Power distribution network site selection determination method and system

Abstract

The invention discloses a power distribution network site selection determination method and system, and the method comprises the steps: segmenting a remote sensing image through a semantic segmentation model and a sliding window, and rapidly obtaining a terrain image, thereby recognizing the terrain features based on a supervised model and an unsupervised model, improving the recognition accuracy of the terrain features, and meanwhile, improving the recognition efficiency. Determining the electric power influence range of each terrain picture in the site selection area based on the electric power demand and the radiation circle in the site selection area to calculate the site selection score, and quantifying the electric power demand and the terrain advantage to ensure the scientificity and rationality of site selection; and the site selection position and the position information are automatically matched based on the site selection score, so that automation and high efficiency of the site selection process are realized, the site selection cost is effectively reduced, and the site selection efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power planning, and in particular to a method and system for determining the site selection of a distribution network site. Background Art

[0002] A distribution network site refers to a facility in a power system that converts high-voltage electricity into low-voltage electricity and distributes it to end users. It is mainly used for power conversion, power distribution, power fault handling and load management. The site selection of distribution network sites plays a key role in improving power transmission efficiency, reducing costs, improving power supply reliability, ensuring system safety, meeting power demand, optimizing load management and promoting regional development. Correct site selection can significantly improve the overall performance and economic benefits of the power system. The site selection of distribution network sites is affected by many factors. Reasonable site selection can reduce the distance of transmission lines, thereby reducing the loss of electricity during transmission.

[0003] The existing site selection method for distribution network sites relies on the experience of engineers or planners to select sites. Relatively stable and suitable locations are selected based on historical data and experience, but the labor cost is high, the time is long, and the efficiency of site selection for distribution network sites is low. Summary of the invention The present invention provides a method and system for determining the site selection of distribution network sites, so as to reduce the site selection cost of distribution network sites and improve the site selection efficiency of distribution network sites.

[0004] In order to solve the above technical problems, an embodiment of the present invention provides a method for determining the site selection of a distribution network site, including: Acquire remote sensing images within the selected area, segment the remote sensing images based on a semantic segmentation model and a sliding window, and acquire a terrain image set and corresponding location information; Inputting the terrain image set into the supervised model and the unsupervised model respectively for feature recognition to obtain the terrain features of each terrain image; Determine a plurality of power demand points and a load center point within the site selection area based on the power demand data within the site selection area; Taking the center point of the terrain image as the midpoint, determining the radiation circle of each terrain image based on a preset radius length and the midpoint; and calculating the load score of each terrain image based on the radiation circle, the power demand point and the load center point; The site selection score of each terrain image is calculated based on the load score and the terrain characteristics, the site selection location is determined based on the site selection score, and the location information of the site selection location is determined to complete the site selection of the distribution network site.

[0005] The present invention segments remote sensing images through a semantic segmentation model and a sliding window to quickly obtain terrain images, thereby jointly identifying terrain features based on supervised and unsupervised models to improve the accuracy of terrain feature recognition. At the same time, the power influence range of each terrain image in the site selection area is determined based on the power demand and radiation circle in the site selection area to calculate the site selection score, quantify the power demand and terrain advantages, and ensure the scientificity and rationality of the site selection; and automatically matches the site selection location and location information based on the site selection score, thereby realizing the automation and efficiency of the site selection process, thereby effectively reducing the site selection cost and improving the site selection efficiency.

[0006] Furthermore, the determining of a plurality of power demand points and a load center point in the site selection area based on the power demand data in the site selection area includes: Acquiring power demand data in the site selection area, wherein the power demand data includes power distribution data and load demand data; constructing a power distribution map based on the power distribution data, and determining a plurality of power demand points based on the power distribution map and a preset power distribution threshold; A load forecasting model is constructed based on the regression model, and the load center point is determined based on the load demand data and the load forecasting model.

[0007] Furthermore, the calculation of the load score of each terrain image based on the radiation circle, the power demand point and the load center point includes: Count the number of power demand points in each radiation circle, and calculate the load distance of each radiation circle based on the midpoint of each radiation circle and the load center point; The load score of each terrain image is calculated based on the number of power demand points, the load distance and the load score calculation formula. Furthermore, the load score calculation formula is specifically:

[0008] in, represents the load score, is the number of power demand points within the current radiation circle, is the total number of power demand points in the site selection area, is the load distance, and They are the power demand point coefficient and the load distance coefficient respectively.

[0009] Furthermore, the calculating the site selection score of each terrain image based on the load score and the terrain feature includes: Acquire terrain features of each terrain image, and match terrain scores based on the terrain features; Based on the load score and the terrain score, a weighted average is performed to calculate the site selection score of each terrain image; specifically:

[0010] in, represents the site selection score, is the load score, Score for terrain, and are terrain score coefficient and load score coefficient, respectively, which are obtained based on regression analysis, and .

[0011] Furthermore, the remote sensing image is segmented based on the semantic segmentation model and the sliding window to obtain a number of terrain images and corresponding location information, including: Constructing an initial semantic segmentation model, and training the initial semantic segmentation model based on remote sensing annotation data, and outputting a trained semantic segmentation model; Initialize a sliding window based on a preset window size and step size, segment the remote sensing image based on the sliding window, and obtain a first segmented image set; The first segmented picture set is input into the semantic segmentation model for feature recognition, terrain information of each first segmented picture is extracted, a terrain picture set is generated, and location information corresponding to each terrain picture is obtained.

[0012] Furthermore, the terrain image set is input into the supervised model and the unsupervised model respectively for feature recognition to obtain the terrain features of each terrain image, including: Building a supervised model based on a vector machine, and inputting a terrain picture set into the supervised model to obtain a first recognition result of the supervised model; Building an unsupervised model based on a generative adversarial network, and inputting a terrain picture set into the unsupervised model to obtain a second recognition result of the unsupervised model; The first recognition result and the second recognition result are integrated to obtain the terrain features of each terrain image.

[0013] In a second aspect, the present invention provides a distribution network site selection determination system, including: a segmentation module, a feature recognition module, a demand determination module, a load score calculation module and a site selection module; The segmentation module is used to obtain remote sensing images within the site selection area, segment the remote sensing images based on the semantic segmentation model and the sliding window, and obtain a terrain image set and corresponding location information; The feature recognition module is used to input the terrain image set into the supervised model and the unsupervised model respectively for feature recognition, so as to obtain the terrain features of each terrain image; The demand determination module is used to determine a plurality of power demand points and a load center point in the site selection area based on the power demand data in the site selection area; The load score calculation module is used to determine the radiation circle of each terrain image based on a preset radius length and the midpoint, taking the center point of the terrain image as the midpoint; and calculate the load score of each terrain image based on the radiation circle, the power demand point and the load center point; The site selection module is used to calculate the site selection score of each terrain picture based on the load score and the terrain characteristics, determine the site selection location based on the site selection score, and determine the location information of the site selection location to complete the site selection of the distribution network site.

[0014] Furthermore, the demand determination module is used to: Acquiring power demand data in the site selection area, wherein the power demand data includes power distribution data and load demand data; constructing a power distribution map based on the power distribution data, and determining a plurality of power demand points based on the power distribution map and a preset power distribution threshold; A load forecasting model is constructed based on the regression model, and the load center point is determined based on the load demand data and the load forecasting model.

[0015] Furthermore, the load score calculation module is used to: Count the number of power demand points in each radiation circle, and calculate the load distance of each radiation circle based on the midpoint of each radiation circle and the load center point; The load score of each terrain image is calculated based on the number of power demand points, the load distance and the load score calculation formula. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 A schematic diagram of a flow chart of a method for determining a site selection for a distribution network provided in an embodiment of the present invention; Figure 2 A schematic diagram of the structure of a distribution network site selection and determination system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0016] The specific implementation of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0017] The terms "first" and "second" and the like in the specification, claims and drawings of this application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0018] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0019] Example 1 See also Figure 1 , Figure 1 The present invention provides a method for determining the site selection of a distribution network site, including steps 101 to 105, which are as follows: Step 101: obtaining remote sensing images within the site selection area, segmenting the remote sensing images based on a semantic segmentation model and a sliding window, and obtaining a terrain image set and corresponding location information; In this embodiment, the remote sensing image is segmented based on the semantic segmentation model and the sliding window to obtain a number of terrain images and corresponding location information, including: Constructing an initial semantic segmentation model, and training the initial semantic segmentation model based on remote sensing annotation data, and outputting a trained semantic segmentation model; Initialize a sliding window based on a preset window size and step size, segment the remote sensing image based on the sliding window, and obtain a first segmented image set; The first segmented picture set is input into the semantic segmentation model for feature recognition, terrain information of each first segmented picture is extracted, a terrain picture set is generated, and location information corresponding to each terrain picture is obtained.

[0020] In this embodiment, an initial semantic segmentation model is constructed, and the initial semantic segmentation model may include U-Net, DeepLab, Mask R-CNN, etc.

[0021] In this embodiment, remote sensing annotation data is obtained, and the remote sensing annotation data includes remote sensing images and corresponding terrain contour annotations and terrain type annotations. Remote sensing images are usually taken by satellites or drones and contain high-resolution ground information. The remote sensing images are preprocessed, including denoising, correction, and standardization, to improve the accuracy of subsequent processing.

[0022] In this embodiment, the sliding window is initialized with a preset window size and step size, and the window is moved pixel by pixel to generate a series of overlapping rectangular areas covering the entire image, and each sliding step generates a small sub-area image.

[0023] In this embodiment, the initial semantic segmentation model is trained based on the sub-region images and their annotation information, the hyperparameters of the semantic segmentation model are adjusted during the training process, and the model is optimized based on the cross entropy loss function to output the trained semantic segmentation model.

[0024] In this embodiment, a remote sensing image of the selected area is obtained, and a sliding window is initialized based on a preset window size and step size, and the remote sensing image is segmented based on the sliding window to obtain a first segmented image set; the first segmented image set is input into the semantic segmentation model for feature recognition, and the terrain information contained in the first segmented image is extracted from the output of the semantic segmentation model to generate corresponding terrain images. These images contain terrain types and terrain contours.

[0025] In this embodiment, the global coordinates or pixel coordinates of the original remote sensing image are used to determine the location information of each terrain image, and the location information is associated with each terrain image, and finally a series of terrain images are generated, each terrain image corresponding to a sub-area in the remote sensing image. These terrain images contain terrain contours and location information.

[0026] In this embodiment, based on the semantic segmentation model and the sliding window method, terrain information can be efficiently and accurately extracted from remote sensing images to generate a number of terrain images and corresponding location information. This process not only improves the accuracy and scientificity of site selection analysis, but also greatly enhances the automation of data processing.

[0027] Step 102: inputting the terrain images into the supervised model and the unsupervised model respectively for feature recognition to obtain the terrain features of each terrain image; In this embodiment, the terrain image set is input into the supervised model and the unsupervised model respectively for feature recognition to obtain the terrain features of each terrain image, including: Building a supervised model based on a vector machine, and inputting a terrain picture set into the supervised model to obtain a first recognition result of the supervised model; Building an unsupervised model based on a generative adversarial network, and inputting a terrain picture set into the unsupervised model to obtain a second recognition result of the unsupervised model; The first recognition result and the second recognition result are integrated to obtain the terrain features of each terrain image.

[0028] In this embodiment, a vector machine constructs a supervised model and determines the optimization parameters of the supervised model. Specifically, the optimization parameter of the supervised model is C, where C is the inverse of the regularization strength. C controls the model's tolerance for misclassification. A larger C value means that the model pays more attention to errors in the training data and tries to reduce these errors, which may increase the complexity of the model and make it easy to overfit. A smaller C value means that the model is more inclined to use a simple model, even if it means more training errors.

[0029] In this embodiment, an unsupervised model is constructed based on a generative adversarial network (GANs), which is a deep learning framework. GANs consists of two main parts: a generator and a discriminator. The unsupervised model is trained based on historical terrain images until its recognition accuracy reaches a preset threshold, and then the trained unsupervised model is output.

[0030] In this embodiment, a first recognition result is obtained based on the supervised model; a second recognition result is obtained based on the unsupervised model; wherein the first recognition result and the second recognition result include terrain features and terrain types of each terrain image, and the terrain types include plain terrain, hilly terrain, mountainous terrain and river valley terrain.

[0031] In this embodiment, if the first recognition result and the second recognition result identify different terrain types for a certain terrain image, the position of the terrain image in the remote sensing image is matched, and the terrain image is regenerated based on the position information for recognition.

[0032] In this embodiment, supervised models and unsupervised models can extract features from different perspectives. The supervised model relies on labeled data and can learn clear terrain classification standards; the unsupervised model discovers natural feature structures through similarities between data.

[0033] In this embodiment, the supervised model may be biased due to inaccurate labeling or insufficient training data, while the unsupervised model does not rely on labels and can provide different perspectives and judgments. By comparing and integrating the results of the two, it helps to improve the accuracy and robustness of the overall recognition.

[0034] In this embodiment, if the classification results of a certain terrain image by the supervised model and the unsupervised model are inconsistent, the position information in the remote sensing image can be used for matching and correction to reduce the error rate of terrain image recognition and improve the accuracy of terrain recognition.

[0035] Step 103: determining a plurality of power demand points and a load center point in the site selection area based on the power demand data in the site selection area; In this embodiment, the determining of a plurality of power demand points and a load center point in the site selection area based on the power demand data in the site selection area includes: Acquiring power demand data in the site selection area, wherein the power demand data includes power distribution data and load demand data; constructing a power distribution map based on the power distribution data, and determining a plurality of power demand points based on the power distribution map and a preset power distribution threshold; A load forecasting model is constructed based on the regression model, and the load center point is determined based on the load demand data and the load forecasting model.

[0036] In this embodiment, power distribution data and load demand data are two key data types in power system planning, operation and management. Power distribution data includes the transmission path and network structure of power from power plants to end users. It specifically includes power grid topology, line characteristics, substation information and facility status. Load demand data is the power demand of the power system, that is, the power consumption of power users at different times and locations. It includes load curves, load demands and load characteristics; In this embodiment, a power distribution map is obtained based on the circuit distribution data, and the main structure and key nodes of the power grid, such as substations, main transmission lines, etc., are identified based on the power distribution map. And using GIS technology, the data such as population distribution and building density are superimposed and analyzed with the power distribution map to determine the power distribution value and power demand value of each key node in the site selection area, and set the node with a power distribution value greater than a preset power distribution threshold as a power demand point.

[0037] In this embodiment, data cleaning and data conversion are performed on historical load demand data to generate training data, a load forecasting model is constructed based on a vector machine regression model, and the load forecasting model is trained based on the training data to obtain a trained load forecasting model.

[0038] In this embodiment, the load demand in the site selection area is identified based on the load prediction model, and the area where the load demand is higher than the preset load threshold is taken as the load demand point, and the load demand points are clustered based on the clustering algorithm to identify the load center point.

[0039] In this embodiment, by identifying the power demand point and the load center point, the distribution station can be located closer to the load center, ensuring the reliability of power supply and reducing the loss during power transmission. The location close to the load center point can shorten the length of the transmission line, reduce the power loss caused by transmission, and improve the overall power supply efficiency. At the same time, the location of the distribution station in the area where the load demand is concentrated can reduce the risk of power outages caused by equipment failure or fault response time.

[0040] Step 104: Taking the center point of the terrain image as the midpoint, determining the power circle of each terrain image based on a preset radius length and the midpoint; and calculating the load score of each terrain image based on the power circle, the power demand point and the load center point; In this embodiment, the calculation of the load score of each terrain image based on the radiation circle, the power demand point and the load center point includes: Count the number of power demand points in each radiation circle, and calculate the load distance of each radiation circle based on the midpoint of each radiation circle and the load center point; The load score of each terrain image is calculated based on the number of power demand points, the load distance and the load score calculation formula.

[0041] In this embodiment, the load score calculation formula is specifically: (1) in, represents the load score, is the number of power demand points within the current radiation circle, is the total number of power demand points in the site selection area, is the load distance, and They are the power demand point coefficient and the load distance coefficient respectively.

[0042] In this embodiment, the power demand point coefficient and the load distance coefficient are constants, which are obtained by analyzing the distribution network site selection data.

[0043] In this embodiment, by counting the number of power demand points within the radiation circle, the system can accurately identify areas with concentrated power demand and help determine which areas need priority power supply. At the same time, by calculating the load distance between the midpoint of the radiation circle and the load center point, it helps to optimize the layout of the distribution station, shorten the power supply distance, and reduce transmission losses. Calculating the load score based on the load score formula can evaluate the power demand and power supply cost of different areas. Selecting an area with a high load score can improve power supply efficiency and reduce energy loss.

[0044] In this embodiment, through accurate load score calculation, it is possible to maximize the return on investment and optimize the construction and operation costs while meeting the power supply demand.

[0045] In this embodiment, by counting the number of power demand points and calculating the load distance, combined with the load score calculation formula, the demand and power supply effect of each area can be accurately evaluated. Selecting a site in an area with a high load score will improve power supply efficiency, optimize energy utilization, and ultimately achieve both economic and environmental benefits.

[0046] Step 105: Calculate the site selection score of each terrain image based on the load score and the terrain characteristics, determine the site selection location based on the site selection score, and determine the location information of the site selection location to complete the site selection of the distribution network site.

[0047] In this embodiment, the calculating of the site selection score of each terrain image based on the load score and the terrain feature includes: Acquire terrain features of each terrain image, and match terrain scores based on the terrain features; Based on the load score and the terrain score, a weighted average is performed to calculate the site selection score of each terrain image; specifically: (2) in, represents the site selection score, is the load score, Score for terrain, and are terrain score coefficient and load score coefficient, respectively, which are obtained based on regression analysis, and .

[0048] In this embodiment, terrain types are identified based on terrain features, and terrain types include plain terrain, hilly terrain, mountainous terrain and river valley terrain. A terrain score table is generated based on the terrain types. Please refer to Table 1. Table 1 is a terrain score table of a distribution network site selection determination method provided in an embodiment of the present invention.

[0049] Table 1

[0050] In this embodiment, the plain terrain, hilly terrain, mountainous terrain and river valley terrain in the terrain type are divided into four levels. Since the plain terrain has a flat terrain and a relatively stable geological structure, it is conducive to the construction and maintenance of distribution network sites. Flat terrain is also conducive to the laying of lines and the layout of pipelines, and is convenient for equipment transportation and installation. Therefore, the plain terrain is an ideal site for distribution network sites; while the hilly terrain is undulating, the geological conditions are relatively complex, and there may be more unstable areas and slopes. These problems may lead to unstable foundations, increase the difficulty and cost of construction, but as long as unstable areas and areas prone to geological disasters are avoided, after appropriate foundation treatment, they can still be used as the site for distribution network sites. The mountainous terrain has complex landforms and changeable geological conditions, and is prone to geological disasters such as landslides and mud-rock flows. High-altitude areas may also face problems with harsh climatic conditions, such as low temperatures and strong winds; the river valley area has a low terrain and is easily invaded by floods. The geological conditions may be relatively soft and the foundation stability is poor, so the terrain scores of mountainous terrain and river valley areas are low.

[0051] In this embodiment, the terrain score is used as the topographic score of each terrain image.

[0052] In another optional embodiment, the terrain score also includes an environmental assessment score.

[0053] In another optional embodiment, the environmental assessment influencing factors include climate conditions (such as temperature, rainfall, wind speed, etc.) and ecological environment (such as protected areas). The environmental assessment score of the location of each terrain image is calculated based on the environmental assessment model and the environmental assessment influencing factors, and finally the environmental assessment score and terrain score are integrated to obtain the terrain score.

[0054] In this embodiment, historical distribution network site selection data are collected on the Internet, and the historical distribution network site selection data are annotated based on expert experience to generate sample data, wherein the sample data includes site selection score, terrain score and load score, and the weight coefficients of the terrain score and the load score are estimated based on the least squares method.

[0055] In this embodiment, load scores and terrain characteristics are used to score the site selection, which relies on a large amount of data analysis and modeling, increases the objectivity of the decision-making process, and reduces the deviation caused by subjective judgment. At the same time, by combining the load score (reflecting power demand) and the terrain score (taking into account the natural environment), it is possible to comprehensively evaluate the impact of different terrains on the construction of power facilities, thereby selecting the optimal area; the scoring model can be flexibly adjusted according to the actual conditions of different regions to ensure that the site selection meets the specific environment and needs.

[0056] In this embodiment, the weights of factors affecting the site selection score can be accurately adjusted through regression analysis, so as to respond to changes in a timely manner and optimize the site selection strategy. The impact of load and terrain characteristics is quantified to make site selection decisions more scientific, and the model can reflect the actual situation and improve the accuracy of analysis. Time and labor costs are saved, thereby improving the efficiency of site selection analysis.

[0057] In this embodiment, the terrain images are sorted based on the site selection scores to generate a site selection list for reference by medical site selection personnel. The terrain image with the highest site selection score is used as the site selection location, and the terrain image is matched with the remote sensing image to determine the location information of the terrain image, thereby completing the site selection of the distribution network. The location information includes the longitude and latitude of the site selection location.

[0058] In this embodiment, through remote sensing images and semantic segmentation models, terrain information can be accurately obtained and detailed terrain images can be generated. This allows the site selection analysis to consider not only power demand, but also actual geographical and topographical features, ensuring the scientificity and rationality of the site selection. At the same time, through sliding window and model recognition technology, a large amount of remote sensing data can be processed quickly and automatically, terrain images can be generated and feature recognition can be performed, greatly improving the efficiency of data processing and analysis. In addition, through supervised and unsupervised models, feature recognition of terrain images can effectively extract terrain features, ensure that the complexity of the terrain is fully considered when selecting the site, and improve the accuracy of the site selection.

[0059] In this embodiment, the power demand points and load center points are determined based on the actual power demand data to ensure that the site selection can effectively cover and serve the actual power demand area and avoid the site selection being out of touch with the actual demand. And by analyzing the influence range of each terrain image through the radiation circle, the potential of each terrain image in power transmission and coverage can be more intuitively evaluated.

[0060] In this embodiment, the site selection score is calculated by weighted average, and the load score and terrain characteristics are comprehensively considered to ensure that the site selection not only meets the power demand, but also takes into account the actual feasibility of construction and maintenance.

[0061] In this embodiment, the automation and efficiency of the site selection process are achieved through the combination of remote sensing images, semantic segmentation models, sliding windows, supervised and unsupervised models, etc. At the same time, through the comprehensive analysis of load scores and terrain characteristics, the scientificity and rationality of site selection are ensured, thereby effectively reducing site selection costs and improving site selection efficiency.

[0062] Please refer to Figure 2 , Figure 2A schematic diagram of the structure of a distribution network site selection and determination system provided by an embodiment of the present invention, comprising: a segmentation module 201, a feature recognition module 202, a demand determination module 203, a load score calculation module 204 and a site selection module 205; The segmentation module 201 is used to obtain remote sensing images within the site selection area, segment the remote sensing images based on the semantic segmentation model and the sliding window, and obtain a terrain image set and corresponding location information; The feature recognition module 202 is used to input the terrain image set into the supervised model and the unsupervised model for feature recognition, and obtain the terrain features of each terrain image; The demand determination module 203 is used to determine a number of power demand points and a load center point in the site selection area based on the power demand data in the site selection area; The load score calculation module 204 is used to determine the radiation circle of each terrain image based on a preset radius length and the midpoint, taking the center point of the terrain image as the midpoint; and calculate the load score of each terrain image based on the radiation circle, the power demand point and the load center point; The site selection module 205 is used to calculate the site selection score of each terrain image based on the load score and the terrain characteristics, determine the site selection location based on the site selection score, and determine the location information of the site selection location to complete the site selection of the distribution network site.

[0063] In this embodiment, the demand determination module is used to: Acquiring power demand data in the site selection area, wherein the power demand data includes power distribution data and load demand data; constructing a power distribution map based on the power distribution data, and determining a plurality of power demand points based on the power distribution map and a preset power distribution threshold; A load forecasting model is constructed based on the regression model, and the load center point is determined based on the load demand data and the load forecasting model.

[0064] In this embodiment, the load score calculation module is used to: Count the number of power demand points in each radiation circle, and calculate the load distance of each radiation circle based on the midpoint of each radiation circle and the load center point; The load score of each terrain image is calculated based on the number of power demand points, the load distance and the load score calculation formula. In this embodiment, the load score calculation formula is specifically:

[0065] in, represents the load score, is the number of power demand points within the current radiation circle, is the total number of power demand points in the site selection area, is the load distance, and They are the power demand point coefficient and the load distance coefficient respectively.

[0066] In this embodiment, the site selection module is used to: Acquire terrain features of each terrain image, and match terrain scores based on the terrain features; Based on the load score and the terrain score, a weighted average is performed to calculate the site selection score of each terrain image; specifically:

[0067] in, represents the site selection score, is the load score, Score for terrain, and are terrain score coefficient and load score coefficient, respectively, which are obtained based on regression analysis, and .

[0068] In this embodiment, the segmentation module is used to: Constructing an initial semantic segmentation model, and training the initial semantic segmentation model based on remote sensing annotation data, and outputting a trained semantic segmentation model; Initialize a sliding window based on a preset window size and step size, segment the remote sensing image based on the sliding window, and obtain a first segmented image set; The first segmented picture set is input into the semantic segmentation model for feature recognition, terrain information of each first segmented picture is extracted, a terrain picture set is generated, and location information corresponding to each terrain picture is obtained.

[0069] In this embodiment, the feature recognition module is used to: Building a supervised model based on a vector machine, and inputting a terrain picture set into the supervised model to obtain a first recognition result of the supervised model; Building an unsupervised model based on a generative adversarial network, and inputting a terrain picture set into the unsupervised model to obtain a second recognition result of the unsupervised model; The first recognition result and the second recognition result are integrated to obtain the terrain features of each terrain image.

[0070] In an embodiment of the present invention, a terminal device is also provided, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, the above-mentioned distribution network site selection and determination method is implemented.

[0071] In an embodiment of the present invention, a computer-readable storage medium is further provided. The computer-readable storage medium includes a stored computer program. When the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned distribution network site site selection determination method.

[0072] Exemplarily, the computer program may be divided into one or more modules, one or more modules are stored in a memory and executed by a processor to implement the present invention. One or more modules may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program in a terminal device.

[0073] The terminal device may be a computing device such as a desktop computer, a notebook, a PDA, and a cloud server. The terminal device may include, but is not limited to, a processor, a memory, and a display. Those skilled in the art will appreciate that the above components are merely examples of the terminal device and do not constitute a limitation on the terminal device. The terminal device may include more or fewer components than the components, or may combine certain components, or different components. For example, the terminal device may also include input and output devices, network access devices, buses, etc.

[0074] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, and uses various interfaces and lines to connect various parts of the entire terminal device.

[0075] The memory can be used to store computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function (such as a sound playback function, a text conversion function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, text message data, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0076] Among them, if the module based on the processing of the multi-device access platform is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0077] The specific embodiments described above further illustrate the purpose, technical solutions and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for determining the site selection of a distribution network, characterized in that: include: Acquire remote sensing images within the selected area, segment the remote sensing images based on a semantic segmentation model and a sliding window, and acquire a terrain image set and corresponding location information; Inputting the terrain image set into the supervised model and the unsupervised model respectively for feature recognition to obtain the terrain features of each terrain image; Determine a plurality of power demand points and a load center point within the site selection area based on the power demand data within the site selection area; Taking the center point of the terrain image as the midpoint, determining the radiation circle of each terrain image based on a preset radius length and the midpoint; and calculating the load score of each terrain image based on the radiation circle, the power demand point and the load center point; The site selection score of each terrain image is calculated based on the load score and the terrain characteristics, the site selection location is determined based on the site selection score, and the location information of the site selection location is determined to complete the site selection of the distribution network site.

2. A method for determining the site selection of a distribution network site according to claim 1, characterized in that: The determining of a plurality of power demand points and a load center point in the site selection area based on the power demand data in the site selection area includes: Acquiring power demand data in the site selection area, wherein the power demand data includes power distribution data and load demand data; constructing a power distribution map based on the power distribution data, and determining a plurality of power demand points based on the power distribution map and a preset power distribution threshold; A load forecasting model is constructed based on the regression model, and the load center point is determined based on the load demand data and the load forecasting model.

3. A method for determining the site selection of a distribution network site according to claim 2, characterized in that: The calculating the load score of each terrain picture based on the radiation circle, the power demand point and the load center point includes: Count the number of power demand points in each radiation circle, and calculate the load distance of each radiation circle based on the midpoint of each radiation circle and the load center point; The load score of each terrain image is calculated based on the number of power demand points, the load distance and the load score calculation formula.

4. A method for determining the site selection of a distribution network site according to claim 3, characterized in that: The load score calculation formula is specifically: ； in, represents the load score, is the number of power demand points within the current radiation circle, is the total number of power demand points in the site selection area, is the load distance, and They are the power demand point coefficient and the load distance coefficient respectively.

5. A method for determining the site selection of a distribution network site according to claim 4, characterized in that: The calculating the site selection score of each terrain image based on the load score and the terrain feature includes: Acquire terrain features of each terrain image, and match terrain scores based on the terrain features; Based on the load score and the terrain score, a weighted average is performed to calculate the site selection score of each terrain image; specifically: ； in, represents the site selection score, is the load score, Score for terrain, and are terrain score coefficient and load score coefficient, respectively, which are obtained based on regression analysis, and .

6. A method for determining the site selection of a distribution network site according to claim 5, characterized in that: The remote sensing image is segmented based on the semantic segmentation model and the sliding window to obtain a plurality of terrain images and corresponding location information, including: Constructing an initial semantic segmentation model, and training the initial semantic segmentation model based on remote sensing annotation data, and outputting a trained semantic segmentation model; Initialize a sliding window based on a preset window size and step size, segment the remote sensing image based on the sliding window, and obtain a first segmented image set; The first segmented picture set is input into the semantic segmentation model for feature recognition, terrain information of each first segmented picture is extracted, a terrain picture set is generated, and location information corresponding to each terrain picture is obtained.

7. A method for determining the site selection of a distribution network site according to claim 6, characterized in that: The step of inputting the terrain image set into the supervised model and the unsupervised model for feature recognition to obtain the terrain features of each terrain image includes: Building a supervised model based on a vector machine, and inputting a terrain picture set into the supervised model to obtain a first recognition result of the supervised model; Building an unsupervised model based on a generative adversarial network, and inputting a terrain picture set into the unsupervised model to obtain a second recognition result of the unsupervised model; The first recognition result and the second recognition result are integrated to obtain the terrain features of each terrain image.

8. A distribution network site selection and determination system, characterized in that: include: Segmentation module, feature recognition module, demand determination module, load score calculation module and site selection module; The segmentation module is used to obtain remote sensing images within the site selection area, segment the remote sensing images based on the semantic segmentation model and the sliding window, and obtain a terrain image set and corresponding location information; The feature recognition module is used to input the terrain image set into the supervised model and the unsupervised model respectively for feature recognition, so as to obtain the terrain features of each terrain image; The demand determination module is used to determine a plurality of power demand points and a load center point in the site selection area based on the power demand data in the site selection area; The load score calculation module is used to determine the radiation circle of each terrain image based on a preset radius length and the midpoint, taking the center point of the terrain image as the midpoint; and calculate the load score of each terrain image based on the radiation circle, the power demand point and the load center point; The site selection module is used to calculate the site selection score of each terrain picture based on the load score and the terrain characteristics, determine the site selection location based on the site selection score, and determine the location information of the site selection location to complete the site selection of the distribution network site.

9. A distribution network site selection and determination system as claimed in claim 8, characterized in that: The demand determination module is used to: Acquiring power demand data in the site selection area, wherein the power demand data includes power distribution data and load demand data; constructing a power distribution map based on the power distribution data, and determining a plurality of power demand points based on the power distribution map and a preset power distribution threshold; A load forecasting model is constructed based on the regression model, and the load center point is determined based on the load demand data and the load forecasting model.

10. A distribution network site selection and determination system according to claim 9, characterized in that: The load score calculation module is used to: Count the number of power demand points in each radiation circle, and calculate the load distance of each radiation circle based on the midpoint of each radiation circle and the load center point; The load score of each terrain image is calculated based on the number of power demand points, the load distance and the load score calculation formula.

Images