Distribution network business expansion intelligent power supply point recommendation method and system based on GIS map

Abstract

The invention relates to the technical field of power grids, and discloses a distribution network business expansion intelligent power supply point recommendation method and system based on a GIS map, and the method comprises the steps: obtaining distribution network equipment parameter data, real-time operation data and geographic space data, and generating a distribution network node line topological graph with an attribute tag; constructing a multi-dimensional power supply point evaluation feature set; based on the multi-dimensional power supply point evaluation feature set, generating a group of Pareto optimal solution sets as a candidate power supply point set by adopting an NSGA-III multi-objective optimization algorithm; inputting the candidate power supply point set and the distribution network node line topological graph into a LightGBM-GAT model, predicting load changes of candidate power supply points in the future 72 hours, calculating a topological matching score of each candidate power supply point, and selecting the power supply point with the highest comprehensive score as an optimal power supply point; in combination with a Dijkstra algorithm, planning an optimal power supply path from an optimal power supply point to a client position on a GIS map, and visually displaying the optimal power supply path; according to the invention, the intellectualization and accuracy of power supply point recommendation are improved.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power grids, in particular to a GIS map-based intelligent power supply point recommendation method and system for distribution network expansion. BACKGROUND

[0002] Distribution network expansion is a core link for power grid enterprises to provide customers with power access services, and the rationality of power supply point recommendation directly affects the economy of power supply schemes and the stability of distribution network operation. At present, power supply point recommendation for distribution network expansion mainly relies on manual experience, and only considers single factors, such as the geographical distance between the power supply point and the customer and the real-time capacity, and ignores key dimensions such as grid topology connectivity, long-term operation and maintenance cost and power supply reliability, which may lead to one-sided schemes. The capacity calculation is static and does not combine with load change trend prediction, so the recommended power supply point may be quickly overloaded due to subsequent load growth. The optimization mode is mainly single-target decision, which is difficult to balance the multi-target demand of distance, capacity utilization rate and construction cost. GIS maps are only used for simple location display and are not deeply integrated with intelligent algorithms, so they cannot efficiently support path planning and result visualization. SUMMARY

[0003] The application aims to solve the above problems, and provides a GIS map-based intelligent power supply point recommendation method and system for distribution network expansion.

[0004] The first aspect of the application provides a GIS map-based intelligent power supply point recommendation method for distribution network expansion, which comprises the following steps: Obtain distribution network equipment parameter data, real-time operation data and geographic space data, and generate a distribution network node line topology graph with attribute labels; Construct a multi-dimensional power supply point evaluation feature set including grid topology features, load features, economic features and reliability features; Based on the multi-dimensional power supply point evaluation feature set, adopt an NSGA-III multi-objective optimization algorithm to minimize the distance, maximize the capacity utilization rate and minimize the construction cost, generate a set of Pareto optimal solutions as a candidate power supply point set; Input the candidate power supply point set and the distribution network node line topology graph into a LightGBM-GAT model, predict the load change of the candidate power supply point in the next 72 hours, calculate the topology matching score of each candidate power supply point, and select the one with the highest comprehensive score as the optimal power supply point; Combine the Dijkstra algorithm to plan the optimal power supply path from the optimal power supply point to the customer location on the GIS map, and visually display it.

[0005] Optionally, in a first implementation of the first aspect of the present invention, the step of acquiring distribution network equipment parameter data, real-time operating data, and geospatial data, and generating a distribution network node line topology map with attribute tags, includes: Obtain power distribution equipment parameter data, real-time operation data, and geospatial data of the area where customers consume electricity; By using the buffer analysis function of GIS maps, geospatial data is filtered to mark power points located in restricted areas or that cannot be accessed due to road obstruction, thus obtaining objects to be excluded. By using the spatial overlay analysis function of GIS map, the parameter data of distribution network equipment, real-time operation data and processed geospatial data are associated to construct a distribution network node line topology map with attribute labels. The attribute labels include the ID, type, rated capacity, real-time load, interval status, geographic coordinates of each power point node, and the impedance and current carrying capacity of each line.

[0006] Optionally, in a second implementation of the first aspect of the present invention, the construction of a multi-dimensional power source evaluation feature set including power grid topology characteristics, load characteristics, economic characteristics, and reliability characteristics includes: Based on the generated distribution network node line topology map, the graph theory analysis method is used to calculate the number of shortest paths from each power source point to the customer's power consumption location. The topology connectivity is obtained by calculating the ratio of the number of connected nodes in the subnet where the power source point is located to the total number of nodes. The real-time available capacity of the power point is calculated based on real-time operating data. The current load rate is obtained by the ratio of real-time load to rated capacity. The average daily maximum load of the power point is calculated by retrieving the historical load data of the past 3 months. By combining the line length in geospatial data with the power grid engineering quota standards, the cost of the cable line from the power source to the customer's power consumption location is calculated. The average annual operation and maintenance cost over the entire life cycle is calculated based on the design service life of the line and equipment, and the total operation and maintenance cost is obtained by summing them up. The incoming line redundancy is obtained by acquiring the total number of incoming lines and the number of spare incoming lines at the power source point. The distribution network fault records of the past year are retrieved, and the ratio of the number of faults of the lines associated with the power source point to the total length of the lines is calculated to obtain the line fault rate.

[0007] Optionally, in a third implementation of the first aspect of the present invention, the multi-dimensional power point evaluation feature set employs the NSGA-III multi-objective optimization algorithm to generate a set of Pareto optimal solutions as a candidate power point set, with the objectives of minimizing distance, maximizing capacity utilization, and minimizing construction cost, including: All power points within the distribution network that are not excluded by geographical constraints are used as the initial solution. A certain number of power points are randomly selected to form the initial population, with each individual representing a candidate power point. The objective function is determined with the goals of minimizing distance, maximizing capacity utilization, and minimizing construction cost. For each individual in the population, the values ​​of the three objective functions are compared, and individuals not dominated by other individuals are grouped into the same non-dominated layer, and the layers are divided in sequence according to the dominance relationship. For each individual in a non-dominated layer, calculate its distance to neighboring individuals in each target dimension, and sum them to obtain the crowding degree; The tournament selection method is adopted, which combines the ranking of the non-dominated layer and the crowding degree to select the better individuals from the current population to form the parent population. The offspring population is generated by simulating binary crossover and polynomial mutation. Repeat the iteration until the number of iterations reaches the preset value. The non-dominated individuals in the population form the Pareto optimal solution set, which is the candidate power source set.

[0008] Optionally, in the fourth implementation of the first aspect of the present invention, the step of inputting the candidate power source set and the distribution network node line topology map into the LightGBM-GAT model, predicting the load changes of the candidate power sources over the next 72 hours, calculating the topology matching score of each candidate power source, and selecting the power source with the highest comprehensive score as the optimal power source includes: The candidate power source set and the distribution network node line topology map are input into the LightGBM-GAT model. The LightGBM layer constructs a weak classifier through gradient boosting decision tree iteration, filters influencing factors through feature importance evaluation, predicts the load change curve of the candidate power source in the next 72 hours, and extracts the predicted maximum load value. Input the distribution network node line topology map into the GAT layer to calculate the topology matching score. Combine the load prediction results of the LightGBM layer with the topology matching score of the GAT layer, calculate the comprehensive score of each candidate power source point according to the preset weight, and select the one with the highest score as the optimal power source point.

[0009] Optionally, in a fifth implementation of the first aspect of the present invention, the step of inputting the distribution network node line topology map into the GAT layer to calculate the topology matching score includes: With candidate power supply points and customer power consumption locations as core nodes and other nodes as auxiliary nodes, the attention coefficients between the core nodes and neighboring nodes are calculated through a self-attention mechanism, and the features of neighboring nodes are aggregated based on the weighted coefficients. Based on the aggregation results, the line current carrying capacity margin compliance rate, voltage drop satisfaction, and topology redundancy are calculated, and the topology matching score is obtained by weighted summation.

[0010] Optionally, in a sixth implementation of the first aspect of the present invention, the step of combining Dijkstra's algorithm to plan the optimal power supply path from the optimal power source to the customer location on a GIS map and visually displaying it includes: Starting from the customer's electricity location and ending at the optimal power source point, the distance from the starting point to all other nodes is initialized to infinity, and the distance from the starting point is set to 0. Iteratively update the shortest distance of each node until the endpoint is reached, then backtrack the path to obtain the shortest power supply path from the customer's location to the optimal power point; In the GIS map, the optimal power source, candidate power source, and customer power location are marked with different colors. The power supply path is drawn according to the planned shortest power supply path, and the line length and current carrying capacity are marked.

[0011] A second aspect of the present invention provides an intelligent power source recommendation system for power distribution network expansion based on GIS maps, the system comprising: The acquisition module is used to acquire power distribution network equipment parameter data, real-time operation data and geospatial data, and generate a power distribution network node line topology map with attribute tags; The module is used to construct a multi-dimensional power source evaluation feature set, including power grid topology characteristics, load characteristics, economic characteristics, and reliability characteristics. The generation module is used to generate a set of Pareto optimal solutions based on a multi-dimensional power point evaluation feature set and employing the NSGA-III multi-objective optimization algorithm, with the objectives of minimizing distance, maximizing capacity utilization, and minimizing construction cost, as a set of candidate power points; The calculation module is used to input the candidate power source set and the distribution network node line topology map into the LightGBM-GAT model, predict the load change of the candidate power source in the next 72 hours, calculate the topology matching score of each candidate power source, and select the one with the highest comprehensive score as the optimal power source. The visualization module is used to combine the Dijkstra algorithm to plan the optimal power supply path from the optimal power source to the customer location on a GIS map and visualize it.

[0012] A third aspect of the present invention provides a GIS map-based intelligent power source recommendation device for power distribution network expansion, the GIS map-based intelligent power source recommendation device comprising a memory and at least one processor, the memory storing instructions; the at least one processor invokes the instructions in the memory to cause the GIS map-based intelligent power source recommendation device to perform the steps of the GIS map-based intelligent power source recommendation method for power distribution network expansion as described in any of the preceding claims.

[0013] A fourth aspect of the present invention provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the steps of the intelligent power point recommendation method for power distribution network expansion based on a GIS map as described in any of the preceding claims.

[0014] The technical solution provided by this invention acquires distribution network equipment parameter data, real-time operation data, and geospatial data, and generates a distribution network node line topology map with attribute tags; constructs a multi-dimensional power source evaluation feature set including power grid topology features, load features, economic features, and reliability features; based on the multi-dimensional power source evaluation feature set, the NSGA-III multi-objective optimization algorithm is used to generate a set of Pareto optimal solutions as candidate power source sets, with the objectives of minimizing distance, maximizing capacity utilization, and minimizing construction costs; the candidate power source set and the distribution network node line topology map are input into the LightGBM-GAT model to predict the load changes of the candidate power sources in the next 72 hours, and calculates the topology matching score of each candidate power source, selecting the one with the highest comprehensive score as the optimal power source; and combines Dijkstra's algorithm with the network topology map. The Dijkstra algorithm plans the optimal power supply path from the optimal power source to the customer location on a GIS map and displays the results visually. This invention constructs a multi-dimensional evaluation feature set to comprehensively cover topology, load, economy, and reliability, avoiding one-sided solutions. It adopts the NSGA-III multi-objective optimization algorithm to achieve a balanced optimization of distance, capacity utilization, and construction cost, generating a Pareto optimal candidate set. The LightGBM-GAT model can accurately predict load changes and calculate topology matching scores, ensuring the dynamic adaptability of power sources and the rationality of topology. Combining the Dijkstra algorithm with GIS visualization, it efficiently plans the optimal power supply path and intuitively displays the results. Overall, it significantly improves the intelligence and accuracy of power source recommendation, reduces manual intervention, shortens the business expansion cycle, and ensures the safe and stable operation of the distribution network. Attached Figure Description

[0015] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention.

[0016] Figure 1 A flowchart of the intelligent power supply point recommendation method for power distribution network expansion based on GIS map provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the intelligent power supply point recommendation system for power distribution network expansion based on GIS map provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of the intelligent power point recommendation device for power distribution network expansion based on GIS map provided in an embodiment of the present invention. Detailed Implementation

[0017] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” or “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, apparatus, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0018] For ease of understanding, the specific process of the embodiments of the present invention is described below. Please refer to [link / reference]. Figure 1 The flowchart of the intelligent power supply point recommendation method for power distribution network expansion based on GIS map provided in this embodiment of the invention includes the following steps: Step 101: Obtain power distribution equipment parameter data, real-time operation data and geospatial data, and generate a power distribution node line topology map with attribute tags; In this embodiment, data acquisition is completed through multi-source data interfaces and database linkage: Distribution network equipment parameter data is retrieved from the power grid resource business platform and equipment lifecycle management system, covering equipment ID, type, rated capacity, and interval status of power sources such as substations, switching stations, and ring main units, as well as static parameters such as impedance, current carrying capacity, laying method, and material specifications of distribution network lines; real-time operating data is obtained from the distribution automation (DA) system and SCADA real-time monitoring system, including real-time load values, power factors, and current load rates of each power source, real-time current, voltage, and power loss of lines, and dynamic data such as the real-time opening and closing status of switching equipment; geospatial data is extracted from the geographic information system (GIS) platform and urban surveying database, including the latitude and longitude coordinates of the customer's electricity location, the road distribution and direction of the area, building boundary range, ecological protection zones and prohibited construction zones, and the geospatial coordinates and spatial relationships of all power sources and lines; after collection, the three types of data are standardized in format and validated for validity, outliers and duplicate data are removed to ensure data field uniformity and accurate logical association.

[0019] In this embodiment, distribution network equipment parameter data, real-time operation data, and geospatial data of the customer's electricity location area are acquired. Using the buffer analysis function of the GIS map, the geospatial data is filtered to identify power supply points located in restricted areas or blocked by roads, thus identifying exclusion targets. Through the spatial overlay analysis function of the GIS map, the distribution network equipment parameter data, real-time operation data, and processed geospatial data are associated to construct a distribution network node line topology map with attribute labels. These attribute labels include the ID, type, rated capacity, real-time load, interval status, geographical coordinates, impedance, and current carrying capacity of each line for each power supply point node.

[0020] Step 102: Construct a multi-dimensional power source evaluation feature set that includes power grid topology characteristics, load characteristics, economic characteristics, and reliability characteristics; In this embodiment, based on the generated distribution network node line topology map, graph theory analysis is used to calculate the number of shortest paths from each power source point to the customer's power consumption location. The topology connectivity is obtained by calculating the ratio of the number of connected nodes in the subnet where the power source point is located to the total number of nodes. The real-time available capacity of the power source point is calculated based on real-time operating data. The current load rate is obtained by the ratio of real-time load to rated capacity. The average daily maximum load of the power source point over the past three months is retrieved as the historical maximum load. The cable line cost from the power source point to the customer's power consumption location is calculated by combining the line length in the geospatial data with the power grid engineering quota standards. The annual average operation and maintenance cost over the entire life cycle is calculated based on the design service life of the lines and equipment, and the total operation and maintenance cost is obtained by summing them up. The incoming line redundancy is obtained by obtaining the total number of incoming lines and the number of spare incoming lines of the power source point. The line failure rate is obtained by retrieving the distribution network fault records of the past year and calculating the ratio of the number of faults of the lines associated with the power source point to the total length of the lines.

[0021] Step 103: Based on the multi-dimensional power point evaluation feature set, the NSGA-III multi-objective optimization algorithm is used to generate a set of Pareto optimal solutions as candidate power point sets, with the objectives of minimizing distance, maximizing capacity utilization, and minimizing construction cost. In this embodiment, all power points within the distribution network that are not excluded by geographical constraints are used as the initial solution. A certain number of power points are randomly selected to form an initial population, with each individual representing a candidate power point. The objective function is determined with the goals of minimizing distance, maximizing capacity utilization, and minimizing construction cost. For each individual in the population, the values ​​of the three objective functions are compared, and individuals not dominated by other individuals are grouped into the same non-dominated layer. The population is then divided into multiple layers according to the dominance relationship. For each individual in the non-dominated layer, the distance between it and its neighboring individuals in each objective dimension is calculated, and the crowding degree is obtained by summing the distances. A tournament selection method is used, combining the non-dominated layer ranking and the crowding degree, to select the better individuals from the current population to form the parent population. The offspring population is generated by simulating binary crossover and polynomial mutation. The iteration is repeated until the number of iterations reaches a preset value. The non-dominated layer individuals in the population form the Pareto optimal solution set, which is the candidate power point set.

[0022] Step 104: Input the candidate power source set and the distribution network node line topology map into the LightGBM-GAT model, predict the load change of the candidate power source in the next 72 hours, calculate the topology matching score of each candidate power source, and select the one with the highest comprehensive score as the optimal power source. In this embodiment, the candidate power source set and the distribution network node line topology map are input into the LightGBM-GAT model. The LightGBM layer iteratively constructs a weak classifier through gradient boosting decision tree, filters influencing factors through feature importance evaluation, predicts the load change curve of the candidate power source for the next 72 hours, and extracts the predicted maximum load value. The distribution network node line topology map is input into the GAT layer to calculate the topology matching score. The load prediction result of the LightGBM layer is combined with the topology matching score of the GAT layer, and the comprehensive score of each candidate power source is calculated according to the preset weight. The one with the highest score is selected as the optimal power source.

[0023] In this embodiment, candidate power supply points and customer power consumption locations are taken as core nodes, and other nodes are taken as auxiliary nodes. The attention coefficients between the core nodes and adjacent nodes are calculated through a self-attention mechanism. The features of neighboring nodes are aggregated based on the weighted coefficients. The line current carrying capacity compliance rate, voltage drop satisfaction, and topology redundancy are calculated based on the aggregation results, and the topology matching score is obtained by weighted summation.

[0024] Step 105: Using Dijkstra's algorithm, plan the optimal power supply path from the optimal power source to the customer's location on the GIS map and visualize it.

[0025] In this embodiment, the customer's power consumption location is taken as the starting point and the optimal power supply point is taken as the ending point. The distance from the starting point to all other nodes is initialized to infinity, and the distance of the starting point is set to 0. The shortest distance of each node is iteratively updated until the ending point is reached. The path is backtracked to obtain the shortest power supply path from the customer's location to the optimal power supply point. In the GIS map, the optimal power supply point, candidate power supply points, and customer's power consumption location are marked with different colors. The power supply path is drawn according to the planned shortest power supply path, and the line length and current carrying capacity are marked.

[0026] Please see Figure 2 A schematic diagram of the structure of the intelligent power supply point recommendation system for power distribution network expansion based on GIS map provided in this embodiment of the invention. The system includes: The acquisition module is used to acquire power distribution network equipment parameter data, real-time operation data and geospatial data, and generate a power distribution network node line topology map with attribute tags; The module is used to construct a multi-dimensional power source evaluation feature set, including power grid topology characteristics, load characteristics, economic characteristics, and reliability characteristics. The generation module is used to generate a set of Pareto optimal solutions based on a multi-dimensional power point evaluation feature set and employing the NSGA-III multi-objective optimization algorithm, with the objectives of minimizing distance, maximizing capacity utilization, and minimizing construction cost, as a set of candidate power points; The calculation module is used to input the candidate power source set and the distribution network node line topology map into the LightGBM-GAT model, predict the load change of the candidate power source in the next 72 hours, calculate the topology matching score of each candidate power source, and select the one with the highest comprehensive score as the optimal power source. The visualization module is used to combine the Dijkstra algorithm to plan the optimal power supply path from the optimal power source to the customer location on a GIS map and visualize it.

[0027] Figure 3This is a schematic diagram of the structure of a GIS map-based intelligent power point recommendation device for power distribution network expansion, provided in an embodiment of the present invention. This GIS map-based intelligent power point recommendation device 300 can vary significantly due to different configurations or performance. It may include one or more central processing units (CPUs) 310 (e.g., one or more processors) and a memory 320, and one or more storage media 330 (e.g., one or more mass storage devices) for storing application programs 333 or data 332. The memory 320 and storage media 330 can be temporary or persistent storage. The program stored in the storage media 330 may include one or more modules (not shown in the diagram), each module including a series of instruction operations on the GIS map-based intelligent power point recommendation device 300. Furthermore, the processor 310 may be configured to communicate with the storage media 330 and execute the series of instruction operations in the storage media 330 on the GIS map-based intelligent power point recommendation device 300 to implement the method provided in the above embodiment.

[0028] The intelligent power point recommendation device 300 for power distribution network expansion based on GIS maps may also include one or more power supplies 340, one or more wired or wireless network interfaces 350, one or more input / output interfaces 360, and / or one or more operating systems 331, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will understand that... Figure 3 The intelligent power point recommendation equipment structure for power distribution network expansion based on GIS map shown does not constitute a limitation on the computer equipment provided by this invention. It may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0029] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform each step of the intelligent power point recommendation method for power distribution network expansion based on GIS maps provided in the above embodiments.

[0030] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described equipment or apparatus / unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0031] If the integrated unit is implemented as 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 technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0032] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for intelligent power source recommendation in power distribution network expansion based on GIS maps, characterized in that, The method includes the following steps: Acquire power distribution network equipment parameter data, real-time operation data, and geospatial data, and generate a power distribution network node line topology map with attribute tags; Construct a multi-dimensional power source evaluation feature set that includes power grid topology characteristics, load characteristics, economic characteristics, and reliability characteristics; Based on the multi-dimensional power point evaluation feature set, the NSGA-III multi-objective optimization algorithm is adopted to generate a set of Pareto optimal solutions as a candidate power point set with the objectives of minimizing distance, maximizing capacity utilization, and minimizing construction cost. Input the candidate power source set and the distribution network node line topology map into the LightGBM-GAT model to predict the load change of the candidate power source in the next 72 hours, calculate the topology matching score of each candidate power source, and select the one with the highest comprehensive score as the optimal power source. By combining Dijkstra's algorithm, the optimal power supply path from the optimal power source to the customer's location is planned on a GIS map and visualized.

2. The intelligent power source recommendation method for power distribution network expansion based on GIS maps as described in claim 1, characterized in that, The process of acquiring distribution network equipment parameter data, real-time operating data, and geospatial data, and generating a distribution network node line topology map with attribute tags, includes: Obtain power distribution equipment parameter data, real-time operation data, and geospatial data of the area where customers consume electricity; By using the buffer analysis function of GIS maps, geospatial data is filtered to mark power points located in restricted areas or that cannot be accessed due to road obstruction, thus obtaining objects to be excluded. By using the spatial overlay analysis function of GIS map, the parameter data of distribution network equipment, real-time operation data and processed geospatial data are associated to construct a distribution network node line topology map with attribute labels. The attribute labels include the ID, type, rated capacity, real-time load, interval status, geographic coordinates of each power point node, and the impedance and current carrying capacity of each line.

3. The intelligent power source recommendation method for power distribution network expansion based on GIS maps as described in claim 1, characterized in that, The construction of the multi-dimensional power source evaluation feature set includes power grid topology characteristics, load characteristics, economic characteristics, and reliability characteristics, including: Based on the generated distribution network node line topology map, the graph theory analysis method is used to calculate the number of shortest paths from each power source point to the customer's power consumption location. The topology connectivity is obtained by calculating the ratio of the number of connected nodes in the subnet where the power source point is located to the total number of nodes. The real-time available capacity of the power point is calculated based on real-time operating data. The current load rate is obtained by the ratio of real-time load to rated capacity. The average daily maximum load of the power point is calculated by retrieving the historical load data of the past 3 months. By combining the line length in geospatial data with the power grid engineering quota standards, the cost of the cable line from the power source to the customer's power consumption location is calculated. The average annual operation and maintenance cost over the entire life cycle is calculated based on the design service life of the line and equipment, and the total operation and maintenance cost is obtained by summing them up. The incoming line redundancy is obtained by acquiring the total number of incoming lines and the number of spare incoming lines at the power source point. The distribution network fault records of the past year are retrieved, and the ratio of the number of faults of the lines associated with the power source point to the total length of the lines is calculated to obtain the line fault rate.

4. The intelligent power source recommendation method for power distribution network expansion based on GIS maps as described in claim 1, characterized in that, The multi-dimensional power point evaluation feature set employs the NSGA-III multi-objective optimization algorithm, aiming to minimize distance, maximize capacity utilization, and minimize construction cost, to generate a set of Pareto optimal solutions as a candidate power point set, including: All power points within the distribution network that are not excluded by geographical constraints are used as the initial solution. A certain number of power points are randomly selected to form the initial population, with each individual representing a candidate power point. The objective function is determined with the goals of minimizing distance, maximizing capacity utilization, and minimizing construction cost. For each individual in the population, the values ​​of the three objective functions are compared, and individuals not dominated by other individuals are grouped into the same non-dominated layer, and the layers are divided in sequence according to the dominance relationship. For each individual in a non-dominated layer, calculate its distance to neighboring individuals in each target dimension, and sum them to obtain the crowding degree; The tournament selection method is adopted, which combines the ranking of the non-dominated layer and the crowding degree to select the better individuals from the current population to form the parent population. The offspring population is generated by simulating binary crossover and polynomial mutation. Repeat the iteration until the number of iterations reaches the preset value. The non-dominated individuals in the population form the Pareto optimal solution set, which is the candidate power source set.

5. The intelligent power source recommendation method for power distribution network expansion based on GIS maps as described in claim 1, characterized in that, The process involves inputting the candidate power source set and the distribution network node line topology map into the LightGBM-GAT model to predict the load changes of the candidate power sources over the next 72 hours, calculating the topology matching score for each candidate power source, and selecting the power source with the highest comprehensive score as the optimal power source. This includes: The candidate power source set and the distribution network node line topology map are input into the LightGBM-GAT model. The LightGBM layer constructs a weak classifier through gradient boosting decision tree iteration, filters influencing factors through feature importance evaluation, predicts the load change curve of the candidate power source in the next 72 hours, and extracts the predicted maximum load value. Input the distribution network node line topology map into the GAT layer to calculate the topology matching score. Combine the load prediction results of the LightGBM layer with the topology matching score of the GAT layer, calculate the comprehensive score of each candidate power source point according to the preset weight, and select the one with the highest score as the optimal power source point.

6. The intelligent power source recommendation method for power distribution network expansion based on GIS maps as described in claim 5, characterized in that, The step of inputting the distribution network node line topology map into the GAT layer to calculate the topology matching score includes: With candidate power supply points and customer power consumption locations as core nodes and other nodes as auxiliary nodes, the attention coefficients between the core nodes and neighboring nodes are calculated through a self-attention mechanism, and the features of neighboring nodes are aggregated based on the weighted coefficients. Based on the aggregation results, the line current carrying capacity margin compliance rate, voltage drop satisfaction, and topology redundancy are calculated, and the topology matching score is obtained by weighted summation.

7. The intelligent power source recommendation method for power distribution network expansion based on GIS maps as described in claim 1, characterized in that, The method of combining Dijkstra's algorithm to plan the optimal power supply path from the optimal power source to the customer location on a GIS map and visually displaying it includes: Starting from the customer's electricity location and ending at the optimal power source point, the distance from the starting point to all other nodes is initialized to infinity, and the distance from the starting point is set to 0. Iteratively update the shortest distance of each node until the endpoint is reached, then backtrack the path to obtain the shortest power supply path from the customer's location to the optimal power point; In the GIS map, the optimal power source, candidate power source, and customer power location are marked with different colors. The power supply path is drawn according to the planned shortest power supply path, and the line length and current carrying capacity are marked.

8. A power supply point recommendation system for power distribution network expansion based on GIS maps, characterized in that, The system includes: The acquisition module is used to acquire power distribution network equipment parameter data, real-time operation data and geospatial data, and generate a power distribution network node line topology map with attribute tags; The module is used to construct a multi-dimensional power source evaluation feature set, including power grid topology characteristics, load characteristics, economic characteristics, and reliability characteristics. The generation module is used to generate a set of Pareto optimal solutions based on a multi-dimensional power point evaluation feature set and employing the NSGA-III multi-objective optimization algorithm, with the objectives of minimizing distance, maximizing capacity utilization, and minimizing construction cost, as a set of candidate power points; The calculation module is used to input the candidate power source set and the distribution network node line topology map into the LightGBM-GAT model, predict the load change of the candidate power source in the next 72 hours, calculate the topology matching score of each candidate power source, and select the one with the highest comprehensive score as the optimal power source. The visualization module is used to combine the Dijkstra algorithm to plan the optimal power supply path from the optimal power source to the customer's location on a GIS map and visualize it.

9. A power supply point recommendation device for power distribution network expansion based on GIS maps, characterized in that, The intelligent power source recommendation device for power distribution network expansion based on GIS map includes a memory and at least one processor. The memory stores instructions. The at least one processor calls the instructions in the memory to cause the intelligent power source recommendation device for power distribution network expansion based on GIS map to perform each step of the intelligent power source recommendation method for power distribution network expansion based on GIS map as described in any one of claims 1-7.

10. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement each step of the intelligent power point recommendation method for power distribution network expansion based on GIS maps as described in any one of claims 1-7.

Images