A meshed distribution network line path automatic planning method based on digital twinning

Abstract

The application discloses a kind of based on digital twinning's meshing distribution network line path automatic planning method, it is related to power grid construction technical field, including the following steps: computing device acquires historical construction distribution network line area's line drawing, and the image of line drawing is blurred and missing information repair completion by generating adversarial network, non-standard specification and parameter identification mark in line drawing, generate non-standard data set, unify line drawing information to corresponding coordinate system, and correct offset coordinate system, export digital twinning model after correction.This can accurately reflect the digital twinning model of distribution network full element state by constructing, using the model makes that path planning and actual geographical environment, equipment parameter and environmental data depth match, reaches the high degree of agreement of planning scheme and actual scene, and designs meshing path design idea, so that subsequent distribution network line path planning can be more intuitive and efficient.

Description

Technical Field

[0001] This invention relates to the field of power construction technology, specifically to an automatic planning method for grid-based distribution network line paths based on digital twins. Background Technology

[0002] Digital twins fully utilize physical models, sensors, operational history, and other data to integrate multi-disciplinary, multi-physical, multi-scale, and multi-probability simulation processes, completing mapping in virtual space to reflect the entire life cycle of the corresponding physical equipment. As a key component of the power system, the distribution network's line planning is a crucial link in ensuring power supply reliability and economy. The scientific and efficient design of its lines directly affects power supply reliability and power quality.

[0003] Existing distribution network route planning methods rely heavily on 2D CAD software and the experience of designers, lacking adaptive adjustment of dynamic weights and cross-system collaborative optimization mechanisms for urban planning. They cannot accurately reflect the status of all elements of the distribution network, resulting in deviations between the route planning scheme and the actual geographical environment, operating conditions and municipal planning. This leads to high route conflict rates, frequent adjustments and poor long-term adaptability, seriously affecting the accuracy and efficiency of distribution network planning. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides an automatic route planning method for gridded distribution networks based on digital twins, which solves the problems mentioned in the background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: an automatic route planning method for grid-based distribution networks based on digital twins, comprising the following steps:

[0006] A1. The computing device collects the line drawings of the historically constructed distribution network line area, and uses a generative adversarial network to repair and complete the blurred and incomplete information of the line drawings, identifies and annotates non-standard specifications and parameters in the line drawings, and generates a non-standard dataset.

[0007] The route drawing information is unified to the corresponding coordinate system, and the offset coordinate system is corrected. After correction, a digital twin model is exported.

[0008] A2. Import the power distribution network line drawings to be constructed using the digital twin model. Compare the power distribution network line drawings with the non-standard dataset to confirm non-standard information and generate the corresponding image model.

[0009] The image model is used to divide the image into grid cells according to a preset precision. The starting point and ending point of the distribution network line are input into the grid cell image, the shortest path is calculated, and candidate paths are generated.

[0010] A3. Obtain municipal planning information, import the municipal planning information into the digital twin model to generate conflict detection drawings, and overlay the conflict detection drawings with the grid cell drawings to detect the conflict relationship of the candidate paths;

[0011] If there is no conflict, it is marked as a pre-selected candidate path; if there is a conflict, it is marked as a deleted path and the process returns to step A2 to regenerate.

[0012] A4. Use the digital twin model to perform a real-world simulation of the pre-selected candidate path to verify its feasibility.

[0013] If the simulation verification is feasible, the distribution network line path is confirmed. If the simulation verification is not feasible, return to step A2 to regenerate until the distribution network line path is confirmed.

[0014] Preferably, the power distribution network line map includes geographical data, power distribution network equipment data, and environmental data. The grid unit map is preset to a precision of 50m×50m for densely built-up urban areas, and 100m×100m for flat areas where there are no specially constructed buildings or equipment on the ground or underground.

[0015] Preferably, the cells of the grid unit drawing are labeled with grid number and function label. The function label includes the industrial area, residential area, commercial area and ecological area corresponding to the grid unit drawing. The grid number adopts the format of area code-row number-column number.

[0016] Preferably, the realistic route selection scenario simulation includes:

[0017] The simulation data of the distribution network line path under extreme environment, peak grid load and load growth scenarios in the next 5 years are used. The simulation of the realistic line selection scenario is performed 3 to 5 times. The average value of the simulation data results is taken to confirm the feasibility of the path selection of the distribution network line path.

[0018] Preferably, the coordinate system adopts the globally unified coordinate system WGS-84 standard, the satellite positioning system adopts the Global Positioning System (GPS), the deviation of the coordinate system after correction does not exceed 0.5m, and the route drawing is a dwg format file.

[0019] Preferably, the municipal planning information includes: road red lines, underground pipeline distribution, rail transit, protected areas and ecological protection areas, and corresponding municipal planning drawings. The conflict detection drawings adopt spatial overlay analysis, that is, the candidate paths are overlaid on the layer of the conflict detection drawings. Based on the overlap analysis of the municipal planning information and the candidate paths in the conflict detection drawings, the obstruction and influence relationship of the paths on the municipal planning information is analyzed.

[0020] Preferably, the evaluation indicators for feasibility verification include: construction cost of distribution network lines, construction difficulty, safety distance compliance rate, and compatibility with municipal planning. When all indicators meet the preset threshold, the evaluation is "yes"; otherwise, when the indicators do not meet the preset threshold, the evaluation is "no".

[0021] Preferably, the shortest path is calculated using Dijkstra's algorithm for the single-source shortest path problem, and the specific steps of Dijkstra's algorithm are as follows:

[0022] Before executing this algorithm, the generation of the grid cell drawing in step A2 is completed, and the starting point s and ending point t of the distribution network line have been clearly identified as the corresponding grid cell marker points;

[0023] A5.1. Given a weighted graph G = (V, E), where V is the set of grid marker points and E is the set of grid edges, for a grid point sV in the graph, the shortest path from s to all other points in G is required. Dijkstra's algorithm requires three data structures: S set: storing the shortest paths of explored points; U set: storing the shortest paths of unexplored points; and D array: storing the shortest distance value of each point.

[0024] A5.2. Initialize the S set to contain only the starting point s of the grid cell drawing, the U set to contain all other points of the grid cell drawing, the shortest path of the starting point s is 0, the value in the D array is 0, and the shortest path value of other points is infinity;

[0025] A5.3 Select the grid marker point u with the smallest value in array D from the set U, and move the grid marker point u from the set U to the set S;

[0026] A5.4. Traverse all adjacent grid markers v of the grid marker u, calculate the temporary distance from the starting point s through the grid marker u to the adjacent grid marker v. If the temporary distance < D[v], update D[v] = temporary distance and record P[v] = u, where P is the predecessor node record array, and the predecessor node record array stores the shortest path predecessor node of each grid marker for subsequent path backtracking.

[0027] A5.5 Repeat step A5.3 until the endpoint t is moved into set S. At this time, D[t] is the shortest path distance from the starting point to the endpoint.

[0028] A5.6 Starting from the endpoint t, backtrack through the predecessor node array and arrange the grid marker point u sequence obtained by backtracking in order to generate candidate paths for the distribution network.

[0029] Preferably, the generative adversarial network is a GAN network, and the training samples of the generative adversarial network include more than 3,000 images of the power distribution network line drawings that are blurry and incomplete. The image information integrity of the power distribution network line drawings after repair and completion is not less than 95%. The identification and annotation of non-standard specifications and parameters in the power distribution network line drawings include: image coordinates, size, equipment model and electrical parameters.

[0030] A computer device, characterized in that the computer device includes a processor and a memory, the memory storing at least one piece of program code, the at least one piece of program code being loaded and executed by the processor to implement the aforementioned automatic planning method for grid-based distribution network line paths based on digital twins.

[0031] This invention provides an automatic route planning method for gridded distribution networks based on digital twins. It has the following beneficial effects:

[0032] (1) By constructing a digital twin model that can accurately reflect the status of all elements of the distribution network, the path planning is deeply matched with the actual geographical environment, equipment parameters and environmental data, achieving a high degree of fit between the planning scheme and the actual scenario. Furthermore, the gridded path design approach makes the subsequent planning of distribution network lines more intuitive and efficient. The gridded path planning model enables automated path planning, effectively reducing the deviation between the path and the actual situation and improving the accuracy of distribution network line planning.

[0033] (2) By using the grid processing of the distribution network line drawings and the digital twin model, the selection of distribution network line paths can be quickly calculated through point-to-point grids, and the shortest line path can be quickly identified within the grid cells. Furthermore, by accessing municipal planning information and simulation verification of the digital twin model, the feasibility of the line path can be verified, thereby selecting the most suitable line path, improving the efficiency of automatic planning of the distribution network line path and the safety and accuracy of the line.

[0034] (3) In the process of route generation, municipal planning data is integrated to conduct conflict detection and adjustment under multiple constraints, so that the candidate route can fully meet the external constraints such as urban spatial layout and functional zoning, achieve the synergistic adaptation between the route and municipal planning, reduce the spatial conflict between the route and municipal facilities, and improve the overall efficiency of power distribution line planning and the safety of route selection. Attached Figure Description

[0035] Figure 1 This is a flowchart illustrating the planning process of an automatic route planning method for gridded distribution networks based on digital twins, according to the present invention.

[0036] Figure 2 This is a flowchart illustrating the planning steps of an automatic route planning method for gridded distribution networks based on digital twins, according to the present invention. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0038] Example 1

[0039] Please see Figure 1-2 This invention provides an automatic route planning method for gridded distribution networks based on digital twins. To achieve the above objectives, this invention is implemented through the following technical solution, including the following steps:

[0040] A1. The computing device collects the line drawings of the historically constructed distribution network line area, and uses a generative adversarial network to repair and complete the blurred and incomplete information of the line drawing image, identify and label non-standard specifications and parameters in the line drawing, and generate a non-standard dataset.

[0041] Unify the route drawing information to the corresponding coordinate system, correct the offset coordinate system, and then export the digital twin model.

[0042] A2. Import the power distribution network line drawings to be constructed using a digital twin model. Compare the power distribution network line drawings with the non-standard dataset to confirm non-standard information and generate the corresponding image model.

[0043] The image model is used to divide the drawing into grid cells according to a preset precision. The starting point and ending point of the distribution network line are input into the grid cell drawing to calculate the shortest path and generate candidate paths.

[0044] A3. Obtain municipal planning information, import the municipal planning information into the digital twin model to generate conflict detection drawings, and overlay the conflict detection drawings with the grid cell drawings to detect the conflict relationship of candidate paths.

[0045] If there is no conflict, it is marked as a pre-selected candidate path; if there is a conflict, it is marked as a deleted path and the process returns to step A2 to regenerate.

[0046] A4. Use a digital twin model to simulate the feasibility of the pre-selected candidate paths.

[0047] If the simulation verification is feasible, then confirm the distribution network line path. If the simulation verification is not feasible, return to step A2 to regenerate until the distribution network line path is confirmed.

[0048] In this embodiment, the route planning of the distribution network line is achieved through a digital twin model and grid-based planning. When creating the digital twin model, the historical geographic data, distribution network equipment data, and environmental data of the distribution network line route are used as training data for the route map. The geographic data consists of geographic images of the distribution network line area, the distribution network equipment data consists of the parameters of the distribution network line and the distribution network equipment, and the environmental data consists of images of buildings, pipelines, terrain, and vegetation in the geographic data. The amount of data is no less than 3,000 sets. The route map is first repaired and completed by a GA network to remove any blurry or incomplete parts.

[0049] When generating the digital twin model, the computing device uses the YOLOv8 object detection algorithm as a pre-trained model. The repaired circuit diagram is imported into the YOLOv8 model, and the YOLOv8 model calling hierarchy is simplified, retaining only the function for feature extraction. This simplified YOLOv8 model reduces the computer load during YOLOv8 model operation, requiring only the extraction of necessary features. The YOLOv8 model identifies non-industry standard equipment and wiring specifications and parameters in the imported circuit diagram. Prior to this, the YOLOv8 model needs to import industry standard equipment and wiring specifications and parameters in advance to identify non-standard specifications and parameters. Then, YOLO... The v8 model learns to identify non-standard specifications and parameters in a large number of route drawings and generates corresponding non-standard datasets for non-standard specification and parameter identification. Then, it combines the geographic data, distribution network equipment data, and environmental data in the learned route drawings and maps them into the YOLOv8 model. When combining the geographic data, distribution network equipment data, and environmental data of each group of route drawings, the coordinate system is aligned and combined. Then, the coordinate system of the corresponding combination is corrected by calling the data of the satellite positioning system to ensure that the image is accurate. After the training and combination are completed, a set of regional models for distribution network route selection and planning can be obtained. The extraction, training and combination of the above large amount of data training process generates the corresponding digital twin model.

[0050] When it is necessary to plan the route of the distribution network, the distribution network route drawings of the current construction can be imported into the digital twin model. Then, the non-standard specifications and parameters of the non-standard dataset are compared with the current distribution network route drawings. The graphics in the distribution network route drawings are identified and the equipment, lines and construction information of the non-standard specifications and parameters in the non-standard dataset are automatically matched. After the above steps, a route planning model of the distribution network area can be obtained and exported. Then, the model is opened on the computer using the digital twin model (CIMPro) application software. The planar image of the model is then divided into grids in the two-dimensional workspace. Each grid is the same size and the scale of the actual geography is 1:50 or 1:100. The scale is adjusted according to the complexity of the terrain. After the two-dimensional image of the model is gridded, each grid is numbered and labeled with a function label.

[0051] Then, the Dijkstra algorithm is used to calculate and design the distribution network line path within the grid, generating a candidate path. The municipal future planning information of the corresponding area is then imported into the digital twin model software and superimposed on the candidate path to identify whether there is any future construction interference. If there is no interference, the candidate path is confirmed. If there is interference, the new shortest path is recalculated to remove the candidate path, and interference checks are performed again until there is no interference. The confirmed interference-free candidate path is then simulated to verify the usage of the distribution network line in the future environment. This verifies whether accelerated aging due to environmental and surrounding power consumption factors will affect its performance and lifespan. Therefore, if the simulation verification is successful, the candidate path is determined as the final construction path. Otherwise, the candidate path is recalculated, and the optimal solution for the distribution network line path is repeatedly verified.

[0052] Example 2

[0053] Specifically: The power distribution network line drawing includes geographical data, power distribution network equipment data and environmental data. The grid unit drawing is preset to a precision of 50m×50m for densely built-up urban areas, and 100m×100m for flat areas where there are no specially constructed buildings or equipment on the ground or underground.

[0054] The cells in the grid unit drawing are labeled with grid number and function label. The function label includes the industrial area, residential area, commercial area and ecological area corresponding to the grid unit drawing. The grid number adopts the format of area code-row number-column number.

[0055] Realistic route selection scenario simulation includes:

[0056] Simulation data of distribution network line routes under extreme environments, peak grid load, and load growth scenarios over the next 5 years are used. The simulation of realistic route selection scenarios is performed 3 to 5 times, and the average value of the simulation data is taken to confirm the feasibility of the distribution network line route selection.

[0057] The coordinate system adopts the globally unified WGS-84 coordinate system standard, and the satellite positioning system adopts the Global Positioning System (GPS). The deviation after coordinate system correction does not exceed 0.5m, and the route drawing is in dwg format.

[0058] The municipal planning information includes: road red lines, underground pipeline distribution, rail transit, protected areas and ecological protection areas, and corresponding municipal planning drawings. The conflict detection drawings adopt spatial overlay analysis, that is, the candidate paths are overlaid on the layer of the conflict detection drawings. Based on the overlap analysis of the municipal planning information and candidate paths in the conflict detection drawings, the obstruction and influence relationship of the paths on the municipal planning information is analyzed.

[0059] The generative adversarial network (GAN) is used. The training samples for the GAN include over 3000 blurry and incomplete power distribution network (PDN) diagrams. The integrity of the repaired PDN diagrams is no less than 95%. The identification and annotation of non-standard specifications and parameters in the PDN diagrams include: image coordinates, dimensions, equipment model, and electrical parameters.

[0060] In this embodiment, the WGS-84 coordinate system has ellipsoidal parameters of 6,378,137 meters for the semi-major axis and 1 / 298.257223563 for the flattening. It is a current standard and is applicable to global geographic positioning. It provides a unified benchmark for the spatial coordinate system alignment of the data in this method. The GPS satellite positioning system receives three-dimensional coordinates from ranging signals from at least four satellites, achieving centimeter-level positioning accuracy. By comparing with the primitive transformation coordinates, linear interpolation is used to correct the deviation, ensuring that the overall deviation after correction does not exceed 0.5m.

[0061] The industrial zone label corresponds to factory clusters, requiring matching the line capacity of high-load equipment; the residential zone label corresponds to concentrated residential areas, requiring priority consideration of the safe distance between lines and buildings and the impact on the landscape; the commercial zone label corresponds to densely populated shops, requiring special attention to avoid underground pipelines (such as gas pipes and communication optical cables); the ecological zone label corresponds to nature reserves, wetland parks, etc., requiring strict restrictions on line crossings. This allows us to identify which grid points within the grid area cannot be used for line path selection, thereby improving the efficiency of path selection.

[0062] The GAN network training and repair uses more than 3,000 power distribution line drawings with blurriness (such as scanning noise, line blurring) and incompleteness (such as missing local line segments, broken symbols). The repair and completion are achieved through adversarial training between the generator and discriminator of the GAN network. The complete drawings verified by humans are used as the benchmark. When the outline and symbols of the repaired graphic elements match the benchmark drawing with a degree of ≥95%, the information is judged to be complete.

[0063] Feasibility verification verifies the construction cost, including materials (cables, towers, and lines), construction (excavation, laying, and labor), and future operation and maintenance costs. The preset threshold is that it should not exceed the construction budget. The construction difficulty is divided into three levels: low, medium, and high, based on the complexity of the terrain and the number of obstacles. The threshold is medium and below. The safety distance compliance rate refers to the percentage of path segments that meet the industry's technical requirements. The threshold is ≥95%. The urban planning compatibility refers to the degree of coordination between the path and the urban plan (such as avoiding the area to be built). The threshold is ≥90%. When all indicators meet the threshold, the evaluation is "yes"; otherwise, it is "no".

[0064] When dividing the grid, a 50m×50m grid is used for densely built-up urban areas, while a 100m×100m grid is used for flat areas where there are no special buildings or equipment on the ground or underground. This reduces the model pressure on the planning grid, while the 50m×50m grid allows for more refined and detailed route planning in densely built-up urban areas.

[0065] Example 3

[0066] Specifically: The shortest path is calculated using Dijkstra's algorithm for the single-source shortest path problem. The specific steps of Dijkstra's algorithm are as follows:

[0067] Before executing this algorithm, the generation of the grid cell drawing in step A2 is completed, and the starting point s and ending point t of the distribution network line have been clearly identified as the corresponding grid cell marker points;

[0068] A5.1. Given a weighted graph G = (V, E), where V is the set of grid marker points and E is the set of grid edges, for a grid point sV in the grid cell graph, the shortest path from s to all other points in G is required. Dijkstra's algorithm requires three data structures: S set: storing the shortest paths of explored points; U set: storing the shortest paths of unexplored points; and D array: storing the shortest distance value of each point.

[0069] A5.2 Initialize the S set to contain only the starting point s of the grid cell drawing, and the U set to contain all other points of the grid cell drawing. The shortest path of the starting point s is 0, the value in the D array is 0, and the shortest path value of other points is infinity.

[0070] A5.3 Select the grid marker u with the smallest value in array D from set U, and move grid marker u from set U to set S;

[0071] A5.4. Traverse all adjacent grid markers v of grid marker u, calculate the temporary distance from the starting point s through grid marker u to the adjacent grid marker v. If the temporary distance < D[v], update D[v] = temporary distance and record P[v] = u, where P is the predecessor node record array, and the predecessor node record array stores the shortest path predecessor node of each grid marker for subsequent path backtracking.

[0072] A5.5 Repeat step A5.3 until the endpoint t is moved into set S. At this point, D[t] is the shortest path distance from the starting point to the endpoint.

[0073] A5.6 Starting from the endpoint t, backtrack through the predecessor node array and arrange the grid marker point u sequence obtained by backtracking in order to generate candidate paths for the distribution network line;

[0074] In this embodiment, the implementation of Dijkstra's algorithm needs to be based on the grid cell drawing. The specific process is as follows: take the center point of each grid cell drawing as a node, the node number corresponds to the grid number, establish directed edges between nodes, the initial weight of the edge is set to the straight-line distance between the two nodes, use the adjacency list to store the graph structure to reduce storage space, use a priority queue to store the nodes to be processed to improve the node extraction efficiency, and optimize the time complexity of the algorithm to O. Set the shortest distance of the starting node to 0, and set the shortest distance of other nodes to infinity. Initialize the priority queue to only contain the starting node, and its priority is 0. Extract the node with the smallest distance from the priority queue, traverse all its adjacent nodes, calculate the distance from the starting point through the smallest node to the adjacent node, if the distance is less than the current shortest distance of the adjacent node, update the shortest distance of the adjacent node, and insert the adjacent node into the priority queue. Repeat this process until the priority queue is empty. After the path selection is completed, start from the end node, backtrack to the starting point through the recorded predecessor nodes to obtain the node sequence of the shortest path, that is, the grid path of the distribution network line.

[0075] By optimizing the implementation details of Dijkstra's algorithm, the efficiency and accuracy of path calculation are improved, ensuring that the shortest path can be found quickly in large-scale grids. The quantitative calculation of the overall path cost takes into account multiple factors, and the dynamic adjustment of the scaling factor enables path planning to adapt to different planning objectives, thereby improving the flexibility and applicability of path solutions.

[0076] Example 4

[0077] Specifically: A computer device, comprising a processor and a memory, wherein the memory stores at least one line of program code, which is loaded and executed by the processor to implement a method for automatic path planning of grid-based distribution network lines based on digital twins.

[0078] It can be implemented wholly or partially through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented wholly or partially as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, they generate, wholly or partially, the flow or function according to embodiments of the present invention. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data processing device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0079] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A meshed distribution network line path automatic planning method based on digital twinning, characterized in that, The method comprises the following steps: A1, the computing device collects the line drawing of the historical construction and distribution network line area, and restores and completes the image of the line drawing through a generative adversarial network to blur and incomplete information, identifies and labels non-standard specifications and parameters in the line drawing, and generates a non-standard data set; The line drawing information is unified to the corresponding coordinate system, and the offset coordinate system is corrected. The corrected digital twin model is derived; A2, the digital twin model is used to import the distribution network line drawing to be constructed. The distribution network line drawing is compared with the non-standard data set to identify non-standard information, and a corresponding image model is generated; The image model is used to divide the grid unit drawing according to the preset accuracy. The starting point and the ending point of the distribution network line are input into the grid unit drawing, the shortest path is calculated, and a candidate path is generated; A3, the municipal planning information is obtained, and the municipal planning information is imported into the digital twin model to generate a conflict detection drawing. The conflict detection drawing is superimposed with the grid unit drawing to detect the conflict relationship of the candidate path; When no conflict is detected, the preselected candidate path is marked; when there is a conflict, the path is deleted and step A2 is returned to regenerate; A4, the digital twin model is used to simulate the authenticity of the preselected candidate path, and the feasibility of the preselected candidate path is verified; When the simulation verification is feasible, the distribution network line path is confirmed. When the simulation verification is not feasible, step A2 is returned to regenerate until the distribution network line path is confirmed.

2. The grid-based distribution network line path automatic planning method based on digital twinning according to claim 1, characterized in that, The distribution network line drawing contains geographic data, distribution network equipment data and environmental data. The grid unit drawing has a preset accuracy of 50m x 50m for urban building dense areas, and 100m x 100m for flat areas without special construction buildings or equipment on the ground and underground.

3. The grid-based distribution network line path automatic planning method based on digital twinning of claim 1, wherein, The grid unit drawing has a grid number and a function label. The function label includes the corresponding industrial area, residential area, commercial area and ecological area of the grid unit drawing. The grid number is numbered in the format of area code-row number-column number.

4. The grid-based distribution network line path automatic planning method based on digital twinning of claim 1, wherein, The authenticity route selection scenario simulation includes: The simulation data of the line in the distribution network line path in extreme environment, peak load of power grid and load growth scenario in the next 5 years. The simulation times of the authenticity route selection scenario simulation are 3-5 times. The simulation data results are averaged to confirm the path selection feasibility of the distribution network line path.

5. The grid-based distribution network line path automatic planning method based on digital twinning of claim 1, wherein, The coordinate system adopts the global unified coordinate system WGS-84 coordinate system standard, and the satellite positioning system adopts the global satellite positioning system GPS system. The deviation of the corrected coordinate system is not more than 0.5m. The line drawing is a dwg format file.

6. The grid-based distribution network line path automatic planning method based on digital twinning of claim 1, wherein, The municipal planning information includes road red line, underground pipeline distribution, rail transit, protected area and ecological protection range and corresponding municipal planning drawings, the conflict detection drawing adopts spatial overlay analysis, that is, the candidate path is superimposed on the layer of the conflict detection drawing, and the blocking and influence relationship of the path on the municipal planning information is analyzed according to the overlap analysis of the municipal planning information in the conflict detection drawing and the candidate path.

7. The grid-based distribution network line path automatic planning method based on digital twinning of claim 1, wherein, The evaluation indexes of the feasibility verification include construction cost, construction difficulty, safety distance compliance rate and municipal planning fitting degree of the distribution network line, and the evaluation is "yes" when each index meets the preset threshold, and the evaluation is "no" when the index does not meet the preset threshold.

8. The grid-based distribution network line path automatic planning method based on digital twinning of claim 1, wherein, The shortest path is calculated by using the Dijkstra algorithm of the single-source shortest path problem, and the algorithm steps of the Dijkstra algorithm are as follows: Before executing the algorithm, the generation of the grid cell drawing in step A2 is completed, and the starting point s and the terminal point t of the distribution network line are clearly corresponding to the mark points of the grid cells; A5.1, there is a weighted graph G=(V, E), wherein V is a set of grid mark points, E is a set of grid edges, and for the grid point sV of the grid cell drawing, the shortest path Dijkstra algorithm requires three data structures, the S set: storing the shortest path of the explored point, the U set: storing the shortest path of the unexplored point, and the D array: storing the shortest distance value of each point; A5.2, the S set only contains the starting point s of the grid cell drawing, the U set contains all the remaining points of the grid cell drawing, the shortest path of the starting point s is 0, the value in the D array is 0, and the shortest path value of the other points is infinite; A5.3, select the grid mark point u with the minimum D array value from the U set, and move the grid mark point u from the U set to the S set; A5.4, traverse all adjacent grid mark points v of the grid mark point u, calculate the temporary distance from the starting point s to the adjacent grid mark point v through the grid mark point u, and when the temporary distance is less than D[v], update D[v]=temporary distance, and record P[v]=u, wherein P is a predecessor node record array, and the predecessor node record array is used for subsequent path backtracking and stores the shortest path predecessor node of each grid mark point; A5.5, repeat step A5.3 until the terminal point t is moved to the S set, at this time, D[t] is the shortest path distance from the starting point to the terminal point; A5.6, start from the terminal point t, backtrace through the predecessor node array, arrange the grid mark point u sequence obtained by backtracking in order, and generate the candidate path of the distribution network line.

9. The grid-based distribution network line path automatic planning method based on digital twinning of claim 1, wherein, The generated adversarial network is a GAN network, training samples of the generated adversarial network include more than 3000 blurred and incomplete distribution network line drawings, the image information integrity of the repaired and completed distribution network line drawings is not less than 95%, and the identification and labeling of non-standard specifications and parameters in the distribution network line drawings include coordinates, sizes, device models and electrical parameters.

10. A computer device, comprising: The computer device includes a processor and a memory, and the memory stores at least one program code, the at least one program code is loaded and executed by the processor to use the computer device to realize the grid-based distribution network line path automatic planning method based on digital twinning as any one of claims 1-9.

Images