semantic understanding model and CNN image recognition model. It belongs to a multimodal fusion model and is divided into three layers: input layer, feature extraction layer and fusion decision layer. Each layer works together to achieve full-dimensional review of document integrity, compliance and authenticity.

[0061] The input layer is responsible for receiving various types of raw data (such as electricity application materials, electricity demand information, and associated auxiliary data) and performing preliminary format adaptation to provide standardized input for subsequent processing. These raw data types can be divided into two main categories: image data and text data. Text data includes ID card text information, property certificate text information, business license text information (for non-residential customers), lists of electrical equipment, and electricity demand information (application capacity, electricity usage type, etc.). Image data includes photos of ID cards, property certificates, business licenses, electricity addresses, and on-site scenes.

[0062] The feature extraction layer is used for text feature extraction and image feature extraction.

[0063] The text feature extraction operation is undertaken by the BERT semantic understanding model. Specifically, it is necessary to perform semantic encoding and keyword extraction (such as core information like ID card validity period, business license annual inspection status, and power capacity) on textual materials such as ID card text information, business license information, electrical equipment list, and electronic forms.

[0064] Image feature extraction is handled by a CNN image recognition model. Specifically, it needs to extract image features (such as ID card anti-counterfeiting marks, address scene features, and photo authenticity features) from image data such as ID card photos, electricity address photos, and on-site photos through convolutional layers and pooling layers. At the same time, it uses OCR technology to extract text information (such as ID card number, name, address, etc.) from the image and convert it into a processable text format.

[0065] The fusion decision layer is used to fuse the text features and image features extracted by the feature extraction layer to obtain fused features. The fusion operation can be implemented based on feature concatenation and attention mechanisms to highlight the importance of key features. After obtaining the fused features, a pre-set rule base (including power industry regulations, Guangdong power grid low-voltage business expansion management specifications, and lists of residential and non-residential data) is used to complete three verification judgments: completeness, compliance, and authenticity, and output the final verification result.

[0066] Specifically, based on the fusion feature, the system can call the preset integrity rule base, compliance rule base, and authenticity rule base to perform three types of audits respectively: (1) Completeness review, including determining whether the extracted key information covers all the necessary items required by the rule base.

[0067] (2) Compliance review, including verifying whether key information (such as the validity period of ID card and the annual inspection status of business license) meets the requirements of the rules base.

[0068] (3) Authenticity verification, including comparing image features with text features or with external data (such as satellite maps) to verify the authenticity of the data.

[0069] Compared to the manual item-by-item verification or simple system verification based on a single dimension in related technologies, this embodiment adopts a multimodal fusion verification method, which has the following advantages over related technologies: First, the review process is more comprehensive. By integrating both text and image features and based on three pre-set rule bases, it achieves a holistic and multi-dimensional review of the completeness, compliance, and authenticity of the data, thus overcoming the shortcomings of manual review, which is prone to omissions, or the one-sidedness of judgment by a single model.

[0070] Secondly, the accuracy of the review is higher. The powerful semantic understanding capability of the BERT model ensures the accuracy of text information extraction, while the CNN model's ability to capture deep image features effectively identifies anomalies such as forgery and tampering. The fusion of these two models, along with the attention mechanism's focus on key features, significantly improves the overall accuracy of the review.

[0071] Finally, the review process is more intelligent and efficient. The entire process requires no human intervention; the system automatically completes all steps from feature extraction to multi-dimensional rule judgment, greatly improving review efficiency, especially suitable for large-scale application processing scenarios during peak business periods.

[0072] In some embodiments, the access point intelligent judgment model can be invoked to determine the access scheme based on electricity demand information and power grid data. The access scheme includes the following steps (1)-(7). The access point intelligent judgment model will be explained first, and then the steps (1)-(7) will be explained in detail.

[0073] The input data for the intelligent access point judgment model mainly comes from multi-source power grid data (i.e., fused power grid data) and customer electricity demand (i.e., electricity demand information). Specifically, it includes power grid topology data processed by the data preprocessing module (such as line route, transformer location and capacity, and existing access point distribution), real-time power grid load data (such as transformer load rate, line current, and load fluctuation), on-site visualized image data (such as on-site images synchronized by the Southern Power Grid Smart View Platform), and customer electricity demand data (such as total installed capacity, sub-item capacity, electricity usage nature, and electricity usage time period).

[0074] The output data of the intelligent access point judgment model is a structured access point judgment result, which specifically includes a list of candidate access points, the feasibility assessment results of each candidate access point (accessible, inaccessible, or requiring adjustment before access), the optimal access point and supporting access scheme (including access point location, line laying method and metering point setting location, etc.), and outputs the judgment basis for each item (such as "the transformer load rate is 75% after access, which meets the 80% safe load threshold", "the access point is 50 meters away from the customer's electricity address, and the line loss is the lowest", etc.).

[0075] The architecture of the intelligent access point judgment model can be found in [reference needed]. Figure 4 As shown, it adopts a combined architecture of graph neural network (GNN) and gradient boosting tree (XGBoost), and is divided into three layers: input layer, feature fusion layer, and decision output layer.

[0076] The input layer is responsible for receiving data from multiple sources and adapting the formats to provide standardized input for subsequent processing.

[0077] The feature fusion layer is divided into a GNN power grid topology analysis unit and an XGBoost capacity assessment unit. The GNN power grid topology analysis unit is responsible for extracting power grid topology features, access point distribution features, and line loss information. The XGBoost capacity assessment unit is responsible for analyzing real-time power grid load data, remaining access point capacity, and load carrying capacity. The features output by the two units can be deeply fused through the feature fusion unit (which uses a weighted fusion algorithm to achieve feature fusion) to improve feature utilization. The decision output layer includes three units: candidate access point screening, feasibility assessment, and optimal decision, which are used to realize the entire process of decision-making from candidate access point screening to the generation of the optimal access scheme.

[0078] The steps in this embodiment are explained as follows: (1) Construct the power grid topology data in the power grid data into a graph structure. The nodes in the graph structure include transformer nodes, existing access point nodes and customer electricity address nodes. The edges in the graph structure represent connecting lines, and the weight of the edges is assigned according to the line length and line loss coefficient.

[0079] Power grid topology data refers to information describing the connection relationships of power grid equipment, such as the route of lines, the location of transformers, and the distribution of existing access points.

[0080] First, the system constructs a graph structure from the preprocessed power grid topology data. In this graph structure, nodes represent transformers, existing connection points, and customer electricity addresses. Based on the specific information they represent, nodes can be further categorized into transformer nodes, existing connection point nodes, and customer electricity address nodes. Edges represent connecting lines, and their weights are assigned based on line length and loss coefficients; generally, longer lines and higher loss coefficients result in higher weights. Abstracting the power grid topology into a graph structure lays the data foundation for subsequent graph neural network processing, enabling the quantification and calculation of the complex interconnections within the power grid.

[0081] (2) The graph structure is processed by the graph neural network algorithm to learn the feature representation of each node, and the graph is traversed based on the customer's electricity address node to filter out candidate access points.

[0082] The system uses a graph neural network algorithm to embed nodes into the graph structure, learning a low-dimensional vector representation for each node to encode its topological information and attributes, focusing on capturing the topological relationships between nodes (such as the affiliation between the access point and the transformer, and the line connectivity between access points). Then, using the customer's electricity address as the core node, a graph traversal algorithm (such as breadth-first search, BFS) is used to filter all existing access points within a preset range (e.g., within 100 meters) of the customer's electricity address. These access points are then sorted by distance from nearest to farthest and line loss from lowest to highest, selecting the top 3-5 access points as candidate access points, while excluding access points with aging lines or near-saturation loads. This step leverages the powerful representational capabilities of graph neural networks to automatically identify the access point with the highest correlation to the customer's address. Compared to blind selection based on experience, this improves the scientific rigor and relevance of candidate access points and reduces the computational burden of subsequent evaluation.

[0083] (3) Obtain standardized feature data for each candidate access point; the standardized feature data includes the remaining capacity of the access point, the real-time load rate of the transformer to which it belongs, the real-time current of the line, the line current carrying capacity, as well as the applied capacity, the load fluctuation coefficient during the power consumption period, and the historical load peak data in the power demand information.

[0084] The system acquires core data associated with each candidate access point as input features. This core data includes the remaining capacity of the access point, the real-time load rate of its associated transformer, the real-time current of the line, the line current carrying capacity, and the applied capacity, load fluctuation coefficient during electricity consumption periods, and historical peak load data from the electricity demand information. These input features are standardized (e.g., normalized to the 0-1 range) to eliminate dimensional differences, ensuring comparability of features of different scales in subsequent model evaluation and avoiding model bias caused by inconsistent data scales.

[0085] (4) Input the standardized feature data of each candidate access point into the access feasibility test model to obtain the feasibility prediction probability of each candidate access point.

[0086] The standardized feature data of each candidate access point is input into a pre-trained access feasibility verification model. This model employs the Gradient Boosting Tree (XGBoost) algorithm, pre-trained on a historical access case dataset. Through multi-round decision tree ensemble inference, it learns the weights of each input feature (e.g., transformer load rate and remaining capacity of the access point have the highest weights) and outputs the feasibility prediction probability (0-1 range) for each candidate access point. Compared to traditional manual estimation relying on simple empirical formulas, this model can comprehensively analyze multi-dimensional dynamic data to accurately quantify the probability of successful access.

[0087] (5) Verify the feasibility prediction probability of each candidate access point according to the preset feasibility probability threshold and power grid safety specifications, and generate the feasibility assessment result of each candidate access point based on the verification result. The feasibility assessment result is that the access point can be accessed, cannot be accessed, or needs to be adjusted before access.

[0088] The system pre-sets feasibility judgment thresholds (e.g., 0.7 is the feasible threshold, 0.3 is the infeasible threshold, and 0.3-0.7 is the threshold requiring adjustment), and combines these with power grid safety standards (e.g., transformer load rate ≤ 80% after connection, line current ≤ rated current carrying capacity) to perform a secondary verification of the predicted probability. For example: If the predicted probability is ≥0.7 and the security specifications are met, it is determined that "access is possible"; If the predicted probability is ≤0.3 or does not meet the security specifications, it is determined as "cannot be accessed"; If the predicted probability is between 0.3 and 0.7, it is determined that "adjustment is required before access" (such as adjusting the customer's electricity usage period or optimizing the line wiring method).

[0089] This step combines machine learning output with hard safety rules, preserving the model's flexibility while ensuring absolute safety in decision-making, thus avoiding the risk of violations that may arise from relying solely on probability.

[0090] (6) Based on the access distance, construction difficulty, access cost, power grid safety and load stability indicators, the optimal access point is determined from the candidate access points that can be accessed according to the feasibility assessment results by using the analytic hierarchy process.

[0091] For candidate access points whose feasibility assessment results are "accessible", the system determines the optimal access point based on five indicators: access distance, construction difficulty, access cost, power grid safety, and load stability, using the analytic hierarchy process.

[0092] The operational process for determining the optimal access point using the analytic hierarchy process (AHP) may include: The first step is to construct a hierarchical model. The model consists of three levels: the target layer, the criteria layer, and the alternative layer. The target layer determines the optimal access point, the criteria layer contains core evaluation indicators (including access distance, construction difficulty, access cost, grid security, and load stability), and the alternative layer comprises candidate access points deemed "fit for access" after feasibility assessment. The second step is to assign weights to the indicators and conduct consistency checks. Specifically, the 1-9 scale method of the Analytic Hierarchy Process (AHP) is used, combined with power grid business specifications and historical access cases, to assign values ​​to each indicator in the criteria layer (with power grid security and load stability having the highest weights, followed by access cost and construction difficulty, and access distance having the lowest weight). Then, a consistency check (consistency index CI < 0.1) is conducted to verify the rationality of the weight allocation. If the check fails, the weights are readjusted. The third step is to calculate the ranking of the schemes. A weighted summation method is used, which combines the weights of the criteria layer and the scores of each candidate access point on each indicator (such as the higher the access distance, the higher the construction difficulty, and the higher the transformer load rate after access). The comprehensive score of each candidate access point is calculated and sorted from high to low according to the comprehensive score. The one with the highest score is the optimal access point. The fourth step is to generate the optimal access scheme. Based on the power grid topology data of the optimal access point location and the customer's electricity demand, a matching access scheme is generated. Specific operations include: a. Determine the cable laying method. Specifically, this can be determined by combining on-site image data (e.g., underground cables are preferred in urban areas, while overhead lines are used in remote areas) to determine the laying path and conductor cross-section specifications; b. Determine the location of the metering points. The principles of "proximity, ease of meter reading, and safety and compliance" should be followed, taking into account the customer's electricity usage scenario (outdoor meter boxes for residential customers, and workshop or office building entrances for non-residential customers) to determine the specific location of the metering points and the model of the meter. Third, supplementary safety requirements need to be specified. Specifically, line protection measures and grounding methods need to be determined to ensure that the connection scheme complies with Guangdong Power Grid's low-voltage power supply specifications. The fifth step is scheme verification and optimization. The generated access scheme is cross-verified with the power grid operation data of the power grid management platform and the on-site images of China Southern Power Grid Smart View. The AI ​​model makes a second judgment on the compliance of the wiring method and the location of the metering point. If there is room for optimization (such as the laying path can be shortened or the location of the metering point can be more convenient), the scheme is automatically adjusted and the optimal access scheme and detailed description are finally output.

[0093] (7) Generate an access scheme based on the optimal access point. The access scheme includes the location of the access point, the method of laying the line, and the location of the metering point.

[0094] Based on the location of the optimal access point, power grid topology data, and customer electricity demand, the system automatically generates a corresponding access plan. The access plan includes the following data: a. Access point location, used to specify the exact location of the optimal access point; b. The method of laying the line can be determined by combining on-site image data (e.g., underground cables are preferred in urban areas, while overhead lines are used in remote areas) to clarify the laying path and conductor cross-section specifications; c. The location of metering points should follow the principles of "proximity, ease of meter reading, and safety and compliance," and the specific location of the metering points and the model of the meter should be determined in combination with the customer's electricity usage scenario.

[0095] Compared to related technologies that rely on human experience to determine access points, this embodiment has the following advantages: First, decision-making accuracy is significantly improved. Specifically, through deep learning of the power grid topology using GNN and comprehensive analysis of multi-dimensional load characteristics using XGBoost, the complex relationships and dynamic operating status of the power grid can be accurately captured, avoiding the subjective bias of human judgment. This greatly improves the rationality of access point selection and the accuracy of capacity assessment.

[0096] Secondly, the power grid's security capabilities have been enhanced. Specifically, the model uses safety indicators such as transformer load rate and line current as core evaluation features, and incorporates power grid safety standards for secondary verification in the feasibility assessment, ensuring that the access scheme always operates within the power grid's safety threshold and effectively avoiding safety hazards caused by improper selection of access points.

[0097] Finally, by comprehensively considering multiple dimensions such as access distance, construction difficulty, cost, safety, and stability through the analytic hierarchy process, optimal decision-making under multiple objectives was achieved, avoiding the overall imbalance caused by optimizing a single indicator. Furthermore, intelligent access point selection reduced the frequency and complexity of manual on-site surveys. Simultaneously, the improved accuracy of decision-making effectively prevented rework in later construction due to unreasonable access point selection, thus lowering overall operating costs.

[0098] In some embodiments, generating survey results based on the access scheme includes: generating a survey report based on the access scheme, power grid data, and electricity demand information; the survey report includes basic customer information, access point analysis results, access scheme details, power grid security assessment, and judgment basis; notifying staff to conduct survey processing based on the survey report; and generating survey results based on the staff's survey confirmation results; the survey results include the final access scheme confirmed by the staff, on-site verification records, and confirmation conclusions.

[0099] First, the system automatically generates a structured survey report based on the access plan (including the optimal access point location, line laying method, and metering point setting location), power grid data (including power grid topology data, real-time load data, and on-site visual image data), and electricity demand information (application capacity, electricity usage nature, etc.).

[0100] The exploration report can be constructed according to a preset template, which includes at least the following: a. Basic customer information, including customer name, electricity address, installation type and capacity, etc.; b. Access point analysis results, including the candidate access point screening process and feasibility assessment details; c. Access scheme details, including core information such as optimal access point location, line laying method, metering point setting, and conductor specifications; d. Power grid safety assessment, including verification results of safety parameters such as transformer load rate and line current after connection; e. Judgment basis, including the core judgment logic of the model output (such as GNN topology analysis conclusions and XGBoost load assessment data).

[0101] Next, the system will automatically generate an exploration report and synchronize it to the mobile marketing app through the end-to-end collaboration module, notifying staff to conduct exploration processing based on the report.

[0102] After receiving the survey report, staff will determine whether on-site verification is necessary based on the complexity of the grid connection plan and the completeness of the power grid data. There are several methods for conducting the survey, including: a. Video Survey Mode. If the access solution is simple and the power grid data is complete (e.g., for urban residential customers with a clear surrounding power grid topology), staff can establish a video connection with the customer via a mobile marketing app, allowing them to remotely view the site conditions in real time. An AI model simultaneously performs image analysis on the video feed to verify that the site conditions match the survey report. Once verified, the access solution is directly confirmed. This mode achieves "zero-site" surveying, significantly reducing survey costs and time.

[0103] b. On-site Survey Mode: If the access scheme is complex or the power grid data is incomplete (e.g., for customers in remote areas with hidden lines), staff need to bring a mobile terminal to the site. The mobile terminal can collect on-site data in real time (such as line parameters and equipment status) and call the access point intelligent judgment model to perform secondary optimization of the access scheme. Once the staff confirms the access scheme in the regenerated survey report, the on-site survey is complete.

[0104] After completing the site survey and confirmation, staff can submit the survey results via the mobile marketing app. The survey results include the following data: a. The final access plan is the final version of the access plan after on-site confirmation or video verification by staff. It may include minor adjustments to the access plan generated by the intelligent judgment model of the access point (such as adjusting the location of the metering point according to the actual situation on site). b. On-site verification records, including on-site photos, video clips, measured data (such as actual route distance, on-site environment photos), as well as key screenshots and AI verification conclusions from the video survey process; c. Confirm the conclusion, including whether the survey is approved or not. If approved, the access plan is confirmed to be effective. If anomalies are found (such as discrepancies between the customer's on-site conditions and the application information, or errors in the power grid data), the anomaly information is uploaded, and the AI ​​processing module automatically generates anomaly handling suggestions (such as "suggest that the customer adjust the list of electrical equipment" or "suggest that the access point be re-selected"). After the processing is completed, the conclusion is reconfirmed.

[0105] This step combines AI-powered intelligent decision-making with human expertise, leveraging the efficiency and accuracy of AI while retaining the flexibility and reliability of on-site verification. For simple scenarios, AI-generated solutions can be directly confirmed through video surveys, significantly reducing the number of on-site inspections. For complex scenarios, on-site verification by humans collaborates with secondary optimization by AI to ensure the feasibility and safety of the solution.

[0106] Compared to related technologies that rely entirely on manual on-site investigation, this embodiment has the following advantages: First, the cost of exploration is significantly reduced: by replacing part of the on-site exploration with video exploration, "zero-site" operation can be achieved for simple scenarios, which significantly reduces travel and time costs. The cost reduction is even more obvious for customers in remote areas.

[0107] Secondly, the exploration cycle is significantly shortened: AI automatically generates exploration reports for staff to use directly, reducing the preparation work before on-site exploration; the video exploration mode enables instant verification without the need for appointments or queuing, shortening the exploration cycle from several days in the traditional mode to hours or even minutes.

[0108] Secondly, the accuracy of decision-making is guaranteed by two factors: the AI ​​model provides accurate preliminary solutions and judgment criteria, which staff then confirm or fine-tune. This approach leverages the efficiency of AI while retaining the professional judgment of human verification, achieving a complementary advantage of intelligence and experience and avoiding the risks that may exist with a single decision-making model.

[0109] Finally, anomaly handling is more efficient: when on-site conditions do not meet expectations, the AI ​​system can generate anomaly handling suggestions in real time to assist staff in making quick decisions. Compared with the traditional model that requires re-surveying or organizing expert discussions, the efficiency of anomaly handling is greatly improved.

[0110] In some embodiments, a construction plan is generated based on the access scheme, material scheduling plan, real-time power grid operation data, and construction team resource data. This includes: acquiring the access scheme, material scheduling plan, real-time power grid operation data, and construction team resource data; determining the construction time based on material delivery time, off-peak power grid load periods, and customer availability time; matching the construction team and personnel responsibilities according to the complexity of the access scheme; generating a standardized construction process based on low-voltage business expansion construction specifications, including the sequence of procedures and the technical requirements of each procedure; verifying the matching degree between construction time and power grid operation status, the matching degree between construction personnel skills and construction difficulty, and the connection between material delivery and construction progress. If conflicts exist, the construction time is automatically adjusted or construction personnel are re-matched to generate a construction plan.

[0111] The system retrieves and integrates the multi-source data required to generate the construction plan. The multi-source data specifically includes: a. Access scheme, which includes the optimal access point location, line laying method (underground cable or overhead line), metering point location, conductor cross-section specifications, etc. b. Material dispatch plan, which includes material delivery time, handover location, material details, etc.; c. Real-time power grid operation data, including line load, equipment maintenance plans, power grid safety thresholds, etc. d. Construction team resource data, which includes information such as the skill level of construction personnel, their on-duty status, and their work areas.

[0112] The system comprehensively considers multiple time constraints and intelligently determines the optimal construction time period. These time constraints include: (1) Material delivery time. The earliest time to start construction shall be the estimated arrival time of materials as specified in the material dispatch plan, to ensure that materials are in place when construction begins; (2) Periods of low grid load. Construction should be carried out during periods with a load factor of ≤50% to avoid affecting the normal power supply of the grid due to operations during construction and to ensure the safe operation of the grid; (3) Customer available time. Extract the available time window from the work order information or confirm it with the customer through the client, respect the customer's time arrangement and improve the customer experience.

[0113] Based on the complexity of the access scheme, the system automatically assigns the optimal construction team using an AI matching algorithm. The assignment logic includes: (1) Complexity assessment. The level of construction difficulty is assessed based on factors such as the line laying method in the access plan (underground cable laying in urban areas is more difficult, while overhead line construction in remote areas is relatively simple) and the working environment; (2) Personnel matching. Based on the skill level of the construction personnel (such as high-voltage operation qualification, line construction experience), on-the-job status and distance to the work area, match the construction team with the corresponding qualifications and the nearest location; (3) Assignment of responsibilities. Clearly define the job responsibilities of each person, including on-site construction command, wiring, meter installation, safety supervision, etc.

[0114] The system automatically generates a standardized construction process based on the Guangdong Power Grid's low-voltage business expansion construction specifications and the details of the access scheme. The process includes the following: (1) Sequence of procedures. Clearly define the order in which each procedure is performed, for example, "cleaning the construction site → laying the line → installing the metering instrument → wiring and debugging → safety inspection"; (2) Technical requirements. The core technical standards for each process shall be marked, such as the turning radius of the line laying, the standard angle for meter installation, and the tightening torque of the wiring terminals.

[0115] The system performs multiple checks on the initially generated construction plan to ensure the coordination of various resources and tasks. These checks include: (1) Verification of the matching degree between construction time and power grid operation status, that is, to check whether the determined construction time avoids the peak load period of the power grid and the equipment maintenance period, so as to prevent the conflict between construction and power grid operation; (2) Verification of the matching degree between construction personnel skills and construction difficulty, that is, confirming whether the skill level of the assigned construction personnel meets the construction difficulty requirements of the access scheme; (3) Verification of the connection between material delivery and construction progress, that is, checking whether the material delivery time is reasonably connected with the construction start time, and whether there is a risk that the construction has started before the materials arrive.

[0116] If any conflicts are detected during the verification process (such as insufficient skills of construction personnel or delays in material delivery), the system will automatically perform optimization and adjustments, such as re-matching construction personnel or adjusting the construction schedule, until all verification items pass. After optimization, a final, standardized construction plan will be generated.

[0117] Compared to the method of manually formulating construction plans in related technologies, this embodiment has the following advantages: First, the scientific nature of the construction plan has been significantly improved: by comprehensively considering multi-dimensional constraints such as material distribution, power grid load, and customer time, the construction time can be intelligently and optimally selected, avoiding time conflicts or inefficiencies caused by single-factor decision-making.

[0118] Secondly, human resource allocation is precise and efficient: based on the intelligent matching of construction difficulty and personnel skills, it ensures "the right person for the right job" and reduces the risk of construction quality problems or safety accidents caused by skill mismatch.

[0119] Secondly, the construction process is standardized and regulated: the sequence of procedures and technical requirements are generated in strict accordance with the low-voltage expansion construction specifications, and the operating standards are solidified into executable instructions, which improves the stability and controllability of construction quality.

[0120] Finally, resource synergy has been greatly enhanced: through multiple verification mechanisms, the close connection between construction time, personnel skills, and material delivery is ensured, avoiding the idle work phenomenon of "people waiting for materials" and "materials waiting for people" in the traditional model, effectively shortening the construction preparation cycle and improving the overall construction efficiency.

[0121] In some embodiments, a metering material demand list is generated based on the survey results and electricity demand information, including: matching meter models with the installed capacity and electricity usage nature included in the final access scheme in the survey results; matching cable or conductor models and lengths with the access point location, line laying method, and conductor cross-sectional specifications included in the final access scheme; matching safety protection materials with the site environmental conditions included in the final access scheme; and integrating the matching results to generate a metering material demand list, which includes the material name, model, specifications, quantity, and purpose.

[0122] The system automatically matches the corresponding meter model based on the applied capacity and electricity usage nature included in the final access plan in the survey results. The applied capacity refers to the total electricity capacity applied for by the customer (in kVA or kW); the electricity usage nature refers to the customer's electricity usage category, which is divided into two main categories: residential customers and non-residential customers. Different categories correspond to different metering methods and electricity pricing policies.

[0123] The system calls the preset meter selection rule library for matching. If the customer is a residential customer and the installed capacity is within the standard range (e.g., below 20kW), a single-phase smart meter is matched; if the customer is a non-residential customer or the installed capacity exceeds the residential standard, a three-phase smart meter is matched. At the same time, the meter protection level is matched based on the location of the metering point (indoor or outdoor) (e.g., outdoor meters need to be rainproof and dustproof).

[0124] The system automatically matches the cable or conductor type and length based on the access point location, line laying method, and conductor cross-sectional specifications included in the final access plan. The access point location refers to the specific coordinates of the optimal access point, used to calculate the laying distance to the customer's electricity address; the line laying method refers to the form of line laying from the access point to the customer's electricity address, including underground cable laying or overhead line laying, with different methods corresponding to different cable types (e.g., armored cables are required for underground work, and insulated conductors are required for overhead work); the conductor cross-sectional specifications refer to the conductor cross-sectional area (unit: square millimeters, mm²) calculated based on the applied capacity and power supply distance.

[0125] Specifically, the system first calculates the laying distance based on the access point location and the customer's electricity address, taking into account a certain margin; then, it matches the corresponding type of cable or wire from the material library according to the conductor cross-section specifications and laying method; finally, it calculates the required length based on the laying distance and generates a cable material list.

[0126] Next, based on the on-site environmental conditions included in the final access solution, safety protection materials are automatically matched. The on-site environmental conditions refer to the installation environment information obtained from the survey results, including special scenarios such as humid areas, outdoor areas, and flammable / explosive areas.

[0127] Then, based on environmental conditions and safety regulations, the system matches appropriate safety protection materials. For example, humid areas require waterproof junction boxes and waterproof tape; outdoor areas require rainproof metering boxes and insulating protective covers; and areas easily accessible to personnel require protective fences and warning signs.

[0128] The system integrates the above matching results to generate a standardized list of required measurement materials. This list includes the following core information: a. Item name, such as "single-phase smart meter", "YJV22 armored cable", "waterproof junction box"; b. Model number, such as "DDZY666", "YJV22-4×35", "Waterproof FJ-01"; c. Specifications, such as "220V, 5(60)A", "35mm², 4-core", "IP65 protection rating"; d. Quantity: The quantity is automatically calculated based on the project requirements, such as 1 electricity meter, 80 meters of cable, and 2 junction boxes; e. Purpose: Briefly describe the specific usage scenario of the material, such as "for new electricity metering installation for customer A", "for laying lines from the access point to the metering point", or "for waterproof protection of outdoor metering boxes".

[0129] After generating the list of metering material requirements, the system can further invoke compliance verification rules to check whether the selected materials comply with the Guangdong Power Grid's low-voltage power supply specifications and match the access plan and power demand, correcting any selection deviations. Simultaneously, it can leverage material inventory data to prioritize models with sufficient stock, reducing the procurement cycle.

[0130] Compared to the method of manually compiling material lists in related technologies, this embodiment has the following advantages: First, the accuracy of selection is significantly improved: intelligent matching is carried out based on key parameters in the access plan (installation capacity, power consumption nature, laying method, environmental conditions), avoiding problems such as model errors and specification deviations caused by insufficient experience or negligence during manual selection, and ensuring that the materials accurately correspond to the actual needs.

[0131] Secondly, material waste is effectively reduced: By automatically calculating the laying distance and matching the precise length by the access point location, the cable waste caused by "more spares and less use" in the traditional mode is reduced, and the material procurement cost is reduced.

[0132] Secondly, proactive safety protection: Safety protection materials are included as an essential component of the list and are automatically matched according to the site conditions, avoiding construction safety hazards caused by the omission of protective materials in the traditional model.

[0133] Finally, the efficiency of list compilation has been greatly improved: the time spent on manual item-by-item querying, filling in and verifying has been reduced to seconds of automatic generation, which significantly improves the efficiency of the material preparation process and shortens the overall business cycle.

[0134] In some embodiments, matching inventory materials with a metered material demand list to generate a material scheduling plan includes: querying inventory data from the metered material platform based on the metered material demand list; if the query result indicates sufficient inventory, generating a material outbound order, determining the transportation method, transportation route, and handover point, and generating a material scheduling plan; if the query result indicates insufficient inventory, triggering the material procurement process and generating a material scheduling plan including the estimated arrival time; the material scheduling plan includes material details, stocking information, transportation handover information, and contingency plans for handling abnormalities.

[0135] The system first sends an inventory query request to the metrology materials platform based on the generated demand list of metrology materials (including material name, model, specifications, quantity, etc.). The metrology materials platform is an information system for unified management of metrology materials inventory, storing information such as the current inventory quantity, storage warehouse location, and inventory status (normal, pending inspection, scrapped) for each material model.

[0136] The system compares each item in the material demand list with the inventory data to obtain the available inventory of each item.

[0137] If the query results show that all materials are in sufficient stock (i.e., available stock quantity ≥ required quantity), the system will automatically perform the following operations: (1) Generate material outbound order, that is, automatically generate a standardized material outbound order based on the demand list and inventory location, determine the name, model, specifications, quantity, storage warehouse, outbound warehouse location and other information of the material to be outbound, and push it to the warehouse management system. Warehouse staff prepare goods according to the outbound order. (2) Determine the mode of transportation, that is, match the optimal mode of transportation based on the characteristics of the materials (such as electric meters being precision instruments and cables being heavy materials), transportation distance, and time requirements (such as using new energy trucks for delivery in urban areas and dedicated logistics lines for remote areas). (3) Determine the transportation route, that is, plan the optimal transportation route based on factors such as traffic conditions and road restrictions to ensure that the materials are delivered on time; (4) Determine the handover point, that is, determine the material handover point (such as the construction site or the customer's designated receiving point) based on the work order information and the customer's electricity address, as well as determine the handover time window and acceptance requirements; (5) Generate a material scheduling plan, that is, integrate the above information to generate a complete material scheduling plan, specifying the material details, stock preparation information (including existing inventory quantity, storage warehouse location, and stock preparation completion time limit), transportation handover information (including transportation mode, transportation route, handover location, handover time window, and acceptance requirements), and contact information of the responsible parties in each link.

[0138] If the query results show that some or all materials are in short supply (i.e., available inventory < required quantity), the system will automatically perform the following operations: (1) Trigger the material procurement process. For materials with insufficient inventory, the system automatically generates a purchase requisition form, determines the name, model, specifications, quantity, and required date of the materials to be purchased, and pushes it to the material procurement department or platform to start the procurement process. (2) Generate a scheduling plan with estimated delivery time. That is, the system estimates the delivery time of the purchased materials (such as "expected delivery in X days") based on information such as historical procurement cycles and supplier supply capacity, and incorporates it into the material scheduling plan. (3) Synchronously generate emergency response plans. In the event of insufficient inventory, the system automatically generates emergency response plans, determines emergency measures (such as prioritizing allocation from nearby warehouses, coordinating with suppliers to expedite delivery, etc.), and pushes them to relevant staff to ensure that there is a basis for responding to shortages of materials.

[0139] Regardless of whether inventory is sufficient, the generated material dispatch plan must include at least the following core information: a. Material details, including material name, model, specifications, quantity, unit, and specific purpose; b. Inventory information, including the current inventory quantity of each material, the location of the storage warehouse, and the deadline for completing the inventory preparation. If there is insufficient inventory, the materials to be purchased and the estimated arrival time will be marked. c. Transportation handover information, including transportation mode, transportation route, estimated transportation time, as well as recipient, handover location, handover time window and acceptance requirements; d. Contingency plans for handling abnormal situations, including emergency measures for situations such as insufficient inventory and transportation delays, to ensure that the construction progress is not affected by the supply of materials.

[0140] Compared to the methods of manually checking inventory and manually formulating material scheduling plans in related technologies, this embodiment has the following advantages: First, scheduling efficiency is greatly improved: the system automatically completes the entire process of inventory query, outbound order generation, and transportation plan planning, reducing the scheduling work that would normally take tens of minutes or even hours to be completed in seconds, significantly improving the processing efficiency of the material scheduling process.

[0141] Secondly, the inventory status is controllable in real time: through real-time data interaction between the system and the material measurement platform, the scheduling plan is ensured to be based on the latest inventory status, avoiding scheduling errors caused by "inventory has changed but not synchronized" that may occur during manual queries.

[0142] Secondly, the ability to handle anomalies has been enhanced: In the event of insufficient inventory, the system automatically triggers the procurement process and generates the estimated delivery time, while also providing an emergency response plan to ensure a rapid response when materials are in short supply. Compared with the traditional method of "manually requesting procurement after discovering a shortage", the timeliness of handling anomalies has been greatly improved.

[0143] Finally, construction support capabilities have been improved: through precise material scheduling plans, it is ensured that materials arrive in a timely manner before construction, avoiding the idle work phenomenon of "materials not arriving and construction waiting" in the traditional model, and ensuring the smooth execution of the construction plan.

[0144] It should be noted that, regarding the various steps included in the intelligent low-voltage industrial expansion, installation, capacity increase / decrease method provided in any of the above embodiments, unless explicitly stated herein, there is no strict order restriction on the execution of these steps; these steps can be executed in other orders. Moreover, at least some of these steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0145] Based on the same inventive concept, this application also provides an intelligent low-voltage industrial expansion and capacity reduction device. In some embodiments, such as Figure 5 As shown, the intelligent low-voltage industrial expansion and capacity reduction device includes the following modules: The intelligent review module 110 is used to receive electricity application materials and electricity demand information sent by the client, conduct multi-dimensional review of the electricity application materials and electricity demand information, and generate a business work order after the review is approved. The access scheme and survey result generation module 120 is used to respond to business work orders, obtain power grid data based on the customer's electricity address, determine the access scheme based on electricity demand information and power grid data, and generate survey results based on the access scheme. The construction control module 130 is used to generate a metering material demand list based on the survey results and electricity demand information, match the inventory materials based on the metering material demand list to generate a material scheduling plan, generate a construction plan based on the access scheme, material scheduling plan, real-time power grid operation data and construction team resource data, collect construction data and perform real-time verification during the construction process based on the construction plan, review the construction completion report when it is received, and execute the power-on operation after the review is passed. The intelligent archiving module 140 is used to collect all process data for intelligent verification in response to the completion of the power-on operation. After the verification is successful, electronic archives are generated and stored.

[0146] In some embodiments, the intelligent review module 110 is specifically used for: performing text recognition on image data in electricity application materials to obtain text information within the image; inputting the text information within the image, along with text data and electricity demand information in the electricity application materials, into a semantic understanding model for semantic encoding and keyword extraction to obtain text features; inputting image data in the electricity application materials into an image recognition model to extract image features; fusing text features and image features, and performing integrity review, compliance review, and authenticity review based on the fused features and a preset integrity rule base, compliance rule base, and authenticity rule base.

[0147] In some embodiments, the access scheme and survey result generation module 120 is specifically used for: constructing a graph structure from the power grid topology data in the power grid data, wherein the nodes in the graph structure include transformer nodes, existing access point nodes, and customer electricity address nodes, and the edges in the graph structure represent connecting lines, with the edge weights assigned based on the line length and line loss coefficient; performing node embedding processing on the graph structure using a graph neural network algorithm to learn the feature representation of each node, and performing graph traversal processing based on the customer electricity address nodes to filter out candidate access points; obtaining standardized feature data for each candidate access point; the standardized feature data includes the remaining capacity of the access point, the real-time load rate of the transformer to which it belongs, the real-time current of the line, the line current carrying capacity, and the applied capacity in the electricity demand information. The model uses load fluctuation coefficients during electricity consumption periods and historical load peak data. Standardized characteristic data of each candidate access point are input into the access feasibility verification model to obtain the feasibility prediction probability of each candidate access point. The feasibility prediction probability of each candidate access point is verified according to a preset feasibility probability threshold and power grid safety regulations. Based on the verification results, a feasibility assessment result is generated for each candidate access point, indicating whether it is feasible to access, not feasible, or requires adjustment before access. Based on access distance, construction difficulty, access cost, power grid safety, and load stability indicators, the optimal access point is determined from the candidate access points whose feasibility assessment result indicates feasibility to access, using the analytic hierarchy process (AHP). An access scheme is generated based on the optimal access point, including the access point location, line laying method, and metering point location.

[0148] In some embodiments, the access scheme and survey result generation module 120 is specifically used to: generate a survey report based on the access scheme, power grid data and electricity demand information, the survey report including basic customer information, access point analysis results, access scheme details, power grid security assessment and judgment basis; notify staff to conduct survey processing based on the survey report, and generate survey results based on the staff's survey confirmation results; the survey results include the final access scheme confirmed by the staff, on-site verification records and confirmation conclusions.

[0149] In some embodiments, the construction management module 130 is specifically used for: acquiring access schemes, material scheduling plans, real-time power grid operation data, and construction team resource data; determining construction time based on material delivery time, power grid off-peak periods, and customer availability time; matching construction teams and personnel responsibilities based on the complexity of the access scheme; generating standardized construction processes according to low-voltage business expansion construction specifications, including the sequence of procedures and the technical requirements of each procedure; verifying the matching degree between construction time and power grid operation status, the matching degree between construction personnel skills and construction difficulty, and the connection between material delivery and construction progress; and automatically adjusting construction time or rematching construction personnel and generating a construction plan if conflicts exist.

[0150] In some embodiments, the construction management module 130 is specifically used to: match the meter model according to the application capacity and power consumption nature included in the final access scheme in the survey results; match the cable or wire model and length according to the access point location, line laying method and conductor cross-section specifications included in the final access scheme; match safety protection materials according to the site environmental conditions included in the final access scheme; and integrate the matching results to generate a metering material demand list, which includes the material name, model, specifications, quantity and purpose.

[0151] In some embodiments, the construction control module 130 is specifically used to: query the inventory data of the metering material platform based on the metering material demand list; if the query result indicates that the inventory is sufficient, generate a material outbound order, determine the transportation method, transportation route and handover point, and generate a material scheduling plan; if the query result indicates that the inventory is insufficient, trigger the material procurement process and generate a material scheduling plan including the estimated arrival time; the material scheduling plan includes material details, stocking information, transportation handover information and anomaly handling plan.

[0152] Specific limitations regarding intelligent low-voltage business expansion and capacity reduction devices can be found in the above-mentioned limitations on intelligent low-voltage business expansion and capacity reduction methods, and will not be repeated here. Each module in the aforementioned intelligent low-voltage business expansion and capacity reduction device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0153] This application also provides a computer device. In some embodiments, the computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it can implement the intelligent low-voltage industrial expansion and capacity reduction method provided in any of the above embodiments.

[0154] In some embodiments, the internal structure diagram of a computer device may be as follows: Figure 6 As shown. The computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores data such as electricity application information and electricity demand information; the specific data stored can also be found in the limitations defined in the above method embodiments. The network interface is used to communicate with external terminals via a network connection. When the computer program is executed by the processor, it implements an intelligent method for expanding, installing, increasing, or reducing capacity in low-voltage industrial applications.

[0155] Those skilled in the art will understand that Figure 6 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0156] This application also provides a computer-readable storage medium, in some embodiments of which a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, it implements the intelligent low-voltage industrial expansion and capacity reduction method provided in any of the above embodiments.

[0157] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0158] Those skilled in the art will understand that implementing all or part of the processes in the above method embodiments can be accomplished by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), Synchlink, DRAM (SLDRAM), memory bus, direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0159] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0160] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. An intelligent method for expanding, increasing, or decreasing the capacity of low-voltage industrial installations, characterized in that: The method includes: Receive electricity application materials and electricity demand information sent by the client, conduct multi-dimensional review of the electricity application materials and electricity demand information, and generate a business work order after the review is approved; In response to the business work order, the system obtains power grid data based on the customer's electricity address, determines an access scheme based on the electricity demand information and the power grid data, and generates survey results based on the access scheme. A metering material demand list is generated based on the survey results and the electricity demand information. Inventory materials are matched with the metering material demand list to generate a material scheduling plan. A construction plan is generated based on the access scheme, the material scheduling plan, real-time power grid operation data, and construction team resource data. Construction data is collected and verified in real time during the construction process based on the construction plan. Upon receiving a construction completion report, the construction completion report is reviewed. After the review is passed, the power-on operation is performed. Upon completion of the power-on operation, the system collects all process data for intelligent verification. Once the verification is successful, an electronic file is generated and stored.

2. The method according to claim 1, characterized in that, The multi-dimensional review of the electricity application materials and the electricity demand information includes: The image data in the electricity application materials are subjected to text recognition to obtain the text information within the image. The text information within the image, the text data in the electricity application materials, and the electricity demand information are input into a semantic understanding model for semantic encoding and keyword extraction to obtain text features. The image data in the electricity application materials are input into an image recognition model to extract image features; The text features and image features are integrated, and integrity audit, compliance audit and authenticity audit are performed based on the integrated features and the preset integrity rule base, compliance rule base and authenticity rule base.

3. The method according to claim 1, characterized in that, The process of determining the access scheme based on the electricity demand information and the power grid data includes: The power grid topology data in the power grid data is constructed into a graph structure. The nodes in the graph structure include transformer nodes, existing access point nodes, and customer electricity address nodes. The edges in the graph structure represent connecting lines, and the weights of the edges are assigned based on the line length and line loss coefficient. The graph structure is processed by a graph neural network algorithm to embed nodes in order to learn the feature representation of each node, and the graph is traversed based on the customer's electricity address node to filter and obtain candidate access points. Obtain standardized feature data for each candidate access point; the standardized feature data includes the remaining capacity of the access point, the real-time load rate of the transformer to which it belongs, the real-time current of the line, the current carrying capacity of the line, as well as the applied capacity, the load fluctuation coefficient during the power consumption period, and the historical load peak data in the power demand information; The standardized feature data of each candidate access point are input into the access feasibility test model to obtain the feasibility prediction probability of each candidate access point. The feasibility prediction probability of each candidate access point is verified according to the preset feasibility probability threshold and power grid safety specifications. Based on the verification results, the feasibility assessment results of each candidate access point are generated. The feasibility assessment results are: can access, cannot access, or need to be adjusted before access. Based on access distance, construction difficulty, access cost, power grid safety and load stability indicators, the optimal access point is determined from the candidate access points that can be accessed according to the feasibility assessment results using the analytic hierarchy process. An access scheme is generated based on the optimal access point, and the access scheme includes the access point location, the line laying method, and the metering point setting location.

4. The method according to claim 1, characterized in that, The generation of survey results based on the access scheme includes: A survey report is generated based on the access scheme, the power grid data, and the electricity demand information. The survey report includes basic customer information, access point analysis results, access scheme details, power grid security assessment, and judgment basis. The staff are notified to conduct an investigation based on the survey report, and a survey result is generated based on the staff's survey confirmation results. The survey result includes the final access plan confirmed by the staff, on-site verification records, and confirmation conclusions.

5. The method according to claim 1, characterized in that, The process of generating a construction plan based on the access scheme, the material scheduling plan, real-time power grid operation data, and construction team resource data includes: Obtain the access scheme, the material scheduling plan, real-time power grid operation data, and construction team resource data; The construction time is determined based on the material delivery time, the off-peak period of the power grid load, and the available time for customers; Match the construction team and personnel responsibilities according to the complexity of the access scheme; A standardized construction process is generated based on the low-voltage business expansion construction specifications. The construction process includes the sequence of procedures and the technical requirements for each procedure. The system verifies the matching degree between the construction time and the power grid operation status, the matching degree between the construction personnel skills and the construction difficulty, and the connection between material delivery and construction progress. If there is a conflict, the construction time is automatically adjusted or the construction personnel are rematched to generate a construction plan.

6. The method according to claim 4, characterized in that, The step of generating a metering material demand list based on the survey results and the electricity demand information includes: The meter model should be matched with the installed capacity and electricity usage characteristics included in the final access scheme in the survey results. Match the cable or conductor type and length according to the access point location, line laying method and conductor cross-section specifications included in the final access scheme; Safety protection materials are matched according to the on-site environmental conditions included in the final access solution; The integrated matching results generate a list of measurement material requirements, which includes the material name, model, specifications, quantity, and purpose.

7. The method according to claim 6, characterized in that, The step of matching inventory materials with the measured material demand list to generate a material scheduling plan includes: Query the inventory data of the measurement materials platform according to the aforementioned measurement materials demand list; If the query results indicate that the inventory is sufficient, a material outbound order is generated, and the transportation method, transportation route and handover point are determined, and a material scheduling plan is generated. If the query results indicate insufficient inventory, the material procurement process is triggered, and a material scheduling plan including the estimated arrival time is generated. The material scheduling plan includes material details, stocking information, transportation handover information, and contingency plans for handling abnormalities.

8. An intelligent low-voltage industrial expansion and capacity reduction device, characterized in that, The device includes: The intelligent review module is used to receive electricity application materials and electricity demand information sent by the client, conduct multi-dimensional review of the electricity application materials and electricity demand information, and generate a business work order after the review is approved; The access scheme and survey result generation module is used to respond to the business work order, obtain power grid data based on the customer's electricity address, determine the access scheme based on the electricity demand information and the power grid data, and generate survey results based on the access scheme. The construction control module is used to generate a metering material demand list based on the survey results and the electricity demand information, match the inventory materials according to the metering material demand list to generate a material scheduling plan, generate a construction plan based on the access scheme, the material scheduling plan, real-time power grid operation data and construction team resource data, collect construction data and perform real-time verification during the construction process based on the construction plan, review the construction completion report when a construction completion report is received, and execute the power-on operation after the review is passed. The intelligent archiving module is used to collect all process data for intelligent verification in response to the completion of the power-on operation. After the verification is successful, electronic archives are generated and stored.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method according to any one of claims 1 to 7.

Images