Dynamic load transfer visualization method and system for power distribution network

Abstract

The invention discloses a power distribution network dynamic load transfer visualization method and system, and belongs to the technical field of power system digitization. The power distribution network dynamic load transfer visualization method comprises the following steps: extracting basic primitives in a drawing; detecting end points, line segments and space distribution of the basic primitives step by step, combining the end points, the line segments and the space distribution into complete electrical equipment primitives, and building an equipment topology library; establishing a mapping relation between the electrical equipment primitive and the text based on a spatial entropy matching model, and matching the electrical equipment primitive and the text with a model in a machine account file; wherein when the load transfer path generation operation is triggered, the load transfer path is searched according to the equipment topology library and the equipment connection relation, and the load transfer path is dynamically visualized. Discrete primitives in a CAD drawing are intelligently aggregated into an electrical equipment entity, the incidence relation between equipment and character labels and between the equipment and machine account data is established, and finally a dynamic load transfer path is generated.

Description

Technical Field

[0001] This invention relates to the field of digital technology for power systems, and in particular to a method and system for visualizing dynamic load transfer in distribution networks. Background Technology

[0002] In power distribution network operation and maintenance management, CAD drawings are the primary carrier for recording the power grid structure. However, existing technologies suffer from three key shortcomings: First, electrical equipment (such as ring main units, circuit breakers, and substations) in CAD drawings are typically composed of multiple discrete line segments, and existing analytical techniques cannot identify these discrete elements as complete devices. Second, the text annotations on the drawings lack a correlation with the equipment and lines, requiring manual establishment of connections, which is inefficient and prone to errors. Finally, existing CAD-to-HTML conversion technology can only achieve static display and cannot support dynamic load transfer simulation. Related patents (such as CN108764242A) use template matching methods to identify equipment, but the recognition rate drops significantly when faced with non-standard designs; while patent US20210089964 relies on manual intervention to establish connections, resulting in insufficient automation. Furthermore, load transfer simulations typically require pre-set templates (such as JP2020156842A), which cannot adapt to the diverse layouts of real CAD drawings.

[0003] It should be noted that the information disclosed in the background section of this invention is intended only to enhance the understanding of the general background of this invention, and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for visualizing dynamic load transfer in a power distribution network, so as to solve the problem in the process of converting power distribution network CAD drawings into dynamically visible load transfer paths.

[0005] To address the aforementioned technical problems, this invention provides a method for visualizing dynamic load transfer in a distribution network, comprising:

[0006] Extract basic graphic elements from the drawings, detect the distribution of endpoints, line segments, and spaces of the basic graphic elements step by step, and combine them into complete electrical equipment graphic elements to build an equipment topology library;

[0007] The mapping relationship between the electrical equipment graphic elements and text is established based on the spatial entropy matching model, and the electrical equipment graphic elements and text are matched with the model in the ledger archive;

[0008] Specifically, when the load transfer path generation operation is triggered, the load transfer path is searched based on the device topology library and device connection relationship, and the load transfer path is dynamically visualized.

[0009] Preferably, the step-by-step detection of the endpoints, line segments, and spatial distribution of basic graphic elements includes:

[0010] The degree of overlap of the endpoints is detected. If the distance between the endpoints is less than a preset tolerance value, the basic primitives are aggregated into the first connecting component; otherwise, the projection intersection detection of the basic primitives continues.

[0011] Calculate the projection of the line segment in the specified direction. If the intersection of the projections is not less than a preset projection value, then the basic primitives are aggregated into a second connection component; otherwise, continue to perform density clustering detection on the basic primitives.

[0012] Construct a spatial density function to classify devices with a density greater than a preset density threshold as device clusters, and otherwise as independent devices.

[0013] Preferably, establishing the mapping relationship between the electrical equipment primitives and text based on the spatial entropy matching model includes:

[0014] Calculate the normalized spatial distance from the text location to the geometric center of the electrical equipment element, search the text location within a preset radius range, and filter the candidate equipment set for each text.

[0015] The association probability distribution of each candidate device is calculated by combining the area weight of electrical equipment elements, the entropy value is calculated using the association probability distribution, and a mapping relationship is established based on the entropy value.

[0016] Preferably, the normalized spatial distance from the text location to the geometric center of the electrical equipment element is calculated:

[0017] dk=∥Pj-Ck∥ / max_bbox_diagonal

[0018] Where dk is the normalized spatial distance of the k-th device, Pj is the position coordinate of the text Tj, Ck is the geometric center of the device, and max_bbox_diagonal is the global normalization factor, which is the maximum value of the diagonal of the border of all electrical device primitives.

[0019] Preferably, the area weight of the electrical equipment elements is calculated using the following formula:

[0020] Wk=Area(Dk)^α

[0021] Where Wk represents the area weight of the k-th electrical equipment element, Area represents the area of ​​the electrical equipment element, Dk is the k-th candidate equipment, and α is a preset weight strength control value; the association probability distribution is calculated using the area weight of the electrical equipment element:

[0022] pk=[exp(-β·dk)·Wk] / [Σ(m=1 to K)exp(-β·dm)·Wm]

[0023] Where pk represents the association probability of the k-th device, dk represents the normalized distance of the k-th device, β is a preset exponential decay control value, and K is the number of candidate devices; the entropy value is calculated using the association probability:

[0024] H(Tj)=-Σ(k=1 to K)pk log2 pk

[0025] Where Tj represents the j-th text, and H(Tj) represents the disorder of the association distribution of the j-th text.

[0026] Preferably, when establishing a mapping relationship based on entropy values, the minimum... k When (dk) < γ or H(Tj) < θ, bind the text Tj to the nearest device; otherwise, start auxiliary matching, where γ is the minimum distance threshold and θ is the entropy threshold.

[0027] Preferably, when matching the electrical equipment graphic elements and text with the model in the ledger archive, the system first retrieves completely identical equipment names from the ledger database through structured query for mapping; for unmatched equipment, the system calculates similarity and generates a candidate set of equipment with a similarity value greater than a preset value for matching.

[0028] Preferably, the system also includes a dynamic visualization engine for triggering a load transfer path generation operation. After triggering the load transfer path generation operation, the dynamic visualization engine searches the equipment topology library, returns an adjacency matrix, and calculates the optimal load transfer path based on the line capacity constraints provided by the ledger system. The system then renders the entire load transfer path process dynamically.

[0029] This invention also provides a dynamic load transfer visualization system for distribution networks, comprising:

[0030] The parsing engine is used to read drawings and extract basic graphic elements;

[0031] The equipment identification module detects and combines the endpoints, line segments, and spatial distribution of basic graphic elements into complete electrical equipment graphic elements, and builds an equipment topology library.

[0032] The association modeling module establishes a mapping relationship between the electrical equipment graphic elements and text based on the spatial entropy matching model, and matches the electrical equipment graphic elements and text with the model in the ledger archive;

[0033] A dynamic visualization engine is used to trigger the load transfer path generation operation, search for load transfer paths based on the device topology library and device connection relationships, and dynamically visualize the load transfer paths.

[0034] The present invention also provides a dynamic load transfer visualization system for distribution networks, which uses the dynamic load transfer visualization method for distribution networks described above to process drawing data and dynamically visualizes the load transfer path when the load transfer path generation operation is triggered.

[0035] In the dynamic load transfer visualization method for power distribution networks provided by this invention, a three-level aggregation mechanism is used to progressively combine the overall equipment from endpoints, line segments, and spatial distribution to solve the problem of identifying complex equipment. Furthermore, a spatial matching model is used to build a mapping relationship, and dynamic visualization of load transfer paths is realized based on the above mapping relationship and equipment topology library, realizing real-time interaction of "clicking on equipment → automatically generating transfer path", with low response delay and the ability to output multiple transfer schemes in parallel.

[0036] The dynamic load transfer visualization system for distribution networks provided by this invention and the dynamic load transfer visualization method for distribution networks provided by this invention belong to the same inventive concept. Therefore, the dynamic load transfer visualization system for distribution networks provided by this invention has at least all the advantages of the dynamic load transfer visualization method for distribution networks provided by this invention, which will not be elaborated here. By intelligently aggregating discrete elements in CAD drawings into electrical equipment entities and establishing the association between equipment and text annotations and ledger data, an HTML visualization system supporting dynamic load transfer interaction is finally generated. Attached Figure Description

[0037] Those skilled in the art will understand that the accompanying drawings are provided to better understand the invention and do not constitute any limitation on the scope of the invention. Wherein:

[0038] Figure 1 This is an architectural diagram of an embodiment of the present invention;

[0039] Figure 2 This is a flowchart of a three-level primitive aggregation processing according to an embodiment of the present invention;

[0040] Figure 3 This is a logic diagram of a spatial entropy matching model according to an embodiment of the present invention;

[0041] Figure 4 This is a flowchart of a dynamic visualization engine according to an embodiment of the present invention;

[0042] Figure 5 This is a detailed flowchart of an embodiment of the present invention. Detailed Implementation

[0043] To make the objectives, advantages, and features of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the drawings are all in a very simplified form and are not drawn to scale, and are only used to facilitate and clarify the explanation of the embodiments of this invention. Furthermore, the structures shown in the drawings are often part of the actual structures. In particular, different figures may emphasize different aspects and may sometimes use different scales.

[0044] As used in this invention, the singular forms “a,” “an,” and “the” include plural objects; the term “or” is generally used to mean “and / or”; the term “a number” is generally used to mean “at least one”; the term “at least two” is generally used to mean “two or more”; furthermore, the terms “first,” “second,” and “third” are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, features defined with "first," "second," and "third" may explicitly or implicitly include one or at least two of those features. The term "proximal" typically refers to the end closer to the operator, and the term "distal" typically refers to the end closer to the patient. "One end" and "the other end," as well as "proximal" and "distal," generally refer to two corresponding parts, including not only endpoints. The terms "installed," "connected," and "linked" should be interpreted broadly. For example, they can be fixed connections, detachable connections, or integral connections; they can be mechanical connections or electrical connections; they can be direct connections or indirect connections through an intermediate medium; they can be internal connections between two elements or interactions between two elements. Furthermore, as used in this invention, the placement of one element on another element generally only indicates a connection, coupling, cooperation, or transmission relationship between the two elements, and the connection, coupling, cooperation, or transmission between the two elements can be direct or indirect through an intermediate element. It should not be construed as indicating or implying a spatial positional relationship between the two elements, i.e., one element can be located arbitrarily inside, outside, above, below, or to one side of another element, unless otherwise explicitly stated. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0045] This invention aims to solve three core problems in the conversion of CAD drawings for power distribution networks: 1) the inability to automatically aggregate discrete elements into complete equipment, especially complex equipment such as ring main units, circuit breakers, and substations; 2) the lack of an automatic association mechanism between drawing text and equipment / lines; and 3) the inability of static displays to support dynamic load transfer simulation. Through innovative element aggregation algorithms, spatial entropy matching models, and dynamic path generation mechanisms, end-to-end conversion from CAD drawings to an interactive HTML visualization system is achieved, providing decision support for load transfer in power distribution networks.

[0046] This invention solves the problem of complex device identification by progressively combining the entire device from endpoints, line segments, and spatial distribution through a three-level aggregation mechanism. It uses a spatial matching model to build mapping relationships and, based on these mapping relationships and a device topology library, achieves dynamic visualization of load transfer paths. This enables real-time interaction of "clicking on a device → automatically generating a transfer path," with minimal response latency and the ability to output multiple transfer schemes in parallel. The solution provided by this invention helps power distribution network operation and maintenance transition from static drawing management to dynamic digital twins, providing efficient response capabilities for load transfer decisions.

[0047] For details, please refer to Figures 1-5 This is a schematic diagram of an embodiment of the present invention. Figure 1 As shown, a method for visualizing dynamic load transfer in a distribution network includes:

[0048] S1: Extract basic graphic elements from the drawings, detect the distribution of endpoints, line segments, and spaces of the basic graphic elements step by step, and combine them into complete electrical equipment graphic elements to build an equipment topology library.

[0049] In one implementation, the drawing is a DWG / DXF file that can be read by CAD software; however, the format of the drawing is not limited to this, and it can also be other types of processing and parsing drawing files.

[0050] Understandably, the process involves reading DWG / DXF files, extracting basic graphic elements such as lines, polylines, and text, and then combining discrete line segments into electrical equipment entities using a three-level graphic element aggregation algorithm. Addressing the issue of discrete line segments failing to be recognized as complete equipment, this invention proposes a three-level aggregation mechanism, breaking through the limitations of traditional template matching and single-level connection detection. This three-level progressive mechanism adapts to different equipment complexities (simple line segments → busbars, complex clusters → ring main units), and density clustering solves the problem of non-connected graphic element aggregation (such as scattered circuit breaker symbols). The processing flow is as follows: Figure 2 As shown.

[0051] The step-by-step detection of the endpoints, line segments, and spatial distribution of basic graphic elements includes:

[0052] The degree of overlap of the endpoints is detected. If the distance between the endpoints is less than a preset tolerance value, the basic primitives are aggregated into the first connecting component; otherwise, the projection intersection detection of the basic primitives continues.

[0053] Calculate the projection of the line segment in the specified direction. If the intersection of the projections is not less than a preset projection value, then the basic primitives are aggregated into a second connection component; otherwise, continue to perform density clustering detection on the basic primitives.

[0054] Construct a spatial density function to classify devices with a density greater than a preset density threshold as device clusters, and otherwise as independent devices.

[0055] Specifically, a preliminary connection is first made, and endpoint overlap detection is performed. The determination criteria are as follows:

[0056] ||P1-P2||<ε

[0057] ε is a preset tolerance value used to represent the error tolerance or neighborhood radius. P1 and P2 represent the two endpoints. Ideally, the tolerance ε ≤ 0.1 units. A component satisfying the above condition is denoted as the first connecting component, i.e.,... Figure 2 Connect component C1 in the middle. Otherwise, perform secondary connections, projection intersection detection, and calculate the projection domain of the line segments:

[0058] Proj(L1)∩Proj(L2)≥δ

[0059] Where δ is the preset projection value, representing the minimum length requirement for projection overlap; Proj is an abbreviation for "Projection," referring to the projected line segment obtained by projecting a line segment onto a specific direction (or coordinate axis); L1 and L2 represent two line segments; more preferably, δ = 0.05L max If the above conditions are met, it is denoted as the second connecting component, i.e., Figure 2 The connecting component C2 in the middle. Otherwise, perform three-level connectivity, spatial density clustering, and construct the following density function:

[0060]

[0061] Where ρ represents density, and the magnitude of σ determines the range of the density function, similar to a smoothing parameter. When σ is large, the density distribution is smoother, and the interaction distance between points is greater; when σ is small, the density distribution is more concentrated in local areas; ρ thres A preset density threshold is used to determine whether a cluster has formed; when ρ > ρ thres When a device is in a cluster, it forms a device cluster; otherwise, it functions as an independent device. Pi and Pj represent two endpoints.

[0062] Understandably, the above ε, δ, ρ thres Both σ and σ can be provided according to the dimensions in the CAD drawings.

[0063] The partial implementation code is as follows:

[0064]

[0065] Through a three-level primitive aggregation algorithm, the technical bottleneck of traditional CAD analysis, which makes it difficult to identify discrete line segments as whole devices, can be overcome. The accuracy of identifying complex devices such as ring main units and circuit breakers can be greatly improved from 71% to 96.2%, significantly improving the efficiency of drawing analysis.

[0066] S2, establish the mapping relationship between the electrical equipment graphic elements and text based on the spatial entropy matching model, and match the electrical equipment graphic elements and text with the model in the ledger archive.

[0067] The spatial entropy matching model, based on information entropy theory, calculates the normalized spatial distance from the text location to the geometric center of the device and combines it with device area weights to form an association probability distribution, using entropy values ​​to measure the concentration of the distribution. Low entropy indicates a clear association that can be directly bound, while high entropy indicates an ambiguous association requiring auxiliary matching. Decision rules combining distance and entropy thresholds achieve efficient and robust matching. It can automatically bind CAD text labels (such as device numbers) to corresponding graphic devices, reducing manual operations and improving design efficiency and accuracy. It is particularly suitable for scenarios with dense equipment or complex layouts, and possesses advantages such as strong robustness (handling uncertainty through entropy values ​​to avoid incorrect binding), strong adaptability (introducing area weights, prioritizing large devices, and normalized distances to adapt to different scales and dense layouts), and high efficiency and flexibility (the decision rules combine fast binding and backup matching, with adjustable parameters to balance accuracy and computational cost). The spatial entropy matching model, by introducing device area weight factors and spatial density factors, establishes a probabilistic association model, maintaining an association accuracy of >93% even in areas of overlapping devices.

[0068] Logical flow as follows Figure 3 As shown. Specifically, establishing the mapping relationship between the electrical equipment primitives and text based on the spatial entropy matching model includes:

[0069] Calculate the normalized spatial distance from the text location to the geometric center of the electrical equipment element, search the text location within a preset radius range, and filter the candidate equipment set for each text.

[0070] The association probability distribution of each candidate device is calculated by combining the area weight of electrical equipment elements, the entropy value is calculated using the association probability distribution, and a mapping relationship is established based on the entropy value.

[0071] The inputs in the processing flow are: text location set {Pj}, device center set {Ck}, and device area set {Area(Dk)}; correspondingly, the outputs are text-device binding relationships or auxiliary matching flags.

[0072] First, preprocess the data and calculate the global normalization factor max_bbox_diagonal (the maximum value of the diagonal of all device bounding boxes). Then, determine the candidate device set for each text Tj by searching for devices within the radius.

[0073] Secondly, for each text Tj, the association metric is calculated:

[0074] a. Calculate the normalized distance: Calculate the normalized spatial distance from the text location to the geometric center of the electrical equipment element for each candidate device Dk using the following formula:

[0075] dk=∥Pj-Ck∥ / max_bbox_diagonal

[0076] Where dk is the normalized spatial distance of the k-th device, Pj is the position coordinate of the text Tj, Ck is the geometric center of the device, and max_bbox_diagonal is the global normalization factor, which is the maximum value of the diagonal of the border of all electrical device elements. In the above formula, the Euclidean distance from the text to the device center is divided by the maximum diagonal length to eliminate the influence of the drawing scale and ensure that the distance value is in the range [0,1].

[0077] Geometric center of device Dk:

[0078] Ck=[(xmin+xmax) / 2,(ymin+ymax) / 2]

[0079] The position of text Tj is Pj=(xj,yj), the area of ​​device Dk is Area(Dk), and the number of candidate devices is K.

[0080] b. Calculate the location weight: For each candidate device Dk, calculate the area weight of the electrical equipment element using the following formula:

[0081] Wk=Area(Dk)^α

[0082] Where Wk represents the area weight of the k-th electrical equipment element, Area represents the area of ​​the electrical equipment element, Dk is the k-th candidate equipment, and α is a preset weight strength control value. It can be understood that the larger the equipment area, the higher the weight (α controls the weight strength), and larger equipment is preferentially associated. Even better, α = 0.3.

[0083] c. Calculate the association probability: For each candidate device Dk, calculate the association probability distribution using the area weights of electrical equipment elements:

[0084] pk=[exp(-β·dk)·Wk] / [Σ(m=1 to K)exp(-β·dm)·Wm]

[0085] Where pk represents the association probability of the k-th device, dk represents the normalized distance of the k-th device, β is a preset exponential decay control value, K is the number of candidate devices, and m = 1 to K is the summation index, representing traversing all candidate devices. The probability pk represents the association probability between text and device k, which is affected by both distance (β controls exponential decay) and area weight; the denominator ensures that the probability is normalized (summing to 1). dm represents the normalized distance between text Tj and candidate device Dm (i.e., dk in the formula, only m is used as the summation index). Wm represents the positional weight of candidate device Dm (i.e., Wk in the formula, again m is used as the summation index).

[0086] d. Calculate the entropy value: using the association probability to calculate the entropy value:

[0087] H(Tj)=-Σ(k=1 to K)pk log2 pk

[0088] Where Tj represents the j-th text, and H(Tj) represents the degree of disorder in the association distribution of the j-th text. The entropy value H(Tj) measures the disorder in the association distribution, with a value range of [0, log2K]. Low entropy (close to 0) indicates a concentrated distribution (clear association), while high entropy (close to log2K) indicates a uniform distribution (fuzzy association).

[0089] Finally, after calculating the entropy value based on the probability distribution {pk}, decision-making and binding are performed. When establishing the mapping relationship based on the entropy value, the minimum condition must be met. k When (dk) < γ or H(Tj) < θ, bind the text Tj to the nearest device; otherwise, initiate auxiliary matching, where γ is the minimum distance threshold and θ is the entropy threshold. The values ​​of α, β, γ, and θ can be set according to the dimensions of the CAD drawing.

[0090] Furthermore, the decision rule in the algorithm code is as follows:

[0091] if min k (dk)<γ or H(Tj)<θ: bind(Tj,argmin k (dk))

[0092] else: use_secondary_match(Tj)

[0093] Even better, γ = 0.1 (minimum distance threshold) and θ = 0.7 (entropy threshold). If either condition is met, the text is bound to the nearest device; otherwise, auxiliary matching (such as content-based fuzzy matching) is initiated.

[0094] Subsequently, the ledger matcher is used to establish the mapping relationship between equipment in CAD drawings and ledgers. When matching the electrical equipment elements and text with the models in the ledger archives, the system first retrieves completely identical equipment names in the ledger database through structured query for mapping. For unmatched equipment, similarity is calculated, and equipment with a similarity value greater than the preset value is generated as a candidate set for matching.

[0095] In one implementation, to achieve matching between CAD equipment and ledger archives, the ledger matcher adopts a dual matching strategy, and its core process is divided into three stages: precise matching, fuzzy matching, and manual verification interface.

[0096] Precise matching stage: Quickly retrieve completely identical device names from the ledger database through structured queries (SQL name = T_i.text) to achieve efficient mapping (e.g., labeling "SWG-5" directly matches ledger record 102).

[0097] The exact matching stage performs fast mapping, as shown in the code below:

[0098]

[0099] Fuzzy matching stage: Multi-dimensional similarity calculation is initiated for unmatched devices, combining text similarity (Levenshtein distance normalization) and spatial location weight (exponential decay function) to generate a candidate set and filter high-confidence results. "High-confidence results" refer to matching results whose similarity (reliability) reaches a preset threshold (Sim>0.85) after calculation in the fuzzy matching stage.

[0100] Error tolerance is implemented during the fuzzy matching stage, as shown in the code below:

[0101] Define the similarity function:

[0102] Sim(T_i,R_k)=ω1·[1-Levenshtein(T_i.norm,R_k.norm) / max_len]+ω2·exp(-‖Loc(T_i)-Loc(R_k)‖ / σ)

[0103] in:

[0104] ω1 = 0.6 (text similarity weight), ω2 = 0.4 (positional weight)

[0105] σ = 10% of the length of the diagonal of the drawing

[0106] Perform a TopK search: Returns R_k values ​​where Sim() > 0.85.

[0107] Manual verification interface: Generates a visual report for low-confidence or conflicting results (such as "XL-3" matching both "Line 3" and "Cable 3") to guide users to quickly confirm.

[0108] In the manual verification interface, a confidence report is generated, and high-conflict results prompt the user for confirmation.

[0109] The aforementioned dual ledger matching strategy combines precise matching and fuzzy matching to effectively solve the problem of differences between CAD annotations and network system naming, achieving a matching accuracy rate of 94.8%.

[0110] S3, wherein when the load transfer path generation operation is triggered, the load transfer path is searched according to the device topology library and device connection relationship, and the load transfer path is dynamically visualized.

[0111] Specifically, it also includes providing a dynamic visualization engine to trigger the load transfer path generation operation. After triggering the load transfer path generation operation, the dynamic visualization engine searches the equipment topology library, returns the adjacency matrix, and calculates the optimal load transfer path based on the line capacity constraints provided by the ledger system. The dynamic visualization engine then renders the entire load transfer path process to dynamically display the process.

[0112] This system innovatively integrates a dynamic visualization engine with topology-capacity dual-constraint path search and SVG real-time animation rendering technology. Breaking away from the traditional single-constraint path search mode, it simultaneously considers network topology and line capacity to accurately generate safe and feasible transfer paths. Combined with real-time visualization of the transfer process using SVG animation, it achieves real-time interaction of "clicking on the device → automatically generating the path," with a response latency of <300ms. It also supports parallel comparison of multiple solutions, significantly improving the efficiency and accuracy of power system load transfer decisions.

[0113] Logic flowchart as follows Figure 4 As shown, the user first triggers the process, such as when a feeder is under maintenance. In order to ensure the normal operation of the power grid, it is necessary to calculate which alternative lines can be used to replace the line under maintenance based on the power grid topology (i.e., using the SVG obtained from the identification results of the CAD drawings) to prevent the load from losing power.

[0114] First, the visualization engine searches the topology library, returning an adjacency matrix (a data structure used to represent the connections between nodes in a graph) and a list of available load transfer devices. Then, based on formulas and pathfinding algorithms, the optimal load transfer route is found. The ledger system returns the line capacity. When an available load transfer path is found, or when capacity is insufficient, the visualization engine renders and dynamically displays the entire load transfer process. The adjacency matrix, a data structure representing the connections between nodes in a graph, is fundamental for analyzing network structure and attributes in scenarios such as power systems, social networks, and transportation networks. It can be conveniently used for path finding in various graph algorithms, such as Depth-First Search (DFS), Breadth-First Search (BFS), and shortest path algorithms (Dijkstra, Floyd-Warshall, etc.).

[0115] In the code implementation, the user inputs the faulty line, and the following code calculates and provides a new load transfer route (i.e., a routing algorithm). The dynamic generation of the load transfer path mainly includes three key steps:

[0116] S3.1, Topology Construction: Convert the device identification results into a graph structure $G=(V,E)$. Vertex set $V={\text{ring mains cabinet},\text{circuit breaker},\text{plant}}$ Edge set $E={\text{busbar},\text{AC line segment}}$

[0117] S3.2, Path Search: After the user selects the starting device $S$ and the ending device $T$, the device connection relationships are obtained, and Dijkstra's algorithm is used to find the optimal path.

[0118]

[0119] S3.2, the core SVG animation code uses a dynamic gradient flow effect and adds a capacity setting:

[0120] Even better, by using the above code, not only can real-time interaction be achieved and the transfer process can be visualized in real time with SVG animation, but the response latency can also be greatly reduced, making the response latency less than 300ms.

[0121] The present invention provides a method and system for visualizing dynamic load transfer in power distribution networks, involving intelligent parsing of power distribution network CAD drawings, automatic modeling of power grid equipment topology, dynamic visualization simulation of load transfer, and digital twin applications of power systems. By intelligently aggregating discrete elements in CAD drawings into electrical equipment entities and establishing associations between equipment and text annotations and ledger data, an HTML visualization system supporting interactive dynamic load transfer is ultimately generated.

[0122] Based on the same technical concept, the present invention also provides a dynamic load transfer visualization system for distribution networks, comprising:

[0123] The parsing engine is used to read drawings and extract basic graphic elements;

[0124] The equipment identification module detects and combines the endpoints, line segments, and spatial distribution of basic graphic elements into complete electrical equipment graphic elements, and builds an equipment topology library.

[0125] The association modeling module establishes a mapping relationship between the electrical equipment graphic elements and text based on the spatial entropy matching model, and matches the electrical equipment graphic elements and text with the model in the ledger archive;

[0126] A dynamic visualization engine is used to trigger the load transfer path generation operation, search for load transfer paths based on the device topology library and device connection relationships, and dynamically visualize the load transfer paths.

[0127] For example, the drawing is a CAD-readable DWG / DXF file, and the parsing engine can be an intelligent CAD parsing engine that parses CAD drawings, used to read DWG / DXF files and extract basic graphic elements such as lines, polylines, and text.

[0128] Therefore, in the device identification module, the distribution of endpoints, line segments, and space of basic graphic elements is detected step by step, including: detecting the degree of endpoint overlap; when the endpoint spacing is less than a preset tolerance value, the basic graphic elements are aggregated into the first connecting component; otherwise, the basic graphic elements are further subjected to projection intersection detection: calculating the projection of the line segment in a specified direction; when the projection intersection is not less than a preset projection value, the basic graphic elements are aggregated into the second connecting component; otherwise, the basic graphic elements are further subjected to density clustering detection: constructing a spatial density function, and classifying those with a density greater than a preset density threshold as device clusters, otherwise as independent devices. This three-level graphic element aggregation algorithm combines discrete line segments into electrical equipment entities, breaking through the limitations of traditional template matching and single-level connection detection. The three-level progressive mechanism adapts to different device complexities (simple line segments → busbars, complex clusters → ring main units), and density clustering solves the problem of non-connecting graphic element aggregation (such as scattered circuit breaker symbols).

[0129] The association modeling module establishes the association relationship between text, equipment, and lines based on the spatial entropy matching model. It introduces equipment area weighting factors and spatial density factors to establish a probabilistic association model, which still maintains an association accuracy of >93% in the overlapping areas of equipment.

[0130] The spatial entropy matching model, based on information entropy theory, calculates the normalized spatial distance from the text location to the geometric center of the device and combines it with device area weights to form an association probability distribution, using entropy values ​​to measure the concentration of the distribution. Low entropy means the association is clear and can be directly bound, while high entropy indicates that the association is ambiguous and requires auxiliary matching. The model then combines distance thresholds and entropy thresholds into a decision rule to achieve efficient and robust matching. It can automatically bind CAD text labels (such as device numbers) to the corresponding graphic devices, reducing manual operations and improving design efficiency and accuracy. It is particularly suitable for scenarios with dense equipment or complex layouts, and possesses advantages such as strong robustness, adaptability, and high efficiency and flexibility.

[0131] Finally, a dynamic visualization engine is used to generate HTML files, which embed SVG graphics and interactive logic, supporting dynamic load transfer simulation. The dynamic visualization engine innovatively integrates topology-capacity dual-constraint path search with real-time SVG animation rendering technology. To achieve matching between CAD equipment and ledger files, the ledger matcher adopts a dual matching strategy, breaking through the traditional single-constraint path search mode, simultaneously considering network topology and line capacity to accurately generate safe and feasible transfer paths. Combined with real-time visualization of the transfer process using SVG animation, it achieves real-time interaction of "clicking on equipment → automatically generating path," with a response latency of <300ms, and supports parallel comparison of multiple schemes, significantly improving the efficiency and accuracy of power system load transfer decisions.

[0132] The system employs a three-level primitive aggregation mechanism, a spatial entropy matching model, a dual-ledger matching strategy, and a dynamic visualization engine. The three-level primitive aggregation mechanism breaks through the limitations of traditional single-level connection detection by using a progressive processing approach of endpoint overlap → projection intersection → density clustering to solve the challenge of identifying complex equipment (ring network cabinet identification rate increased from 71% to 96%). The spatial entropy matching model introduces equipment area weighting factors and spatial density factors to establish a probabilistic association model, maintaining an association accuracy of >93% even in overlapping equipment areas. The dual-ledger matching strategy combines precise matching and fuzzy matching to effectively solve the problem of discrepancies between CAD annotations and distribution network system naming, achieving a matching accuracy of 94.8%. The dynamic visualization engine (or dynamic topology response engine) enables real-time interaction of "clicking on a device → automatically generating a transfer path," with a response latency of <300ms and supports parallel comparison of multiple schemes.

[0133] The present invention provides a dynamic load transfer visualization system for distribution networks, which uses the dynamic load transfer visualization method for distribution networks described above to process drawing data and dynamically visualizes the load transfer path when the load transfer path generation operation is triggered.

[0134] This invention overcomes the technical bottleneck of traditional CAD parsing, where discrete line segments are difficult to identify as complete devices, through a three-level primitive aggregation algorithm. It significantly improves the identification accuracy of complex devices such as ring main units and circuit breakers from 71% to 96.2%, dramatically increasing drawing parsing efficiency. An innovative association mechanism based on a spatial entropy matching model achieves 93.8% accuracy in automatic text-to-device binding in densely populated areas, overcoming the mismatch problem caused by spatial overlap in traditional nearest neighbor methods, thus improving drawing parsing efficiency by nearly 8 times. The dual ledger matching strategy uses text-space joint similarity calculation to effectively solve the naming differences between CAD annotations and distribution network ledgers, achieving an automatic matching coverage rate of 94.8% and reducing the need for manual intervention to below 6%. The dynamic visualization engine enables end-to-end conversion from CAD drawings to an interactive HTML system, supporting dynamic path generation based on topological constraints and real-time capacity calculation, with a response latency of less than 300ms, and can output multiple load transfer schemes in parallel. This solution helps distribution network operation and maintenance transform from static drawing management to dynamic digital twins, providing efficient response capabilities for load transfer decisions.

[0135] The above description is only a description of preferred embodiments of the present invention and is not intended to limit the scope of the present invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure shall fall within the protection scope of the present invention.

Claims

1. A method for visualizing dynamic load transfer in a distribution network, characterized in that, include: Extract basic graphic elements from the drawings, detect the distribution of endpoints, line segments, and spaces of the basic graphic elements step by step, and combine them into complete electrical equipment graphic elements to build an equipment topology library; The mapping relationship between the electrical equipment graphic elements and text is established based on the spatial entropy matching model, and the electrical equipment graphic elements and text are matched with the model in the ledger archive; Specifically, when the load transfer path generation operation is triggered, the load transfer path is searched based on the device topology library and device connection relationship, and the load transfer path is dynamically visualized.

2. The method for visualizing dynamic load transfer in a distribution network according to claim 1, characterized in that, The step-by-step detection of the endpoints, line segments, and spatial distribution of basic graphic elements includes: The degree of overlap of the endpoints is detected. If the distance between the endpoints is less than a preset tolerance value, the basic primitives are aggregated into the first connecting component; otherwise, the projection intersection detection of the basic primitives continues. Calculate the projection of the line segment in the specified direction. If the intersection of the projections is not less than a preset projection value, then the basic primitives are aggregated into a second connection component; otherwise, continue to perform density clustering detection on the basic primitives. Construct a spatial density function to classify devices with a density greater than a preset density threshold as device clusters, and otherwise as independent devices.

3. The method for visualizing dynamic load transfer in a distribution network according to claim 1, characterized in that, Establishing the mapping relationship between the electrical equipment primitives and text based on the spatial entropy matching model includes: Calculate the normalized spatial distance from the text location to the geometric center of the electrical equipment element, search the text location within a preset radius range, and filter the candidate equipment set for each text. The association probability distribution of each candidate device is calculated by combining the area weight of electrical equipment elements, the entropy value is calculated using the association probability distribution, and a mapping relationship is established based on the entropy value.

4. The method for visualizing dynamic load transfer in a distribution network according to claim 3, characterized in that, Calculate the normalized spatial distance from the text location to the geometric center of the electrical equipment element: dk=∥Pj-Ck∥ / max_bbox_diagonal Where dk is the normalized spatial distance of the k-th device, Pj is the position coordinate of the text Tj, Ck is the geometric center of the device, and max_bbox_diagonal is the global normalization factor, which is the maximum value of the diagonal of the border of all electrical device primitives.

5. The method for visualizing dynamic load transfer in a distribution network according to claim 3, characterized in that, The area weight of the electrical equipment elements is calculated using the following formula: Wk=Area(Dk)^α Where Wk represents the area weight of the k-th electrical equipment element, Area represents the area of ​​the electrical equipment element, Dk is the k-th candidate equipment, and α is a preset weight strength control value; the association probability distribution is calculated using the area weight of the electrical equipment element: pk=[exp(-β·dk)·Wk] / [Σ(m=1 to K)exp(-β·dm)·Wm] Where pk represents the association probability of the k-th device, dk represents the normalized distance of the k-th device, β is a preset exponential decay control value, and K is the number of candidate devices; the entropy value is calculated using the association probability: H(Tj)=-Σ(k=1 to K)pk log2 pk Where Tj represents the j-th text, and H(Tj) represents the disorder of the association distribution of the j-th text.

6. The method for visualizing dynamic load transfer in a distribution network according to claim 5, characterized in that, When establishing a mapping relationship based on entropy values, if mink(dk) < γ or H(Tj) < θ, the text Tj is bound to the nearest device; otherwise, auxiliary matching is initiated, where γ is the minimum distance threshold and θ is the entropy threshold.

7. The method for visualizing dynamic load transfer in a distribution network according to claim 1, characterized in that, When matching the electrical equipment graphics and text with the model in the ledger file, the system first retrieves the completely identical equipment names in the ledger database through structured query and maps them; for unmatched equipment, the similarity is calculated, and equipment with a similarity value greater than the preset value is generated as a candidate set for matching.

8. The method for visualizing dynamic load transfer in a distribution network according to claim 1, characterized in that, It also includes a dynamic visualization engine for triggering load transfer path generation operations. After triggering the load transfer path generation operation, the dynamic visualization engine searches the equipment topology library, returns the adjacency matrix, and calculates the optimal load transfer path based on the line capacity constraints provided by the ledger system. The dynamic visualization engine then renders the entire load transfer path process to dynamically display the process.

9. A dynamic load transfer visualization system for a power distribution network, characterized in that, include: The parsing engine is used to read drawings and extract basic graphic elements; The equipment identification module detects and combines the endpoints, line segments, and spatial distribution of basic graphic elements into complete electrical equipment graphic elements, and builds an equipment topology library. The association modeling module establishes a mapping relationship between the electrical equipment graphic elements and text based on the spatial entropy matching model, and matches the electrical equipment graphic elements and text with the model in the ledger archive; A dynamic visualization engine is used to trigger the load transfer path generation operation, search for load transfer paths based on the device topology library and device connection relationships, and dynamically visualize the load transfer paths.

10. A dynamic load transfer visualization system for a power distribution network, characterized in that, The dynamic load transfer visualization method for distribution networks as described in any one of claims 1-8 is used to process the drawing data, and the load transfer path is dynamically visualized when the load transfer path generation operation is triggered.

Images