Distribution network topology panoramic perception method, system and medium based on AI technology

Abstract

The present application relates to a method, system and medium for panoramic perception of distribution network topology based on AI technology. The method includes collecting and integrating line common information model (CIM) model files and equipment inventory information in the distribution network homology system; performing structural analysis and field correction on the collected data; extracting the main feeder path and dividing the main line into segments, while identifying the positions of key control switches and segment boundaries; supporting branch hierarchical management and modeling; the graph model can be collapsed or expanded for display; realizing closed-loop splicing and overall connectivity analysis of multiple lines; realizing panoramic perception of distribution network; identifying problems and in-depth analysis of the non-standardized structure of the current graph model; realizing multi-dimensional feature display of the graph model and panoramic view output. The present application realizes automatic recognition and optimization of the distribution network topology structure, significantly improves the accuracy of the graph model, generation efficiency and business adaptability, and provides unified support for digital operation and maintenance and intelligent management of the distribution network.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent construction of distribution network graph models in power systems and their graphical expression in multiple business scenarios, and in particular to a method, system and medium for panoramic perception of distribution network topology based on AI technology, which is suitable for multi-source heterogeneous data fusion, automatic identification of topological structure, construction of main and branch line models, mining of operating parameters and visualization output of graph models in scenarios such as planning, scheduling, and operation and maintenance of distribution networks. Background Art

[0002] The current distribution network structure continues to grow in complexity, with trunk lines featuring multi-source feeders and interconnected grids, extensive branch lines, and flexible substation access. The topology is becoming increasingly atypical and diverse. Distribution network equipment and wiring relationships are often stored in CIM models, diagram images, and text-based equipment ledgers. These data sources are heterogeneous, the structure is inconsistent, and there are redundant device identifiers and non-standard field representations. This data suffers from issues such as ambiguous naming, missing attributes, and abnormal connection relationships. This leads to significant discrepancies and low accuracy in topology recognition results based on diagrams. Traditional diagram generation relies on manual diagram recognition and routing, requiring manual symbol splicing and path annotation. This is inefficient and error-prone, lacking a topology logic verification mechanism, and the inclusion of numerous non-critical devices such as disconnectors, fuses, and conductor segments in the diagrams compromises structural clarity. Existing methods lack the ability to intelligently identify trunk paths, branch connections, and device segments, and lack the ability to automatically extract key operating indicators. This makes it difficult to adapt to the diverse diagram requirements of various scenarios, such as planning, scheduling, and operation and maintenance. This limits the business adaptability and sustainability of distribution network diagrams, hindering the advancement of intelligent distribution networks. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a method, system and medium for panoramic perception of distribution network topology based on AI technology, intelligently generate targeted topology optimization suggestions; build a structured graph data model, support multi-view output and cross-business system calls, and significantly improve the accuracy, uniformity and business adaptability of graph construction.

[0004] To achieve the above objectives, this application provides the following technical solutions:

[0005] In a first aspect, an embodiment of the present application provides a method for panoramic perception of distribution network topology based on AI technology, the method comprising the following steps:

[0006] Step 1: Collect and integrate the line common information model (CIM) model files and equipment ledger information in the distribution network homologous system. By building a unified data interface standard, we can achieve synchronous extraction and analysis of multi-source heterogeneous data and provide a logical association basis for equipment for subsequent graph structure construction.

[0007] Step 2: Based on the AI ​​big model and CIM standard specifications, the collected data is structurally analyzed and fields are corrected to extract key structural elements including device connection relationships, switch types, substation transformer attributes, voltage levels, and connected users, generating a data set with logical consistency.

[0008] Step 3: Build the topology of the distribution network diagram based on the connection relationship. Combined with the interconnection switch equipment ledger, a depth-first traversal algorithm is used to extract the main feeder path and divide the main line into sections. At the same time, the key control switch positions and section boundaries are marked.

[0009] Step 4: Extract branch connection points from the mainline topology, cluster and classify the branch structures based on path depth and load capacity indicators, and evaluate their electrical independence and topological influence to support branch hierarchical management and modeling;

[0010] Step 5: Based on the identification and subordinate relationships of the switchgear within the station, the internal topology of the switchgear, box-type transformer, and ring main unit is merged and simplified into virtual aggregate nodes to form a structured diagram. Statistics on line segment capacity, switchgear, user connection relationships, and connection capacity are simultaneously presented, and the diagram can be collapsed or expanded to provide a unified topology foundation for multiple business scenarios.

[0011] Step 6: After completing the single-line structure, based on the tie-switch ledger information and feeder trunk topology, identify the logical connection relationship between multiple feeders and construct a feeder ring structure with the tie-switch as the boundary to achieve closed-loop connection of multiple lines and overall connectivity analysis. Integrate typical daily measurement data to achieve perception of load distribution and heavy overload conditions, and achieve panoramic perception of the distribution network.

[0012] Step 7: Combining the topology modeling results with main and branch path information, the AI ​​large model is used to learn distribution network planning guidelines and typical design cases. This allows for problem identification and in-depth analysis of the current non-standardized structure of the model, assisting in determining the rationality and economic efficiency of equipment layout. Combining a knowledge base of historical transformation cases with typical daily measurement data, targeted topology optimization recommendations are intelligently generated to support intelligent optimization of trunk paths and assist in decision-making for transformation.

[0013] Step 8. The topology model and the measurement data are integrated and encapsulated into an interactive graphic data model, which supports customized reading of line structure and operation parameters according to different professional needs, realizes multi-dimensional feature display and panoramic view output of the graphic model, and supports downstream system linkage calls through interfaces.

[0014] The step 2 further comprises:

[0015] Step 2.1: Extract elements from equipment labels, diagram annotations, and ledger fields using the named entity recognition algorithm, use the equipment type classifier for normalized identification and standard mapping, and unify the coding rules;

[0016] Step 2.2: Build a multidimensional tag set for device attributes, decouple the structure and function of circuit breakers, disconnectors, busbars, transformers, fuses, and conductors, and classify their functions. Aggregate and simplify uncontrollable devices based on production needs.

[0017] Step 2.3: Combine the AI ​​big model with expert rules to perform reasoning, completion, and confidence scoring for missing fields, outliers, and ambiguous naming, and output a data set with a complete closed-loop structure to ensure the uniqueness and correctness of devices and connection points during the topology modeling process.

[0018] The implementation of step 3 can be specifically divided into the following steps:

[0019] Step 3.1: Using the substation outgoing line switch as the starting point of the graph search and the tie switch in the area as the end point, a graph neural network and a heuristic path selection algorithm are used to extract all possible mainline candidate paths;

[0020] Step 3.2: Automatically identify the logical control segments within the main line through the topological partitioning algorithm, and mark the control device nodes at the beginning and end of each segment for subsequent simulation control, load isolation logic analysis and diagram annotation.

[0021] The implementation of step 4 can be specifically divided into the following steps:

[0022] Step 4.1: After removing the devices and edge sets from the trunk path, perform a depth-first traversal in the residual subgraph starting from the trunk node to extract the reachable branch paths and establish the branch graph structure;

[0023] Step 4.2: Count the path depth, number of devices, and cumulative capacity of each branch, build a branch load distribution model, and divide the branches into large branches and small branches based on heuristic weights;

[0024] Step 4.3: Identify the substation transformer and boundary switch at the end of the branch, extract their connection capacity, operation attributes and node identification, generate branch structure summary information, and automatically align it with the topology model.

[0025] The implementation of step 5 can be specifically divided into the following steps:

[0026] Step 5.1: Identify the station building relationship of the equipment based on the equipment name, type identification and location information, and establish a station building-sub-equipment mapping table;

[0027] Step 5.2: Perform node aggregation processing on all device connection nodes in the same station building, unify multiple internal nodes into a virtual central node for topology model expression and scaling presentation;

[0028] Step 5.3: Under the premise of ensuring the logical connectivity between the station building and external equipment, adopt the "black box" modeling method to retain the external interface relationship and improve the clarity and interactivity of the model;

[0029] Step 5.4: Count the capacity of each segment of the trunk, the length of the path, and the distribution of contact points, and extract the structural relationship of important nodes on the trunk path;

[0030] Step 5.5: Extract branch line connection depth, connected users, and branch load distribution to construct a complete load topology distribution map;

[0031] Step 5.6: Summarize the operating attributes of each transformer in each substation, including voltage level, wiring method, total installed capacity and operating status, to form a structured information list.

[0032] The image output method of step 6 includes:

[0033] Step 6.1: Based on the trunk path structure and tie switch ledger information, identify cross-line connection nodes and construct a feeder node mapping relationship and splicing boundary list;

[0034] Step 6.2: Using the tie switch as the logical connection point, a graph structure expansion and connection algorithm is used to generate the feeder ring topology, supporting closed-loop modeling and interconnection verification of multiple lines.

[0035] Step 6.3: Load typical daily measurement data and associate current, voltage, and load fields with topological nodes and paths to enhance the topology's ability to perceive actual operating status.

[0036] Step 6.4: Combine the structural splicing with typical daily measurement information to complete a comprehensive assessment of the connectivity of the panoramic topology and the integrity of the load hanging machine, and output the structurally consistent feeder loop analysis results.

[0037] The implementation of step 7 can be specifically divided into the following steps:

[0038] Step 7.1: Use the AI ​​model to learn the knowledge base of distribution network planning and design guidelines, construction specifications, and historical transformation plans to extract key rules for path identification, control node layout, and structural compliance;

[0039] Step 7.2: Based on the existing backbone structure, combine path connectivity analysis with typical daily load data to identify non-standard construction issues and operational bottlenecks, and locate abnormal fields and connection defects;

[0040] Step 7.3: Based on the diagnostic results and the matching rules of the standard guidelines, the abnormal paths are structurally modified to form a trunk path structure that meets the planning specifications;

[0041] Step 7.4: Combine typical load evolution trends with existing line capacity structures to generate structured topology optimization recommendations and propose guiding expansion, segmentation, or switching optimization strategies to support intelligent transformation and assist decision-making.

[0042] The output results of step 7.5 can be directly embedded in the graphics model system and visualized in multiple business views to support scheduling strategy preview and solution deployment verification.

[0043] The image output method of step 8 includes:

[0044] Step 8.1: Integrate the topology structure with typical daily measurement data to construct an interactive data model that includes node attributes, path status, and time series characteristics;

[0045] Step 8.2: Configure the diagram parameter view template based on the business needs of operation, planning, and scheduling, and support customized display of line structure, capacity segmentation, and load density elements;

[0046] Step 8.3: Encapsulate the graph model output results into a standardized interface data format to support downstream system calls and dynamic view linkage refresh;

[0047] Step 8.4: The model structure supports traceability and structural change backtracking mechanisms to ensure consistent expression and timely updates of model views under different business requirements.

[0048] In the second aspect, an embodiment of the present application provides a distribution network topology panoramic perception system based on AI technology, including a memory and a processor, wherein the memory includes a program of a distribution network topology panoramic perception method based on AI technology, and the program of the distribution network topology panoramic perception method based on AI technology is executed by the processor to implement the above steps.

[0049] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, which stores program code. When the program code is executed by a processor, it implements the steps of the distribution network topology panoramic perception method based on AI technology as described above.

[0050] Compared with the existing technology, the beneficial effects of the present invention are: addressing the problems of heterogeneous distribution network graph data, complex topology identification, inaccurate trunk path extraction, and insufficient multi-business adaptability. This method uses an artificial intelligence model to parse CIM model files and equipment ledgers, identify equipment types and connection relationships, and correct non-standard modeling and attribute missing problems; extracts trunk paths and branch distributions based on graph structure construction, identifies key equipment nodes such as circuit breakers, load switches, and transformers; automatically calculates key operating indicators such as line capacity, connected load, and contact point configuration; combines AI large models to learn distribution network design guidelines, identifies problems and conducts in-depth analysis of the current graph model's non-standardized structure, and intelligently generates targeted topology optimization suggestions; constructs a structured graph model data model, supports multi-view output and cross-business system calls, and significantly improves the accuracy, uniformity, and business adaptability of graph model construction. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0052] Figure 1 It is a flow chart of the method provided by an embodiment of the present invention.

[0053] Figure 2 It is a single-line diagram of a homologous system of an actual distribution network line.

[0054] Figure 3 This is a simplified single-line diagram of a distribution network provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings. It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings.

[0056] The terms "comprises," "comprising," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not preclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0057] The terms "first," "second," etc. are only used to distinguish one entity or operation from another entity or operation, and are not to be understood as indicating or implying relative importance, nor are they to be understood as requiring or implying any actual relationship or order between these entities or operations.

[0058] This embodiment is implemented through the following technical solutions: a method for panoramic perception of distribution network topology based on AI technology, the main implementation process is as shown in the attached Figure 1 As shown, the following steps are included:

[0059] S1. Collect and integrate line Common Information Model (CIM) files and equipment inventory data from the distribution network homology system to establish a unified data interface standard. Through interface calls, extract elements such as feeder number, equipment type, connection relationship, and voltage level, complete the synchronous extraction and formatting of heterogeneous data, and provide structured data support for topology construction.

[0060] S2. Perform structured analysis of collected data based on AI big models and rule bases, identify device attributes and their connection relationships, and correct problems such as field inconsistencies, missing device information, and modeling errors.

[0061] S2.1. Perform entity recognition and classification normalization on the extracted data, classifying circuit breakers and load switches as operational equipment, transformers as power supply equipment, and uncontrollable nodes such as busbars, fuses, and disconnectors as auxiliary equipment. Use rule-based models to merge similar objects and reduce modeling redundancy.

[0062] S2.2: To address issues such as missing fields and ambiguous naming, we call on the AI ​​model to perform semantic reasoning and content completion, perform association verification based on device context and topological location, and output a structured closed-loop dataset with confidence scores.

[0063] S3. Build the distribution network topology based on device connection relationships. Model devices as edges and connection nodes as points to construct an undirected graph model G. Using a depth-first traversal algorithm, starting with the substation outgoing equipment, recursively generate a network-wide path set based on the connection relationships and extract the main feeder path.

[0064] S3.1. Using the substation outgoing line switch as the starting point for graph search and the tie switch within the area as the target end point, a set of candidate mainline paths is extracted by combining a graph neural network with a heuristic path algorithm. The paths record the start and end nodes, control nodes, and path characteristic indicators.

[0065] S3.2. Perform topological partition analysis on the extracted path to identify control sections on the main line. Each section is bounded by two control devices, and the first and last device nodes of the section are marked to ensure consistency between the path control logic and the diagram structure annotation.

[0066] S4. After removing the edges corresponding to the main line equipment, perform a depth-first traversal in the residual undirected graph G' to extract the paths reachable from the trunk nodes and form an initial branch set.

[0067] S4.1. Remove all mainline devices from the trunk path edge set. In the residual subgraph G', perform a depth-first search starting from the trunk node to identify reachable branch paths.

[0068] S4.2. Count the structural characteristics of each branch, including path depth (number of hops), number of devices involved, and connection capacity, and establish a branch attribute matrix;

[0069] S4.3. Set clustering rules: When a branch meets any of the following conditions—path depth ≥ 3, connected capacity ≥ 2000kVA, number of connected transformers ≥ 3—it is classified as a “large branch,” and the rest are “small branches.”

[0070] S4.4. Mark the substation transformer and boundary switch nodes connected to the branch end, extract their capacity, operating status and node number information, and automatically bind them to the diagram model nodes;

[0071] S4.5. Build a branch index and structure summary to generate a multi-level branch topology description, providing a data foundation for subsequent graph visualization organization and business-side branch strategy recommendations;

[0072] S5. Based on the relationship between equipment identification attributes and spatial distribution, identify the station building range and perform equipment aggregation processing within the station to achieve structural abstraction and topology clarification.

[0073] S5.1. Create a "station-sub-equipment" mapping list ZF_list based on equipment naming rules, type identifiers (such as "switch station," "box-type transformer," "ring main unit," etc.), and spatial location information to group equipment within the station.

[0074] S5.2. Analyze the connection nodes within each station building: If the node is only connected to equipment within the station building, it is marked as a "virtual internal node"; if it is connected to cross-station building or backbone equipment, it is marked as an "external interface node";

[0075] S5.3. Aggregate all “virtual internal nodes” into a station building central node and display it in the diagram as a “virtual aggregate node” to uniformly express the internal structure of the station building;

[0076] S5.4. Update the main diagram topology. All devices in the same station building maintain logical connectivity with the outside world through this virtual node, achieving a black-box representation of the structure.

[0077] S5.5. The station building can be folded or unfolded in the model, supporting simultaneous display of key equipment layout and load attachment views, improving model clarity and interactivity;

[0078] S5.6. Extract the control node distribution, segment length, voltage level, and connection capacity between trunk path segments, construct a "mainline segment capacity model," and analyze path connectivity and operational rationality.

[0079] S5.7. Establish a connection topology mapping table for each branch line, calculate the branch line length, number of equipment, total capacity of connected transformers, typical connected load, etc., mark the large / small branch types, and calculate the load distribution;

[0080] S5.8. Integrate the voltage level, connection mode, operating status, total capacity and other operating attributes of each transformer in each substation, output a structured summary table, and form a structured information list;

[0081] S6. After the single-line structure modeling is completed, the interconnection switch ledger and trunk path information are combined to identify the logical splicing boundaries between multiple feeders and build a panoramic topology network with a closed structure.

[0082] S6.1. Extract the connection nodes corresponding to all tie switches from the topology diagram, establish a feeder-feeder connection map, mark the starting and intersection locations of each feeder, and generate a logical splicing list;

[0083] S6.2. Using the tie switch as the topological splicing point, use the graph structure augmentation algorithm to expand the boundary nodes of each feeder to construct a "feeder ring" structure. Perform loop integrity and node connectivity checks to ensure the overall topology is closed.

[0084] S6.3. Access typical daily operating measurement data (such as voltage, current, and power), align and map it to topological nodes and path segments, and form a topological graph model with temporal characteristics.

[0085] S6.4. Combine the spliced ​​structure with typical daily measurement data to perform a full-network connectivity analysis on the topology diagram, assessing the integrity of path load coverage, node power supply independence, and the rationality of tie-switch redundancy arrangements, and output a consistent, panoramic feeder topology view.

[0086] S7. After the trunk and branch paths are extracted, the distribution network planning standard guidelines and typical design cases are learned based on the AI ​​big model, and multiple paths are evaluated and ranked.

[0087] S7.1. Call the AI ​​big model to load the distribution network design guidelines, standard construction specifications, and historical transformation knowledge base, extract control rules such as structure identification, node configuration, and topology compliance judgment, and build a trunk path compliance analysis framework;

[0088] S7.2. Based on the existing diagram model, combine path connectivity, node density, typical load distribution and measurement status to locate non-standard construction problems and abnormal problems such as jump point disconnection in the structure;

[0089] S7.3. Perform structural correction operations on the identified problem paths, referencing standard specifications and expert rules, automatically recommending a structurally compliant line topology, and completing the diagram model update;

[0090] S7.4. Generate structural topology optimization recommendations based on load evolution trends, equipment operating status, and capacity layout, including typical strategies such as output expansion, switch reconfiguration, and path rerouting, to support future distribution network construction planning.

[0091] S7.5. The above suggestions and optimized structures are embedded in the graphical model system, supporting on-demand calling, visual display, and strategy verification in multiple business views, thus realizing a closed-loop decision-making process from diagnosis to deployment.

[0092] S8. Integrate and encapsulate the topology structure and measurement data to build an interactive graph data model to support the viewing requirements and dynamic linkage of multiple business sides.

[0093] S8.1. Unify the trunk and branch structure, connection relationships, node attributes, and typical daily measurement data to build a graphical data model with path status, node load, and time series characteristics, and support temporal analysis and dynamic attribute refresh;

[0094] S8.2. Load different diagram template configurations based on business needs such as operation, planning, and scheduling, supporting interactive settings such as layer control, node attribute visibility, and indicator filtering to meet differentiated view requirements;

[0095] S8.3. The graph model data results are encapsulated in a unified standardized interface format (such as RESTful API + JSON) for downstream scheduling, operation and maintenance system calls, and support for view cascading and parameter linkage updates;

[0096] S8.4. Support a traceability mechanism for structural changes and data updates. When the model structure, measurement data, or equipment parameters change, the change records can be compared with the history to ensure the consistency of model expression and the controllability of updates.

[0097] Attachment Figure 2 The single-line diagram of the distribution network generated by the traditional homologous system in a certain area is shown, which has problems such as diverse equipment, chaotic structure, and difficulty in reading. Figure 3 This is a simplified diagram of the present invention constructed based on the above steps. Only the main control equipment and typical branch loads are retained in the diagram. The structure is clear, the logic is complete, and it is easy to understand and operate.

[0098] An embodiment of the present application provides a distribution network topology panoramic perception system based on AI technology, including a memory and a processor. The memory includes a program of a distribution network topology panoramic perception method based on AI technology. When the program of the distribution network topology panoramic perception method based on AI technology is executed by the processor, the above steps are implemented.

[0099] An embodiment of the present application provides a computer-readable storage medium, which stores program code. When the program code is executed by a processor, the steps of the distribution network topology panoramic perception method based on AI technology are implemented as described above.

[0100] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0101] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0102] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0103] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0104] In a typical configuration, a computing device includes one or more processors (CPUs), input / output interfaces, network interfaces, and memory.

[0105] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0106] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can be implemented using any method or technology to store information. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change RAM (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory computer-readable media such as modulated data signals and carrier waves.

[0107] The above description is merely an embodiment of the present application and is not intended to limit the scope of protection of the present application. For those skilled in the art, various modifications and variations of the present application are possible. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the scope of protection of the present application.

Claims

1. A method for panoramic perception of distribution network topology based on AI technology, characterized in that: The method comprises the following steps: Step 1: Collect and integrate the line common information model (CIM) model files and equipment ledger information in the distribution network homologous system. By building a unified data interface standard, we can achieve synchronous extraction and analysis of multi-source heterogeneous data and provide a logical association basis for equipment for subsequent graph structure construction. Step 2: Based on the AI ​​big model and CIM standard specifications, the collected data is structurally analyzed and fields are corrected to extract key structural elements including device connection relationships, switch types, substation transformer attributes, voltage levels, and connected users, generating a data set with logical consistency. Step 3: Build the topology of the distribution network diagram based on the connection relationship. Combined with the interconnection switch equipment ledger, a depth-first traversal algorithm is used to extract the main feeder path and divide the main line into sections. At the same time, the key control switch positions and section boundaries are marked. Step 4: Extract branch connection points from the feeder trunk topology, cluster and classify the branch structures based on path depth and load capacity indicators, and evaluate their electrical independence and topological influence to support branch hierarchical management and modeling; Step 5: Based on the identification and subordinate relationships of the switchgear within the station, the internal topology of the switchgear, box-type transformer, and ring main unit is merged and simplified into virtual aggregation nodes to form a structured graph model. This data includes statistics on line segment capacity, switchgear, user connection relationships, and connection capacity, and supports the folding or expanding of the graph model, providing a unified topology foundation for multiple business scenarios. Step 6: After completing the single-line structure, based on the tie-switch ledger information and feeder trunk topology, identify the logical connection relationship between multiple feeders and construct a feeder ring structure with the tie-switch as the boundary to achieve closed-loop connection of multiple lines and overall connectivity analysis. Integrate typical daily measurement data to achieve perception of load distribution and heavy overload conditions, and achieve panoramic perception of the distribution network. Step 7: Combining the topology modeling results with main and branch path information, the AI ​​large model is used to learn distribution network planning guidelines and typical design cases. This allows for problem identification and in-depth analysis of the current non-standardized structure of the model, assisting in determining the rationality and economic efficiency of equipment layout. Combining a knowledge base of historical transformation cases with typical daily measurement data, targeted topology optimization recommendations are intelligently generated to support intelligent optimization of trunk paths and assist in decision-making for transformation. Step 8. The topology model and the measurement data are integrated and encapsulated into an interactive graphic data model, which supports customized reading of line structure and operation parameters according to different professional needs, realizes multi-dimensional feature display and panoramic view output of the graphic model, and supports downstream system linkage calls through interfaces.

2. The method for panoramic perception of distribution network topology based on AI technology according to claim 1 is characterized in that: The step 2 further comprises: Step 2.1: Extract elements from equipment labels, diagram annotations, and ledger fields using the named entity recognition algorithm, use the equipment type classifier for normalized identification and standard mapping, and unify the coding rules; Step 2.2: Build a multidimensional tag set for device attributes, decouple the structure and function of circuit breakers, disconnectors, busbars, transformers, fuses, and conductors, and classify their functions. Aggregate and simplify uncontrollable devices based on production needs. Step 2.3: Combine the AI ​​big model with expert rules to perform reasoning, completion, and confidence scoring for missing fields, outliers, and ambiguous naming, and output a data set with a complete closed-loop structure to ensure the uniqueness and correctness of devices and connection points during the topology modeling process.

3. The method for panoramic perception of distribution network topology based on AI technology according to claim 1, characterized in that: The implementation of step 3 can be specifically divided into the following steps: Step 3.1: Using the substation outgoing line switch as the starting point of the graph search and the tie switch in the area as the end point, a graph neural network and a heuristic path selection algorithm are used to extract all possible mainline candidate paths; Step 3.2: Automatically identify the logical control segments within the main line through the topological partitioning algorithm, and mark the control device nodes at the beginning and end of each segment for subsequent simulation control, load isolation logic analysis and diagram annotation.

4. The method for panoramic perception of distribution network topology based on AI technology according to claim 1, characterized in that: The implementation of step 4 can be specifically divided into the following steps: Step 4.1: After removing the devices and edge sets from the trunk path, perform a depth-first traversal in the residual subgraph starting from the trunk node to extract the reachable branch paths and establish the branch graph structure; Step 4.2: Count the path depth, number of devices, and cumulative capacity of each branch, build a branch load distribution model, and divide the branches into large branches and small branches based on heuristic weights; Step 4.3: Identify the substation transformer and boundary switch at the end of the branch, extract their connection capacity, operation attributes and node identification, generate branch structure summary information, and automatically align it with the topology model.

5. The method for panoramic perception of distribution network topology based on AI technology according to claim 1, characterized in that: The implementation of step 5 can be specifically divided into the following steps: Step 5.1: Identify the station building relationship of the equipment based on the equipment name, type identification and location information, and establish a station building-sub-equipment mapping table; Step 5.2: Perform node aggregation processing on all device connection nodes in the same station building, unify multiple internal nodes into a virtual central node for topology model expression and scaling presentation; Step 5.3: Under the premise of ensuring the logical connectivity between the station building and external equipment, a "black box" modeling method is used to retain the external interface relationship and improve the clarity and interactivity of the model; Step 5.4: Count the capacity of each segment of the trunk, the length of the path, and the distribution of contact points, and extract the structural relationship of important nodes on the trunk path; Step 5.5: Extract branch line connection depth, connected users, and branch load distribution to construct a complete load topology distribution map; Step 5.6: Summarize the operating attributes of each transformer in each substation, including voltage level, wiring method, total installed capacity and operating status, to form a structured information list.

6. The method for panoramic perception of distribution network topology based on AI technology according to claim 1, characterized in that: The image output method of step 6 includes: Step 6.1: Based on the trunk path structure and tie switch ledger information, identify cross-line connection nodes and construct a feeder node mapping relationship and splicing boundary list; Step 6.2: Using the tie switch as the logical connection point, a graph structure expansion and connection algorithm is used to generate the feeder ring topology, supporting closed-loop modeling and interconnection verification of multiple lines. Step 6.3: Load typical daily measurement data and associate current, voltage, and load fields with topological nodes and paths to enhance the topology's ability to perceive actual operating status. Step 6.4: Combine the structural splicing with typical daily measurement information to complete a comprehensive assessment of the connectivity of the panoramic topology and the integrity of the load hanging machine, and output the structurally consistent feeder loop analysis results.

7. The method for panoramic perception of distribution network topology based on AI technology according to claim 1, characterized in that: The implementation of step 7 can be specifically divided into the following steps: Step 7.1: Use the AI ​​model to learn the knowledge base of distribution network planning and design guidelines, construction specifications, and historical transformation plans to extract key rules for path identification, control node layout, and structural compliance; Step 7.2: Based on the existing backbone structure, combine path connectivity analysis with typical daily load data to identify non-standard construction issues and operational bottlenecks, and locate abnormal fields and connection defects; Step 7.3: Based on the diagnostic results and the matching rules of the standard guidelines, the abnormal paths are structurally modified to form a trunk path structure that meets the planning specifications; Step 7.4: Combine typical load evolution trends with existing line capacity structures to generate structured topology optimization recommendations and propose guiding expansion, segmentation, or switching optimization strategies to support intelligent transformation and assist decision-making. Step 7.5: The output results can be directly embedded in the graphics model system and visualized in multiple business views to support scheduling strategy preview and solution deployment verification.

8. The method for panoramic perception of distribution network topology based on AI technology according to claim 1, characterized in that: The image output method of step 8 includes: Step 8.1: Integrate the topology structure with typical daily measurement data to construct an interactive data model that includes node attributes, path status, and time series characteristics; Step 8.2: Configure the diagram parameter view template based on the business needs of operation, planning, and scheduling, and support customized display of line structure, capacity segmentation, and load density elements; Step 8.3: Encapsulate the graph model output results into a standardized interface data format to support downstream system calls and dynamic view linkage refresh; Step 8.4: The model structure supports traceability and structural change backtracking mechanisms to ensure consistent expression and timely updates of model views under different business requirements.

9. A distribution network topology panoramic perception system based on AI technology, characterized by: It includes a memory and a processor, the memory includes a program of a distribution network topology panoramic perception method based on AI technology, and when the program of the distribution network topology panoramic perception method based on AI technology is executed by the processor, the steps of the method according to any one of claims 1 to 8 are implemented.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores program code, and when the program code is executed by the processor, the steps of the distribution network topology panoramic perception method based on AI technology are implemented as described in any one of claims 1 to 8.

Images