Method for constructing power distribution network planning digital twin system based on oblique photography technology

Abstract

The invention relates to a power distribution network planning digital twinborn system building method based on an oblique photography technology, and the method is characterized in that the method comprises the following steps: 1, power distribution region three-dimensional model reconstruction, 2, digital twinborn scene construction, and 3, power distribution network planning path optimization. According to the method, the oblique photography technology and the point cloud data collected by the unmanned aerial vehicle are utilized to construct the high-precision three-dimensional model of the power distribution network area, and the working efficiency of power distribution network planning is remarkably improved. The oblique photography technology can shoot ground images from multiple angles to generate a high-resolution three-dimensional model, the point cloud data provides accurate terrain information, and the combination of the oblique photography technology and the point cloud data ensures the high accuracy and detail richness of the model. Compared with traditional manual measurement and paper maps, the method has the advantages that the efficiency of data acquisition and model construction is greatly improved, meanwhile, the high consistency of the model and the actual environment is ensured, and the accuracy and feasibility of power distribution network planning are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system planning and construction, and in particular to a method for constructing a digital twin system for distribution network planning by using oblique photography. Background Art

[0002] In modern power systems, the planning and management of distribution networks are of great significance to ensure the safety and stability of power supply. Traditional distribution network planning methods usually rely on manual measurements and paper maps, which is not only inefficient but also difficult to meet the increasingly complex and sophisticated power network needs. In recent years, with the development of digital twin technology, this situation has been significantly improved. Digital twin refers to a virtual scene constructed by digitizing the real-world scene through a variety of advanced technologies such as physical models, Internet of Things technology, big data analysis, and virtual simulation, realizing the data and operation interoperability between the virtual and real worlds. With the help of digital twin technology, users can remotely understand the status of entities through the virtual world and perform operations with feedback, greatly improving the efficiency and accuracy of management and decision-making.

[0003] Compared with the traditional two-dimensional plane display method, the three-dimensional scene visual experience provided by digital twin technology is more intuitive and realistic, which can effectively improve the user's management efficiency of the target object and simplify maintenance work. For example, in the fields of smart cities and intelligent manufacturing, digital twins have shown strong application potential. However, although there have been preliminary attempts in the field of distribution network planning, there are still some problems, such as the display effect is not realistic enough and the virtual-real interaction function is limited.

[0004] As an emerging spatial information technology, oblique photogrammetry technology can collect ground image data from different angles simultaneously by carrying multiple sensors on a flying platform. After a professional data processing process, it can obtain accurate location information and rich texture data of ground objects and generate high-precision, high-realistic three-dimensional real-scene models. This technology has shown broad application prospects in many fields such as urban planning and disaster emergency response, especially in distribution network planning, which can provide a solid data foundation for the digital twin platform and further enhance the practicality and user experience of the system. Summary of the invention

[0005] In order to overcome the problems of poor implementation and frequent changes in the actual planning of existing distribution networks, the present invention proposes a method for building a digital twin scene for distribution network planning based on digital twin oblique photography technology. The technical scheme of the present invention is as follows: (1) reconstruction of a three-dimensional model of the distribution area; (2) construction of a digital twin scene; (3) optimization of the distribution network planning path.

[0006] The reconstruction of the three-dimensional model of the power distribution area specifically includes:

[0007] (1-1) Flight path planning and environmental assessment: Determine the mission objectives of drone oblique photography and apply for flight permission from the airspace management department. Conduct a comprehensive environmental assessment of the planning area, including factors such as topography, climate conditions, and obstacle distribution to ensure flight safety. Use professional GIS software to design the flight path of the drone based on the assessment results to ensure that all areas where data needs to be collected are covered. Path design needs to consider factors such as the drone's flight speed, flight altitude, and shooting angle to obtain the best image data. Evenly distribute a number of image control points in the planning area, mark them by spraying on the ground or selecting landmark features, and use RTK equipment for precise measurement to improve the accuracy of the model.

[0008] (1-2) Image data acquisition: Choose suitable weather conditions, equip a drone with a high-resolution camera, and ensure that the camera is working properly and the RTK module is connected correctly. Start the drone for aerial photography according to the predetermined flight path. During the shooting process, monitor the drone status in real time to ensure the integrity and accuracy of data collection. After the aerial photography is completed, export and back up the captured image data immediately to prevent data loss.

[0009] (1-3) 3D model reconstruction and restoration: Import the image data obtained by oblique photography into the image processing software, perform pre-processing such as denoising and splicing on the collected impact data to ensure data quality. Import the pre-processed image data into the 3D modeling software (ContextCapture), set the parameters and submit the aerial triangulation calculation to generate a high-precision 3D real scene model. Finally, the model is output in two file formats, OSGB and OBJ, to meet the needs of different application scenarios.

[0010] The construction of the digital twin platform specifically includes:

[0011] (2-1) System architecture and module design: Based on the WebGIS framework (such as Cesium or MapTalks), the architecture of the digital twin platform is constructed to support multi-terminal access. A user-friendly interface is designed to provide a variety of interactive function modules, such as model browsing, equipment query, and path planning. The optimized 3D model and distribution network data are loaded onto the platform to display the 3D real scene. The existing distribution network data (such as substation location, transmission line direction, equipment attributes, etc.) is imported into the 3D real scene model to achieve seamless data integration.

[0012] The distribution network planning path optimization specifically includes:

[0013] (3-1) Simulation environment establishment and parameter definition: The simulation environment of the study area was established using the grid method, which was divided into a 100x100 grid, with each grid size of 50m x 50m. Each unit was treated as a node, numbered and with coordinates determined. Then, key parameters including the starting point and the end point were defined, such as the construction difficulty coefficient, the cost of poles and wires, etc., and the calculation method of the total path length, difficulty and construction cost was set.

[0014] (3-2) Constraint setting: Different construction difficulty coefficients are set according to the complexity of the terrain, ranging from flat terrain suitable for power line planning to extremely complex detour requirements. At the same time, it is stipulated that the path must be a connected path from the starting point to the end point, and the Euclidean distance formula is used to ensure the accuracy of the path length calculation, ensuring that all path planning meets the actual construction conditions and technical requirements.

[0015] (3-3) Application of path planning algorithm: The A* algorithm is combined with the ant colony optimization algorithm to study the distribution network path planning in the research area. The ultimate optimization goal is to find a path that minimizes the weighted sum of the total planning difficulty and planning cost of the path. First, the A* algorithm is used to preliminarily generate a path as the initial solution of the ant colony optimization algorithm. After multiple iterations, the final optimal path is generated. The generated path is evaluated in terms of total length, planning cost, and planning difficulty to ensure the optimality and reliability of the path.

[0016] Compared with the prior art, the present invention has the following beneficial effects:

[0017] The present invention uses oblique photography technology and point cloud data collected by drones to construct a high-precision three-dimensional model of the distribution network area, significantly improving the work efficiency of distribution network planning. Oblique photography technology can capture ground images from multiple angles to generate high-resolution three-dimensional models, while point cloud data provides accurate terrain information. The combination of the two ensures the high accuracy and richness of details of the model. Compared with traditional manual measurement and paper maps, the present invention greatly improves the efficiency of data collection and model construction, while ensuring a high degree of consistency between the model and the actual environment, and improving the accuracy and feasibility of distribution network planning.

[0018] The present invention combines the A* algorithm and the ant colony optimization algorithm for path planning, making full use of the advantages of these two algorithms. The A* algorithm can quickly find the initial optimal path, while the ant colony optimization algorithm further optimizes the path through multiple rounds of iterations to ensure the optimality and robustness of the path. In addition, the digital twin platform supports dynamic adjustment and optimization of the path, and users can modify the path at any time according to actual conditions, which improves the flexibility and adaptability of the system. In the event of a failure or change, the system can respond quickly, provide new path planning and optimization suggestions, and ensure the stable operation of the distribution network.

[0019] In summary, the present invention realizes efficient, accurate and intelligent distribution network planning through the combination of oblique photography technology and digital twin technology. During the specific implementation process, each step is strictly carried out in accordance with the technical solution to ensure the reliability and practicality of the system. It is hoped that the present invention can provide a new solution for distribution network planning and promote technological progress in the power industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0021] Figure 1 It is a schematic diagram of the process of the present invention

[0022] Figure 2 This is a schematic diagram of the final effect of real-life 3D modeling.

[0023] Figure 3 This is the schematic diagram of the comprehensive homepage of the digital twin system

[0024] Figure 4 This is a schematic diagram of the regional information presentation of the digital twin system.

[0025] Figure 5 This is a schematic diagram of the digital twin system path planning and self-selection path

[0026] Figure 6 This is a schematic diagram of the preset route for path planning of the digital twin system.

[0027] Figure 7 This is a schematic diagram of the information pop-up window of the digital twin system route planning interface

[0028] Figure 8 This is a preview diagram of the digital twin system circuit.

[0029] Fig. 9 This is a flow chart of the A* algorithm.

[0030] Fig.10 It is a schematic diagram of the specific implementation process of distribution network planning path optimization. DETAILED DESCRIPTION

[0031] In order to more clearly understand the core idea, technical features and expected effects of the present invention, the present invention will be described in detail below in conjunction with the accompanying drawings and embodiments. It should be particularly noted that, under the premise of not conflicting with the core idea of ​​the present invention, the various embodiments and their features in this application can be used in combination with each other.

[0032] In the following description, the specific implementation details of the present invention will be described in detail so as to fully understand the technical solution of the present invention. However, it should be understood that the present invention is not limited to the specific implementation methods described below, but can be implemented in other different ways, so the protection scope of the present invention is not limited by the following specific embodiments.

[0033] Example: See Figures 1 to 8 The present invention relates to a method for constructing a distribution network planning digital twin system using unmanned aerial vehicle (UAV) oblique photography to construct a three-dimensional model, comprising: (1) reconstruction of a three-dimensional model of a distribution area; (2) construction of a digital twin scene; and (3) distribution network planning path optimization.

[0034] The step (1) specifically comprises:

[0035] (1-1) Flight path planning and environmental assessment: To ensure the successful execution of the UAV oblique photography mission, the specific mission objectives and expected results were first determined, and a detailed flight plan was submitted to the local airspace management department to apply for the necessary flight permit. Next, a comprehensive environmental assessment of the planned area was conducted, taking into account factors such as topography, climatic conditions (such as wind speed and humidity), and obstacle distribution to ensure flight safety and data collection effectiveness. Using professional geographic information system (GIS) software, the UAV flight path was designed based on the results of the environmental assessment. The path was designed to cover all key areas where data needs to be collected, while optimizing parameters such as flight speed, altitude, and shooting angle to obtain high-quality image data. To improve the accuracy of the 3D model, a number of image control points were set in the planned area. These image control points were marked by ground spraying or selecting landmarks, and were accurately measured using a real-time dynamic differential global positioning system (RTK) device to ensure the position accuracy of each image control point, thereby improving the spatial accuracy of the final 3D model. Before the actual flight, a final check was also conducted to confirm that the UAV and its onboard high-resolution camera and RTK module were working properly and that all settings met the predetermined standards. In addition, we also arranged special personnel to monitor the status of the drone and the changes in the surrounding environment in real time to ensure that any emergency situations can be responded to in a timely manner during the entire flight process, to ensure flight safety and to ensure the integrity of data collection. Through the above careful preparation and strict control, we successfully completed the data collection task of the drone oblique photography, laying a solid foundation for the subsequent construction of an accurate three-dimensional model of the power distribution area.

[0036] (1-2) Image data collection: Use DJI M3E drone for data collection. Conduct flight surveys in the planning area, use markup language to draw the aerial photography area, and ensure that the target area is covered. After the drone is loaded with the markup language KML, it can automatically plan routes in the area; before flying, check the surrounding environment to determine whether the on-site environment meets the flight requirements, understand and comply with local drone flight regulations, and ensure that there are no no-fly restrictions in the flight area. When the conditions are met, assemble the drone. After the drone self-checks, check whether the network RTK module is connected normally. Fly according to the oblique photogrammetry scheme configured by DJI M3E to complete the image collection of the planning area. After the collection is completed, calibrate the camera parameters and prepare the camera's internal orientation elements (such as focal length, sensor size) and external orientation elements (such as shooting position and attitude information).

[0037] (1-3) 3D model reconstruction and repair: The original data is exported from the drone and preprocessed, including removing redundant data and checking data integrity. The Context Capture software is used to load data, associate the corresponding camera metadata to set pose, perform the first aerial triangulation, adjust feature point matching parameters, perform the second aerial triangulation, create a new 3D solid model, and generate original 3D real scene model data in context capture 3MX and OBJ formats. To ensure the model accuracy and flexible adjustment of the reconstruction range of the target area, the final output is data in context capture 3MX (default project file storage format) and OBJ (3D model file format). The first aerial triangulation is used to calculate the exterior orientation elements of each photo. After the first aerial triangulation calculation is completed, the image matching quality and reprojection error are checked, and the feature point matching parameters are adjusted or control points are manually added if necessary. The second aerial triangulation is used to improve the calculation accuracy of the exterior orientation elements of the photo and correct the model. After the 3D solid model is generated, the model is evaluated for accuracy, including checking the model's geometry and texture quality. Due to the technical characteristics of oblique photography, some details of the model are lost. Although the accuracy of the model can be improved by flying at a lower altitude and high-level stratification, the local texture of the model is missing due to factors such as mutual occlusion between objects, and there are model holes and texture distortion. Therefore, it is necessary to repair the model and use Blender software to optimize the original 3D real-life model. Import the OBJ format file into Blender, and restore the structure of the hollow area in the model by manual modeling or texture filling. Among them, manual modeling needs to draw the missing geometric shapes based on high-definition satellite maps and drone aerial image data; texture filling needs to extract textures from other similar areas and apply them to the hollow area. Texture mapping adjustment is performed for the texture distortion part of the model, and textures can be re-collected for areas with poor texture quality. After the model repair is completed, the geographical features of the planning area are ensured to be clearly identifiable. After the model is repaired, the model is optimized by reducing the number of model faces to improve the running efficiency and rendering effect of the model. Continue to bake and render the repaired 3D model of the planning area in Blender. Before rendering, set the material of the model, including parameters such as color, glossiness, and transparency. After the material settings are completed, configure the world environment and local lighting, adjust the position, color, intensity and other parameters of the light source according to the project requirements, and simulate the real lighting effect. During the rendering process, adjust the rendering quality, resolution, anti-aliasing and other parameters according to the project requirements. Render the 3D scene required by the visualization platform and export the 3D model in the general FBX format. Finally, the required model is obtained. Figure 2 .

[0038] The step (2) specifically comprises:

[0039] (2-1) System architecture and module design: In order to support the efficient planning of distribution network, the present invention builds a multifunctional digital twin platform based on WebGIS framework (such as Cesium or MapTalks). The platform adopts modular design and integrates multiple functional modules to meet different operation requirements, including but not limited to model browsing, equipment query, path planning, etc. The user interface is friendly and intuitive, supporting multi-terminal access (PC, tablet, etc.), so that operators can view and manage the three-dimensional real-life model of the distribution area anytime and anywhere. The platform also pays special attention to user experience, adopts the yellow-green color scheme of the power grid ecology with a dark background, reduces visual fatigue, and is suitable for long-term operation. In addition, data is presented in multiple dimensions through charts, lists, and status identifiers, helping users to quickly capture key information and make decisions. The system function adopts a modular design form, mainly with the functions of power grid path planning and integrated regional information. Based on this, four modules are designed, namely the homepage module, regional information module, path planning module, and interaction and decision-making module. The system built on the WebGIS framework supports multi-terminal access. Operators can view the real-life model of the power grid area through PC, tablet and other devices, and perform operations such as path planning and information query.

[0040] (2-1-1) Comprehensive homepage design: The comprehensive homepage is designed to allow operators to quickly grasp the basic information of the distribution area. Figure 3 .In terms of content design, the fixed panel on the left is designed to display the basic information of the target distribution area, the regional overview, including terrain, population, area, etc.; the environmental monitoring module displays the climate indicators of the target area, which is convenient for the selection of laying lines; the grid information design concisely displays the overall distribution of each line and wiring specifications in the target area. The right panel specifically displays the statistics of the field scene and the main transmission nodes such as substations, and also contains the alarms and notifications of block information, such as fault alarms, risk warning notifications and custom alarm settings. The bottom data module displays the overall load operation status of the region. The three-dimensional scene display module in the central area is designed to display the three-dimensional location distribution and status of the site real-life model, transmission towers, and substations, and provide detailed equipment attribute data. Clicking the block provides free roaming function in the three-dimensional scene, the ability to observe paths and equipment from different perspectives, helping operators understand the distribution of distribution lines from an overall perspective, and supports the selection, rotation and scaling of equipment to facilitate operators to view equipment information.

[0041] (2-1-2) Visual and interactive design: The system interface uses the yellow-green color scheme of the power grid ecology, with a dark background, which creates contrast and reduces visual fatigue, making it suitable for long-term operation. Data is presented in multiple dimensions through charts, lists, and status identifiers. The line chart dynamically displays the time trend of the load status; the icon form intuitively displays the equipment operation status for quick identification.

[0042] (2-1-3) Regional information module: This module provides block information and power equipment information, see Figure 4 . The left panel displays the relevant data parameters that need to be considered when laying the line in the target block. Including field scene data, environmental climate information of the target area, and line load information of the target area. The right panel further displays the power equipment information, including the data of the substation equipment in the project block (such as voltage, current, temperature, etc.), tower and cable status data. The central area is centered on the real-life model, and provides data monitoring and operation information around the core. When the mouse is hovered or clicked, detailed information of the target object is displayed, such as farmland area, residential area, commercial area, etc., line name, etc. Visually load the terrain information of the area, and obstacle avoidance highlights the location of obstacles in path planning. Modal pop-ups and fixed windows are used to display key information prompts, such as government land, gas stations, etc., to provide reliability and adaptability support for path planning.

[0043] (2-1-4) Path planning and simulation module: This module provides custom pipeline laying, line connection and planning simulation functions, which are divided into three operation functions: path selection, pipeline preset and level switching. Users can select the route to be planned in the scene model, create custom pipeline information, and after the pipeline is laid, the system automatically connects the newly created pipeline to form a preliminary route, which can be previewed or adjusted in full view. The preset route scene supports previewing, adjusting and displaying multiple routes of saved pipeline routes, and pops up the estimated details of route parameters. Level switching supports switching between global view and local view to display resource distribution.

[0044] (2-1-5) Interactive decision support module: This module provides a preview of the preset plan, which is divided into three functions: path selection, optimal path, and level switching. The path selection module displays the length, type, cost, and loss information of each path for operators to quickly compare; the optimal path module displays alternative paths and provides accuracy feedback in combination with the government's established planning area; the level switching module provides a bird's-eye view, supports the switching of planning previews between the global view and the bird's-eye view, and improves visual planning performance.

[0045] The step (3) specifically includes:

[0046] (3-1) Simulation environment establishment and parameter definition: In order to effectively optimize the distribution network planning path, it is necessary to first establish a simulation environment for the study area. In this process, we selected a specific area in Shaoxing City as the research object. The map data of this area was obtained through high-resolution satellite images and GIS systems to ensure the accuracy and detail of the map.

[0047] (3-1-1) Grid processing: We used the grid method to model the study area. Specifically, the study area was divided into several 50m x 50m square grid cells. The center point of each grid cell is regarded as a node, which is used to represent the key position in the path. This processing method not only simplifies the computational complexity, but also provides sufficient accuracy to meet the actual engineering needs. In order to facilitate subsequent path planning and optimization calculations, we numbered each grid cell. The numbering rule is in row and column order, that is, the grid cell in the i-th row and the j-th column is numbered (i, j). The coordinates (x, y) of the node can be calculated by the grid number.

[0048] (3-1-2) Node adjacency method: Each node can be connected to the eight nodes around it, including four directions of up, down, left, right, and four diagonal directions. This eight-connected method ensures the flexibility and diversity of path planning and helps to find a better path solution.

[0049] (3-1-3) Parameter definition: G = (V, E): represents the grid map of the study area, where V is the set of grid cells and E is the set of edges connecting the center points of two adjacent grid cells. s = (x s ,y s ) and t=(x t ,y t ):represents the coordinates of the starting point and the end point of the path respectively. d(v): represents the planning difficulty coefficient of the grid unit v=(x,y), and the value is 0-1, 1.5 or 2. c r : represents the cost of each electric pole. w : Indicates the cost of each meter of wire. c : Indicates the cost of cable per meter. d and w c : Represents the weights of planning difficulty and planning cost respectively P: Represents the path from the starting point s to the end point t. The path consists of the coordinates of the center points of a series of grid cells, denoted as P = {(x1, y1), (x2, y2)…(x n ,y n )}. L(P): indicates the total length of path P. D(P): indicates the total planning difficulty of path P. C(P): indicates the total construction cost of path P. n r : represents the number of electric poles in path P. L w: represents the total length of the wires in path P. L c : Indicates the total length of the cable in path P.

[0050] (3-2) Setting Constraints

[0051] (3-2-1) Construction Difficulty Coefficient: Construction Difficulty Value Setting: In order to more accurately reflect the impact of actual terrain and environmental conditions on path planning, we assign a construction difficulty value d(v) to each grid cell. These difficulty values ​​are derived by domain experts based on a comprehensive evaluation of the actual terrain, environment, and other factors that affect planning, ensuring the reliability and practicality of the data. The specific difficulty value range is 0-1, 1.5, or 2, and the specific meanings are as follows:

[0052] 0≤d(v)≤1: indicates flat terrain, suitable for planning with wires

[0053] d(v)=1.5: Indicates that the terrain is relatively complex and is suitable for planning with cables

[0054] d(v)=2: indicates that the terrain is extremely complex and power distribution planning cannot be carried out, and a detour is required.

[0055] (3-2-2) Path connectivity: Path P must be a connected path from the starting point s to the end point t:

[0056]

[0057] (3-2-3) Path length calculation: The total length of path P, L(P), is the sum of the lengths of all edges on the path.

[0058]

[0059] (3-2-4) Optimization goal: In order to make the path planning more reasonable, we define the total planning difficulty D(P) and planning cost C(P) of the path, and introduce the corresponding weight coefficient w d and w c The ultimate goal is to find a path P that minimizes the weighted sum of the total planning difficulty and planning cost of the path. The objective function F(P) is defined as:

[0060] min F(P)=w d ·D(P)+w C ·C(P)

[0061] The planning difficulty D(P) is the sum of the difficulty values ​​of the grid cells of each path on the path:

[0062]

[0063] The planning cost C(P) is determined by the number of poles, the length of wires and the length of cables:

[0064] C(P)=c r ·n r +c w ·L w +c c ·L c

[0065] (3-2-5) Use the A* algorithm to generate an initial path tendency: We first use the A* algorithm to generate an initial path. The A* algorithm is a heuristic search algorithm that finds the shortest path by combining the actual distance and the estimated distance. Fig. 9 , the specific steps are as follows:

[0066] (3-2-6) Initialization: After gridding the study area, assign a planning difficulty coefficient d(x,y) to each node. Then use the Euclidean distance as the heuristic function h(n), that is, Where (x t ,y t ) is the coordinate of the target node, (x n ,y n ) is the coordinate of the current node. The actual cost g(n) from the starting node 8 to the current node n needs to consider the planning difficulty coefficient d(v), that is, g(n)=g(parent(n))+d(v)×distance(parent(n),n), where distance(parent(n),n) is the distance between two adjacent nodes. The open list and closed list are initialized. The open list (OpenList) is used to store the nodes to be processed, and the closed list (Closed List) is used to store the nodes that have been processed.

[0067] (3-2-7) Execution: Set the starting node s = (x s ,y s ) into the open list. When the open list is not empty, repeat the following steps. First, select the node n with the lowest f(n) value from the open list. If n is the target node t=(x t ,y t), then end the search and backtrack the path. Remove n from the open list and add it to the closed list. For each neighbor node m of n, if m is not accessible (difficulty coefficient is 2), skip this node; if m is already in the closed list, skip this node; calculate the actual cost g(n) from s to m, calculate the heuristic cost h(n) of m, and calculate the total cost f(m) of m, f(m) = g(m) + h(m). If m is not yet in the open list, add it to the open list and set n as the parent node of m. If m is already in the open list, but the path to m through n is better than the previously recorded path, update the parent node of m to n and update g(m) and f(m) of m. If the target node t is found, backtrack the path from s to t, and this path is the required preliminary path. If the open list is empty and the target node is not found, it means that there is no feasible path.

[0068] (3-2-8) Path backtracking: Once the target node t is added to the closed list, we can start from t and trace back to the starting node s along the parent node pointer of each node to build a complete path.

[0069] (3-3) Application of path planning algorithm: In order to further optimize the path, we used the ant colony optimization algorithm (ACO). The ant colony optimization algorithm is a meta-heuristic algorithm based on the foraging behavior of ants in nature and is suitable for solving combinatorial optimization problems. Fig.10 , the specific steps are as follows:

[0070] (3-3-1) Initialization: Set the number of ants N = 100, the pheromone volatility rate ρ = 0.5, the pheromone importance factor α = 1, the heuristic information factor β = 1, and initialize the pheromone concentration τ on all paths to 0.1.

[0071] (3-3-2) Path generation: Each ant starts from the starting point s and selects the next node according to the difficulty coefficient and pheromone concentration from the current node to the adjacent node until it reaches the end point t. The probability of selecting the next node is p ij Calculated by the following formula:

[0072]

[0073] Among them, τ ij is the pheromone concentration from node i to node j, η ij is the heuristic information from node i to node j (3-3-3) Path evaluation: Calculate the total planning difficulty D(P) and planning cost C(P) of the path generated by each ant. Calculate the total cost of each path F(P) = w d ·D(P)+w c ·C(P).

[0074] (3-3-4) Pheromone update: Update the pheromone concentration on the path. The pheromone concentration of the high-quality path increases, and the pheromone concentration of the low-quality path decreases. The pheromone update formula is:

[0075] τ ij =(1-ρ)·τ ij +Δτ ij

[0076] in, is the amount of pheromone left by the kth ant on the path (i, j), and the calculation formula is

[0077]

[0078] Among them, Q is a constant used to control the size of the pheromone increment.

[0079] (3-3-5) Iteration: Repeat the above steps until the predetermined number of iterations is reached or the stopping condition is met (for example, the path has not been significantly improved in several consecutive iterations) to obtain the final optimal planning path.

[0080] The description of each embodiment in the above embodiments has different emphases. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0081] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for building a digital twin system for distribution network planning based on oblique photography technology, characterized in that: The method comprises the following steps, Step 1: Reconstruct the 3D model of the power distribution area. Step 2: Digital twin scenario construction, Step 3: Distribution network planning path optimization.

2. According to claim 1, a method for building a digital twin system for distribution network planning based on oblique photography technology is characterized in that: Step 1: Reconstruct the 3D model of the power distribution area, including: (1-1) Flight path planning and environmental assessment: Determine the mission objectives of the drone oblique photography, apply for flight permission from the airspace management department, conduct a comprehensive environmental assessment of the planned area, including topography, climate conditions, and obstacle distribution factors to ensure flight safety, use professional GIS software, design the drone's flight path based on the assessment results, and ensure that all areas where data needs to be collected are covered. The path design needs to consider the drone's flight speed, flight altitude, and shooting angle factors to obtain the best image data. Evenly distribute a number of image control points in the planning area, mark them by spraying on the ground or selecting landmarks, and use RTK equipment for precise measurement to improve the accuracy of the model. (1-2) Image data acquisition: Choose suitable weather conditions, equip the drone with a high-resolution camera, ensure that the camera is working properly and the RTK module is connected correctly, start the drone for aerial photography according to the predetermined flight path, monitor the drone status in real time during the shooting process to ensure the integrity and accuracy of data collection, and immediately export and back up the captured image data after the aerial photography is completed to prevent data loss. (1-3) 3D model reconstruction and restoration: Import the image data obtained by oblique photography into the image processing software, perform denoising and splicing preprocessing on the collected impact data to ensure data quality, import the preprocessed image data into the 3D modeling software (ContextCapture), set the parameters and submit the aerial triangulation calculation to generate a high-precision 3D real-life model, and finally output the model in two file formats, OSGB and OBJ, to meet the needs of different application scenarios.

3. According to the method for building a digital twin system for distribution network planning based on oblique photography technology in claim 1, it is characterized in that: Step 2: Digital twin scenario construction, including: (2-1) System architecture and module design: Based on the WebGIS framework, the digital twin platform architecture is constructed to support multi-terminal access, design a friendly user interface, and provide a variety of interactive function modules including model browsing, equipment query, and path planning. The optimized three-dimensional model and distribution network data are loaded onto the platform to realize the display of three-dimensional real scenes. The existing distribution network data is imported into the three-dimensional real-scene model to achieve seamless data integration.

4. According to the method for building a digital twin system for distribution network planning based on oblique photography technology in claim 1, it is characterized in that: Step 3: Distribution network planning path optimization, as follows: (3-1) Simulation environment establishment and parameter definition: The simulation environment of the study area was established using the grid method, which was divided into a 100x100 grid. The size of each grid was 50 m x 50 m. Each unit was regarded as a node, numbered and determined the coordinates. Then, key parameters including the start and end points were defined. (3-2) Constraint setting: Different construction difficulty coefficients are set according to the complexity of the terrain, ranging from flat terrain suitable for wire planning to extremely complex detour requirements. At the same time, it is stipulated that the path must be a connected path from the starting point to the end point, and the Euclidean distance formula is used to ensure the accuracy of the path length calculation, ensuring that all path planning meets the actual construction conditions and technical requirements. (3-3) Application of path planning algorithm: The A* algorithm is combined with the ant colony optimization algorithm to study the distribution network path planning in the study area. The ultimate optimization goal is to find a path that minimizes the weighted sum of the total planning difficulty and planning cost of the path. First, the A* algorithm is used to preliminarily generate a path as the initial solution of the ant colony optimization algorithm. After multiple iterations, the final optimal path is generated. The generated path is evaluated in terms of total length, planning cost, and planning difficulty to ensure the optimality and reliability of the path.

5. According to claim 3, a method for building a digital twin system for distribution network planning based on oblique photography technology is characterized in that: Step (2-1) The system architecture and module design are as follows: (2-1-1) Comprehensive homepage design: The comprehensive homepage is designed to allow operators to quickly grasp the basic information of the distribution area. In terms of content design, the fixed panel on the left is designed to display the basic information of the target distribution area, regional overview, including terrain, population, and area. The environmental monitoring module displays the climate indicators of the target area, which is convenient for the selection of laying lines; the grid information design concisely displays the overall distribution of each line and wiring specifications in the target area. The right panel specifically displays the statistics of the actual scene and the main transmission nodes such as substations, and also contains the alarms and notifications of the block information. The bottom data module displays the overall load operation status of the area. The three-dimensional scene display module in the central area is designed to display the three-dimensional location distribution and status of the site real-life model, transmission towers, and substations, and provide detailed equipment attribute data. Clicking the block provides the free roaming function in the three-dimensional scene, the ability to observe the path and equipment from different perspectives, helping the operator to understand the distribution of the distribution lines from an overall perspective, and supports the selection, rotation and scaling of equipment to facilitate the operator to view equipment information. (2-1-2) Visual and interactive design: The system interface color scheme uses the yellow-green color scheme of the power grid ecology, with a dark background to create contrast and reduce visual fatigue, making it suitable for long-term operation. Data is presented in multiple dimensions through charts, lists, and status identifiers. The line chart dynamically displays the time trend of the load status; the icon form intuitively displays the equipment operation status, which is easy to quickly identify. (2-1-3) Regional information module: This module provides block information and power equipment information. The left panel displays the relevant data parameters that need to be considered when laying the line in the target block, including field scene data, environmental climate information of the target area, and line load information of the target area. The right panel further displays power equipment information, including data of substation equipment in the project block, tower and cable status data. The central area is centered on the real-life model, and provides data monitoring and operation information around the core. When the mouse is hovered or clicked, detailed information of the target object is displayed, including farmland area, residential area, commercial area, line name, and visual loading area terrain information. Obstacle avoidance highlights the location of obstacles in path planning, and uses modal pop-up windows and fixed windows to display key information prompts, providing reliability and adaptability support for path planning. (2-1-4) Path planning and simulation module: This module provides custom pipeline laying, line connection and planning simulation functions, which are divided into three operation functions: path selection, pipeline preset and level switching. Users can select the route to be planned in the scene model and create custom pipeline information. After the pipeline is laid, the system automatically connects the newly created pipeline to form a preliminary route, and can preview or adjust it in a panoramic view. The preset route scene supports previewing, adjusting and displaying multiple routes of saved pipeline routes, popping up the estimated details of route parameters, and level switching supports switching between global view and local view to display resource distribution. (2-1-5) Interactive decision support module: This module provides a preview of preset plan plans, which is divided into three functions: path comparison, optimal path and level switching. The path comparison module displays the length, type, cost and loss information of each path for operators to quickly compare; the optimal path module displays alternative paths and provides accuracy feedback in combination with the government's established planning areas; the level switching module provides a bird's-eye view and supports the switching of planning previews between the global view and the bird's-eye view, thereby improving visual planning performance.

6. According to claim 4, a method for building a digital twin system for distribution network planning based on oblique photography technology is characterized in that: (3-1) Simulation environment establishment and parameter definition, as follows: (3-1-1) Grid processing: The grid method is used to model the study area. Specifically, the study area is divided into several 50m×50m square grid cells. The center point of each grid cell is regarded as a node to represent the key position in the path. Each grid cell is numbered in the order of rows and columns, that is, the grid cell in the i-th row and the j-th column is numbered (i, j). The coordinates (x, y) of the node can be calculated by the grid number. (3-1-2) Node adjacency: Each node is connected to the eight nodes around it, including four directions: up, down, left, right, and four diagonal directions. (3-1-3) Parameter definition: G = (V, E): represents the grid map of the study area, where V is the set of grid cells, E is the set of edges connecting the center points of two adjacent grid cells, and s = (x s ,y s ) and t=(x t ,y t ):represent the coordinates of the starting point and the end point of the path, d(v): represents the grid unit v=( x， y) planning difficulty coefficient, which can be 0-1, 1.5 or 2, c r : represents the cost of each electric pole, c w : represents the cost per meter of wire, c c : represents the cost of cable per meter, w d and w c : Represents the weights of planning difficulty and planning cost respectively P: Represents the path from the starting point s to the end point t. The path consists of the coordinates of the center points of a series of grid cells, denoted as P = {(x1, y1), (x2, y2)…(x n ,y n )}, L(P): represents the total length of path P, D(P): represents the total planning difficulty of path P, C(P): represents the total construction cost of path P, n r : represents the number of poles in path P, L w : represents the total length of the wires in path P, L c : Indicates the total length of the cable in path P.

7. According to claim 4, a method for building a digital twin system for distribution network planning based on oblique photography technology is characterized in that: (3-2) Constraint setting, as follows: (3-2-1) Construction Difficulty Coefficient: Construction Difficulty Value Setting: Each grid cell is assigned a construction difficulty value d(v). These difficulty values ​​are determined by field experts based on a comprehensive assessment of the actual terrain, environment, and other factors that affect planning, ensuring the reliability and practicality of the data. The specific difficulty value range is 0-1, 1.5, or 2. The specific meanings are as follows: 0≤d(v)≤1: indicates that the terrain is flat and suitable for planning with wires. d(v)=1.5: indicates that the terrain is relatively complex and is suitable for planning with cables. d(v)=2: indicates that the terrain is extremely complex and power distribution planning cannot be carried out, and a detour is required. (3-2-2) Path connectivity: Path P must be a connected path from the starting point s to the end point t: (3-2-3) Path length calculation: The total length of path P, L(P), is the sum of the lengths of all edges on the path. (3-2-4) Optimization goal: Define the total planning difficulty D(P) and planning cost C(P) of the path, and introduce the corresponding weight coefficient w d and w c , the ultimate goal is to find a path P that minimizes the weighted sum of the total planning difficulty and planning cost of the path. The objective function F(P) is defined as: min F(P)=w d ·D(P)+w C ·C(P) The planning difficulty D(P) is the sum of the difficulty values ​​of the grid cells of each path on the path: The planning cost C(P) is determined by the number of poles, the length of wires and the length of cables: C(P)=c r ·n r +c w ·L w +c c ·L c (3-2-5) Use the A* algorithm to generate an initial path tendency: We first use the A* algorithm to generate an initial path. The A* algorithm is a heuristic search algorithm that finds the shortest path by combining the actual distance and the estimated distance. The specific steps are as follows: (3-2-6) Initialization: After gridding the study area, assign a planning difficulty coefficient d(x, y) to each node, and then use the Euclidean distance as the heuristic function h(n), that is, Where (x t ,y t ) is the coordinate of the target node, (x n ,y n ) is the coordinate of the current node. The actual cost g(n) from the starting node 8 to the current node n needs to consider the planning difficulty coefficient d(v), that is, g(n)=g(parent(n))+d(v)×distance(parent(n),n), where distance(parent(n),n) is the distance between two adjacent nodes. The open list and the closed list are initialized. The open list (OpenList) is used to store the nodes to be processed, and the closed list (Closed List) is used to store the nodes that have been processed. (3-2-7) Execution: Set the starting node s = (x s ,y s ) into the open list. When the open list is not empty, repeat the following steps. First, select the node n with the lowest f(n) value from the open list. If n is the target node t=(x t ,y t ), then the search ends, the path is backtracked, n is removed from the open list, and added to the closed list. For each neighbor node m of n, if m is not accessible (the difficulty coefficient is 2), this node is skipped; If m is already in the closed list, skip this node; Calculate the actual cost g(n) from s to m, calculate the heuristic cost h(n) of m, calculate the total cost f(m), f(m) = g(m) + h(m), if m is not in the open list, add it to the open list and set n as the parent node of m, if m is already in the open list, but the path to m through n is better than the previously recorded path, update the parent node of m to n and update g(m) and f(m) of m, if the target node t is found, then backtrack the path from s to t, this path is the required preliminary path, if the open list is empty and the target node is not found, it means that there is no feasible path, (3-2-8) Path backtracking: Once the target node t is added to the closed list, start from t and trace back to the starting node s along the parent node pointer of each node to build a complete path.

8. According to claim 4, a method for building a digital twin system for distribution network planning based on oblique photography technology is characterized in that: (3-3) Application of path planning algorithm: The specific steps are as follows: (3-3-1) Initialization: Set the number of ants N = 100, the pheromone volatility ρ = 0.5, the pheromone importance factor α = 1, the heuristic information factor β = 1, and the pheromone concentration τ on all paths = 0.

1. (3-3-2) Path generation: Each ant starts from the starting point s and selects the next node according to the difficulty coefficient and pheromone concentration from the current node to the adjacent node until it reaches the end point t. The probability of selecting the next node is p ij Calculated by the following formula: Among them, τ ij is the pheromone concentration from node i to node j, η ij is the heuristic information from node i to node j (3-3-3) Path evaluation: Calculate the total planning difficulty D(P) and planning cost C(P) of the path generated by each ant, and calculate the total cost of each path F(P) = w d ·D(P)+w c C(P), (3-3-4) Pheromone update: Update the pheromone concentration on the path. The pheromone concentration on the high-quality path increases, while the pheromone concentration on the low-quality path decreases. The pheromone update formula is: t ij =(1-ρ)·τ ij +ΔT ij in, is the amount of pheromone left by the kth ant on the path (i, j), and the calculation formula is Among them, Q is a constant used to control the size of the pheromone increment. (3-3-5) Iteration: Repeat the above steps until the predetermined number of iterations is reached or the stopping condition is met (for example, the path has not been significantly improved in several consecutive iterations) to obtain the final optimal planning path.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements a method for building a digital twin system for distribution network planning based on oblique photography technology as described in any one of claims 1 to 8 above.

10. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the computer instruction is executed by the processor, a method for building a digital twin system for distribution network planning based on oblique photography technology is implemented as described in any one of claims 1-8.

Images