A method and system for selecting access nodes of charging stations in power distribution networks based on multi-indicator integration

Abstract

This invention relates to the field of charging station access node selection technology, and particularly to a method and system for selecting charging station access nodes in a distribution network using multi-index fusion. The method involves: obtaining the node loss coefficient of each candidate node; obtaining the predicted peak power curve of the target charging station within a preset time window using a preset peak power predictor; obtaining the node peak power curve of the candidate node within the preset time window, and obtaining the node redundancy power curve based on the node's design parameters and the node peak power curve; correcting the node redundancy power curve based on the transmission loss coefficient to obtain the node's available power curve; obtaining the node power adaptation coefficient based on the time adaptation coefficient and spatial adaptation coefficient; calculating the node adaptation coefficient of multiple candidate nodes based on the node loss coefficient and the node power adaptation coefficient; and selecting the access node. The nodes selected by this invention can ensure the stable operation of the charging station and effectively control transmission costs and losses.

Description

Technical Field

[0001] This invention relates to the field of charging station access node screening technology, and in particular to a method and system for screening charging station access nodes in a power distribution network by integrating multiple indicators. Background Technology

[0002] In the construction of charging stations in power distribution networks, the selection of access nodes is crucial to operational efficiency, network stability, and economics. Traditional screening methods often consider only a single dimension, such as distance or single power data. While some methods incorporate power factors, they fail to fully account for key elements affecting power demand, such as traffic flow and time-of-day information. Furthermore, when assessing node power supply capacity, they do not take into account transmission losses and the limitations imposed by the node's design capacity on actual power supply. This leads to selected nodes that are prone to defects in cost, losses, or power adaptability, resulting in large deviations in power prediction, inaccurate assessments of node power supply capacity, and decisions that are highly subjective due to a lack of quantitative basis, ultimately impacting network stability and charging station operation. Summary of the Invention

[0003] This invention addresses the technical problems in existing technologies, such as nodes being selected that are prone to defects in cost, loss, or power adaptability, large power prediction deviations, inaccurate assessment of node power supply capacity, and highly subjective decision-making due to a lack of quantitative basis, which affect the stability of the distribution network and the operation of charging stations. It provides a method and system for selecting distribution network charging station access nodes by integrating multiple indicators to solve these problems.

[0004] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0005] In a first aspect, the present invention provides a method for screening access nodes of a power distribution network charging station by integrating multiple indicators, comprising: calculating the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station; calculating the node loss coefficient of each candidate node based on the line cost coefficient and transmission loss coefficient; obtaining the predicted peak power curve within a preset time window based on the traffic flow and current time information of the area where the target charging station is located, using a preset peak power predictor; obtaining the node peak power curve of the candidate node within the preset time window based on the historical load data of the candidate nodes, and obtaining the node redundancy power curve based on the node's design parameters and the node peak power curve; correcting the node redundancy power curve based on the transmission loss coefficient to obtain the node available power curve; comparing the node available power curve with the predicted peak power curve to obtain the time adaptation coefficient and spatial adaptation coefficient, weighting and fusing them as the node power adaptation coefficient; calculating the node adaptation coefficient of multiple candidate nodes based on the node loss coefficient and node power adaptation coefficient of multiple candidate nodes, and selecting the node with the largest node adaptation coefficient as the access node.

[0006] Optionally, the process involves calculating the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station, and calculating the node loss coefficient based on the line cost coefficient and transmission loss coefficient. This includes: establishing a transmission line cost model based on a geographic information system; obtaining multiple optimal paths for constructing transmission lines from the target charging station to the multiple candidate nodes based on the transmission line cost model and the locations of the target charging station and the multiple candidate nodes; calculating the cost of constructing multiple lines for each of the multiple optimal paths, and obtaining multiple line cost coefficients based on the multiple line costs; calculating the transmission loss coefficient for each candidate node based on the length of the multiple optimal paths; and calculating the node loss coefficient for each candidate node based on the line cost coefficient and transmission loss coefficient.

[0007] Optionally, based on traffic flow and current time information in the area where the target charging station is located, a predicted peak power curve within a preset time window is obtained using a preset peak power predictor. This includes: acquiring real-time traffic flow data and current time information in the area where the target charging station is located, wherein the time information includes at least time, date type, and weather; acquiring historical traffic flow, historical time information, and corresponding historical peak power of similar charging stations within a historical time period as historical sample information; training a preset peak power predictor based on the historical sample information; and inputting the real-time traffic flow data and current time information into the peak power predictor to obtain a predicted peak power curve within the preset time window, wherein the predicted peak power curve is composed of predicted peak power at multiple time points.

[0008] The process of acquiring historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations within a historical period as historical sample information includes: using a vehicle counting device to acquire the electric vehicle traffic flow per unit time within a preset range of the target charging station as potential passenger flow; acquiring multiple existing charging stations, acquiring their potential passenger flow for each, and comparing it with the potential passenger flow of the target charging station to obtain multiple passenger flow deviation coefficients; setting a potential passenger flow deviation threshold, acquiring existing charging stations that meet the passenger flow deviation threshold as similar charging stations; and acquiring historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations within a historical period as historical sample information.

[0009] Optionally, the peak power curve of the candidate node within a preset time window is obtained, and a node redundancy power curve is obtained based on the node's design parameters and the node peak power curve. The node redundancy power curve is then corrected based on the transmission loss coefficient to obtain the node available power curve. This includes: obtaining historical load data and historical time period information of the candidate node; analyzing the load patterns of the candidate node based on machine learning; predicting multiple predicted peak powers of the candidate node within the preset time window based on the current time period information; plotting the node peak power curve based on the multiple predicted peak powers; obtaining the maximum design capacity of the candidate node based on the node's design parameters; converting the node peak power curve based on the maximum design capacity to obtain the node redundancy power curve; and correcting the node redundancy power curve based on the transmission loss coefficient to obtain the node available power curve.

[0010] Optionally, the available power curve of a node is compared with the predicted peak power curve to obtain a time adaptation coefficient and a spatial adaptation coefficient. These are then weighted and fused to obtain the node power adaptation coefficient. This includes: grouping the predicted peak power curve and the available power curve of the node into the same two-dimensional coordinate system, where the horizontal axis represents the preset time window; calculating the ratio of the time the available power curve of the node is higher than the predicted peak power curve to the preset time window, as the time adaptation coefficient; calculating the excess power supply area and the total demand area based on the curve characteristics of the available power curve and the predicted peak power curve in the two-dimensional coordinate system, and using the ratio of the excess power supply area to the total demand area as the spatial adaptation coefficient; and weightedly fusing the time adaptation coefficient and the spatial adaptation coefficient to obtain the node power adaptation coefficient.

[0011] Specifically, based on the curve characteristics of the node's available power curve and the predicted peak power curve in a two-dimensional coordinate system, the excess power supply area and the total demand area are calculated respectively. The ratio of the excess power supply area to the total demand area is used as the spatial adaptation coefficient. This includes: using a numerical integration method to calculate the area enclosed by the portion of the node's available power curve that is higher than the predicted peak power curve within a preset time window, which is taken as the excess power supply area; calculating the total area enclosed by the predicted peak power curve and the time axis within the preset time window, which is taken as the total demand area; and calculating the ratio of the excess power supply area to the total demand area to obtain the spatial adaptation coefficient.

[0012] Secondly, the present invention provides a multi-index fusion-based distribution network charging station access node screening system, comprising:

[0013] The node loss coefficient acquisition module is used to calculate the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station, and to calculate the node loss coefficient of each candidate node based on the line cost coefficient and transmission loss coefficient.

[0014] The peak power curve prediction module is used to obtain the predicted peak power curve within a preset time window based on the traffic flow and current time information of the area where the target charging station is located, through a preset peak power predictor.

[0015] The available power curve acquisition module is used to acquire the node peak power curve of the candidate node within a preset time window based on the historical load data of the candidate node, and to obtain the node redundancy power curve based on the node design parameters and the node peak power curve. Based on the transmission loss coefficient, the node redundancy power curve is corrected to obtain the node available power curve.

[0016] The node selection module compares the available power curve of a node with the predicted peak power curve to obtain the time adaptation coefficient and the spatial adaptation coefficient, and then uses the weighted fusion as the node power adaptation coefficient. Based on the node loss coefficient and node power adaptation coefficient of multiple candidate nodes, the module calculates the node adaptation coefficient of multiple candidate nodes and selects the node with the largest node adaptation coefficient as the access node.

[0017] Thirdly, this application provides an electronic device, comprising:

[0018] Memory, used to store the first computer program;

[0019] The processor is used to read and execute the first computer program, thereby realizing the multi-index fusion method for screening access nodes of power distribution network charging stations in the first aspect.

[0020] Fourthly, this application provides a storage medium storing a second computer program, which, when executed by a processor, implements the multi-index fusion method for screening access nodes of power distribution network charging stations in the first aspect.

[0021] By implementing this invention, it is possible to calculate the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station, and to calculate the node loss coefficient of each candidate node based on the line cost coefficient and transmission loss coefficient. The node loss coefficient can quantify the performance of the candidate node in terms of transmission cost and loss, providing an objective and accurate reference for cost and loss dimensions for subsequent selection of access nodes, and avoiding decision-making bias caused by considering only a single factor.

[0022] By implementing this invention, it is possible to obtain a predicted peak power curve within a preset time window based on traffic flow and current time information in the area where the target charging station is located, using a preset peak power predictor. The predicted peak power curve can provide advance knowledge of the power demand of the target charging station at different time points, providing accurate demand-side data support for subsequent judgment on whether the candidate nodes can meet the power supply, and ensuring the stability of power supply during the operation of the charging station.

[0023] By implementing this invention, it is possible to obtain the peak power curve of the candidate node within a preset time window based on the historical load data of the candidate node, and obtain the node redundancy power curve based on the node's design parameters and the node peak power curve. Based on the transmission loss coefficient, the node redundancy power curve is corrected to obtain the node available power curve. The node available power curve comprehensively considers the node's own design capacity and transmission loss, and truly reflects the effective power that the candidate node can provide to the charging station in actual operation. This provides reliable supply-side data for subsequent power adaptability judgment and avoids resource waste or insufficient power supply due to overestimation or underestimation of the node's power supply capacity.

[0024] By implementing this invention, it is possible to compare the available power curve of a node with the predicted peak power curve to obtain the time adaptation coefficient and the spatial adaptation coefficient, and then use the weighted fusion as the node power adaptation coefficient. Based on the node loss coefficient and node power adaptation coefficient of multiple candidate nodes, the node adaptation coefficient of multiple candidate nodes is calculated, and the node with the largest node adaptation coefficient is selected as the access node. The node adaptation coefficient comprehensively considers the loss and power adaptation of the candidate nodes. Using this coefficient as a screening criterion, the access node with the best comprehensive performance in terms of cost loss and power supply adaptability can be selected, ensuring the scientific and rational selection of access nodes and improving the operating efficiency of the distribution network and the operating benefits of charging stations.

[0025] In summary, by implementing this invention, the selected access nodes can not only meet the power requirements of the target charging station within a preset time window, ensuring stable operation of the charging station, but also effectively control transmission costs and losses, improving the overall operating efficiency and economy of the distribution network. This avoids the problem of unreasonable access node selection caused by traditional screening methods due to single-factor considerations and inaccurate data, providing strong technical support for the scientific decision-making of charging station access nodes in the distribution network. By implementing this invention, it is possible to calculate the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station, and based on these coefficients, calculate the node loss coefficient for each candidate node. The node loss coefficient quantifies the performance of the candidate node in terms of transmission costs and losses, providing an objective and accurate reference for subsequent access node selection based on cost and loss dimensions, avoiding decision-making biases caused by considering only a single factor.

[0026] By implementing this invention, it is possible to obtain a predicted peak power curve within a preset time window based on traffic flow and current time information in the area where the target charging station is located, using a preset peak power predictor. The predicted peak power curve can provide advance knowledge of the power demand of the target charging station at different time points, providing accurate demand-side data support for subsequent judgment on whether the candidate nodes can meet the power supply, and ensuring the stability of power supply during the operation of the charging station.

[0027] By implementing this invention, it is possible to obtain the peak power curve of the candidate node within a preset time window based on the historical load data of the candidate node, and obtain the node redundancy power curve based on the node's design parameters and the node peak power curve. Based on the transmission loss coefficient, the node redundancy power curve is corrected to obtain the node available power curve. The node available power curve comprehensively considers the node's own design capacity and transmission loss, and truly reflects the effective power that the candidate node can provide to the charging station in actual operation. This provides reliable supply-side data for subsequent power adaptability judgment and avoids resource waste or insufficient power supply due to overestimation or underestimation of the node's power supply capacity.

[0028] By implementing this invention, it is possible to compare the available power curve of a node with the predicted peak power curve to obtain the time adaptation coefficient and the spatial adaptation coefficient, and then use the weighted fusion as the node power adaptation coefficient. Based on the node loss coefficient and node power adaptation coefficient of multiple candidate nodes, the node adaptation coefficient of multiple candidate nodes is calculated, and the node with the largest node adaptation coefficient is selected as the access node. The node adaptation coefficient comprehensively considers the loss and power adaptation of the candidate nodes. Using this coefficient as a screening criterion, the access node with the best comprehensive performance in terms of cost loss and power supply adaptability can be selected, ensuring the scientific and rational selection of access nodes and improving the operating efficiency of the distribution network and the operating benefits of charging stations.

[0029] In summary, by implementing this invention, the selected access nodes can not only meet the power requirements of the target charging station within a preset time window and ensure the stable operation of the charging station, but also effectively control transmission costs and losses, improve the overall operating efficiency and economy of the distribution network, and avoid the problem of unreasonable access node selection caused by the single consideration of factors and inaccurate data in traditional screening methods. This provides strong technical support for the scientific decision-making of access nodes for charging stations in the distribution network. Attached Figure Description

[0030] Figure 1 A flowchart illustrating the multi-index fusion method for selecting access nodes of power distribution network charging stations provided by this invention;

[0031] Figure 2 A schematic diagram of the structure of the multi-index fusion power distribution network charging station access node screening system provided by the present invention;

[0032] Figure 3 A schematic diagram of the structure of an electronic device provided by the present invention;

[0033] Figure 4 This is a schematic diagram of a storage medium provided by the present invention.

[0034] In the attached diagram, the components represented by each number are as follows:

[0035] The module includes a node loss coefficient acquisition module 11, a peak power curve prediction module 12, an available power curve acquisition module 13, an adapter node selection module 14, an electronic device 300, a memory 310, a processor 320, a first computer program 311, a storage medium 400, and a second computer program 410. Detailed Implementation

[0036] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0037] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0038] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0039] Example 1, as Figure 1 As shown in the figure, this embodiment of the invention provides a method for screening access nodes of power distribution network charging stations by integrating multiple indicators, including:

[0040] S100: Calculate the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station, and calculate the node loss coefficient of each candidate node based on the line cost coefficient and transmission loss coefficient.

[0041] S200: Based on the traffic flow and current time information of the target charging station area, the system obtains the predicted peak power curve within a preset time window through a preset peak power predictor.

[0042] S300: Based on the historical load data of the candidate nodes, obtain the node peak power curve of the candidate nodes within a preset time window, and obtain the node redundancy power curve based on the node design parameters and the node peak power curve. Based on the transmission loss coefficient, correct the node redundancy power curve to obtain the node available power curve.

[0043] S400: Compare the available power curve of the node with the predicted peak power curve to obtain the time adaptation coefficient and spatial adaptation coefficient, and use the weighted fusion as the node power adaptation coefficient; Based on the node loss coefficient and node power adaptation coefficient of multiple candidate nodes, calculate the node adaptation coefficient of multiple candidate nodes, and select the node with the largest node adaptation coefficient as the access node.

[0044] In step S100 of this application embodiment, the calculation of the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station, and the calculation of the node loss coefficient based on the line cost coefficient and transmission loss coefficient, includes:

[0045] Establish a transmission line cost model based on geographic information systems;

[0046] Based on the transmission line cost model and the locations of the target charging station and multiple candidate nodes, multiple optimal paths for constructing transmission lines from the target charging station to the multiple candidate nodes are obtained;

[0047] Calculate the cost of constructing multiple optimal routes for each route, and obtain multiple route cost coefficients based on the costs of these routes.

[0048] Based on the lengths of multiple optimal paths, the transmission loss coefficient of each candidate node is calculated.

[0049] The node loss coefficient for each candidate node is calculated based on the line cost coefficient and transmission loss coefficient.

[0050] In this embodiment of the application, the core purpose of step S100 is to quantitatively evaluate the access feasibility between each candidate node and the target charging station from two key dimensions: economy and energy efficiency, and finally generate a comprehensive index of "node loss coefficient" to provide basic data support for the subsequent selection of the optimal access node.

[0051] To achieve the above objectives, it is first necessary to establish a transmission line cost model based on a Geographic Information System (GIS). Specifically, this requires collecting basic cost parameters for transmission line construction, such as the unit price of conductor materials, construction cost per unit length, crop compensation cost per unit area, and terrain adjustment coefficient. Then, geographic data of the target area is entered into the GIS system. This geographic data includes the latitude and longitude of the candidate nodes and the target charging station, the terrain type of the intermediate area, and the land use type. The terrain type includes plains, mountains, and water areas, and the land use type includes arable land and construction land. Next, a cost calculation formula is established, for example: Line Cost = Conductor Length × Conductor Unit Price + Construction Cost × Terrain Adjustment Coefficient + Land Compensation Cost × Land Area, forming a transmission line cost model that can automatically call upon geographic data to calculate costs.

[0052] Next, based on the transmission line cost model and the locations of the target charging station and multiple candidate nodes, it is necessary to obtain multiple optimal paths for constructing transmission lines from the target charging station to the multiple candidate nodes.

[0053] Specifically, the location coordinates of the target charging station and all candidate nodes need to be imported into the GIS system, and path constraints need to be set, such as prohibited areas and minimum turning radius. Then, the GIS's "shortest path" or "lowest cost path" algorithm is called to generate one or more transmission line paths that meet the constraints for each candidate node. From the generated transmission line paths, infeasible paths are eliminated, and paths with "shortest length, less terrain obstruction, and lower land compensation cost" are retained as the "best paths".

[0054] Then, the costs of constructing multiple lines along multiple optimal paths need to be calculated separately, and multiple line cost coefficients are obtained based on these costs. Geographical parameters such as the length, terrain type, and land type of each optimal path are input into the transmission line cost model, which automatically calculates the line cost for each optimal path. For example, the optimal path cost for candidate node A is 1 million yuan, and the optimal path cost for candidate node B is 800,000 yuan.

[0055] Then, the line costs of all candidate nodes are normalized to generate line cost coefficients. For example, the line cost coefficient of a candidate node = the line cost of that candidate node / the maximum line cost of all candidate nodes. If the maximum line cost is 1 million yuan, then the line cost coefficient of candidate node A is 1.0, and the line cost coefficient of candidate node B is 0.8. The smaller the line cost coefficient, the lower the line construction cost of that candidate node.

[0056] Furthermore, based on the lengths of multiple optimal paths, the transmission loss coefficient for each candidate node is calculated. Specifically, the actual conductor length of each optimal path needs to be extracted from the GIS system; then, based on the physical formula for transmission loss, such as theoretical power loss = I²Rt, where R = conductor resistance × length, combined with the estimated transmission current and estimated transmission power of the candidate nodes, the theoretical power loss value and theoretical transmission power of each path are calculated. The theoretical transmission power can be estimated as the rated power or average demand power of the target charging station.

[0057] Furthermore, combining the theoretical power loss value and the theoretical transmission power, the transmission loss rate for each optimal path is calculated, where the transmission loss rate = theoretical power loss value / theoretical transmission power × 100%. This transmission loss rate is used as the transmission loss coefficient for the candidate nodes. Finally, the transmission loss coefficient for each candidate node is obtained using the above method.

[0058] In step S200 of this embodiment, based on the traffic flow and current time information of the area where the target charging station is located, a predicted peak power curve within a preset time window is obtained through a preset peak power predictor, including:

[0059] Obtain real-time traffic flow data for the area where the target charging station is located, as well as the current time period information, which includes at least time, date type, and weather.

[0060] Historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations within a historical period are obtained as historical sample information. Based on the historical sample information, a preset peak power predictor is trained.

[0061] The real-time traffic flow data and current time period information are input into the peak power predictor to obtain the predicted peak power curve within a preset time window. The predicted peak power curve is composed of the predicted peak power at multiple time points.

[0062] In this embodiment of the application, the purpose of step S200 is to predict the peak power change trend of the target charging station within a future preset time window, so as to provide key basis for subsequent evaluation of the power supply capacity adaptability and grid load scheduling after the candidate node is connected.

[0063] First, it is necessary to obtain real-time traffic flow data of the area where the target charging station is located, as well as the current time period information, which includes at least time, date type, and weather.

[0064] Specifically, real-time traffic flow data can be obtained from traffic monitoring equipment around the target charging station, navigation platform APIs, or traffic management department databases. This includes data such as the number of vehicles passing by the roads around the charging station every 15 minutes and traffic density. At the same time, the current specific time, date type (weekday / weekend / holiday), and weather conditions (sunny / rainy / snowy / high temperature, etc.) can be collected. This information will affect users' travel and charging habits. For example, traffic flow increases during holidays, while inclement weather may reduce outdoor activities.

[0065] Next, it is necessary to obtain the historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations within a historical period, as historical sample information, and then use the historical sample information as the basis for the analysis.

[0066] In step S200 of this application embodiment, historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations within a historical time period are obtained as historical sample information, including:

[0067] Using vehicle counting equipment, the electric vehicle traffic per unit time within a preset range of the target charging station is obtained as potential passenger traffic;

[0068] Multiple existing charging stations were identified, their potential passenger flow was obtained for each, and compared with the potential passenger flow of the target charging station to obtain multiple passenger flow deviation coefficients.

[0069] A potential passenger flow deviation threshold is preset, and existing charging stations that meet the passenger flow deviation threshold are identified as similar charging stations.

[0070] Historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations are obtained as historical sample information.

[0071] In this embodiment of the application, the core purpose of obtaining historical sample information is to accurately screen out “similar charging stations” with similar operating characteristics to the target charging station, and obtain their historical data as training samples to provide reliable basic data for the subsequent construction of a peak power predictor.

[0072] To achieve the above objectives, it is first necessary to use vehicle counting equipment to obtain the electric vehicle traffic flow per unit time within a preset range of the target charging station, as potential passenger flow. This involves deploying vehicle counting equipment, such as smart cameras, infrared sensors, and loop detectors, within a preset range around the target charging station, such as major roads and parking lot entrances within a 1-kilometer radius. The number of electric vehicles passing through this range per unit time is collected by the vehicle counting equipment and defined as the "potential passenger flow" of the target charging station. For example, collecting the number of electric vehicles passing through this range every hour or every 30 minutes reflects the scale of vehicles that may be heading to the charging station.

[0073] Next, it is necessary to acquire data from multiple existing charging stations, obtain their potential passenger flow for each, and compare it with the potential passenger flow of the target charging station to obtain multiple passenger flow deviation coefficients. Specifically, multiple existing charging stations can be selected, and the same vehicle counting equipment and counting method can be used to collect the "potential passenger flow" of each existing charging station. Then, the "passenger flow deviation coefficient" between each existing charging station and the target charging station can be calculated. The calculation method can be set as: Passenger flow deviation coefficient = |Potential passenger flow of existing charging station - Potential passenger flow of target charging station| / Potential passenger flow of target charging station. The smaller the passenger flow deviation coefficient, the closer the potential passenger flow of the two stations is.

[0074] Next, a potential passenger flow deviation threshold needs to be preset, and existing charging stations that meet the threshold are identified as similar charging stations. Specifically, a "potential passenger flow deviation threshold" is preset, such as 10% or 20%, which can be adjusted according to the actual scenario. For example, setting the potential passenger flow deviation threshold to 20% means that stations with a "deviation coefficient ≤ 20%" are considered similar charging stations.

[0075] From the passenger flow deviation coefficient results calculated in the aforementioned steps, existing charging stations that meet the potential passenger flow deviation threshold are selected and defined as "similar charging stations". The passenger flow scale of these stations is close to that of the target station, and it is inferred that their charging demand patterns are also similar.

[0076] Finally, it is necessary to obtain historical traffic flow, historical time period information, and corresponding historical peak power for similar charging stations over historical periods, as historical sample information. Specifically, this involves extracting historical traffic flow, historical time period information, and corresponding historical peak power under these conditions from the operational database of similar charging stations. The historical peak power represents the maximum charging power of similar charging stations within each historical time period.

[0077] Furthermore, a preset peak power predictor needs to be trained based on the historical sample information.

[0078] Considering the task type of the peak power predictor, a long short-term memory (LSTM) network can be used to build the peak power predictor.

[0079] The peak power predictor is structured as follows: input layer, hidden layer, fully connected layer, and output layer. The input layer receives traffic flow data and time period information. The hidden layer consists of two LSTM layers. The first LSTM layer has 32 neurons with the tanh activation function and returns "yes," transmitting complete sequence information to the second layer. The second LSTM layer has 16 neurons with the tanh activation function and returns "no," outputting only the features from the last time step. The fully connected layer has one layer with 8 neurons and the ReLU activation function. The output layer has the number of neurons corresponding to the preset time window. For example, to predict the peak power for the next 24 hours, it outputs 24 neurons with the linear activation function, outputting continuous peak power values.

[0080] For the peak power predictor parameters, the batch size is set to 32. The learning rate is set to 0.001. The time step can be set to 24, meaning each sample contains historical data from the past 24 time units, used to predict power in future periods. The regularization coefficient is set to 0.0001, using L2 regularization to prevent overfitting. The dropout rate is set to 0.2.

[0081] For training the peak power predictor, the training data consists of the historical sample information. The total sample size is no less than 10,000, divided into training and validation sets in an 8:2 ratio. The initial training epochs are set to 100. When the mean squared error loss value of the validation set no longer decreases for 10 consecutive epochs, and the difference between the loss value of the training set and the loss value of the validation set is less than 0.01, the model is considered to have converged, and the peak power predictor is obtained.

[0082] Finally, the real-time traffic flow data and current time period information are input into the peak power predictor to obtain the predicted peak power curve within a preset time window. The predicted peak power curve consists of the predicted peak power at multiple time points. The peak power predictor outputs the predicted peak power at multiple time points within a preset time window, such as within the next 24 or 48 hours, based on a model mapping relationship. These power values ​​are arranged in chronological order to form the "predicted peak power curve," for example: predicted peak of 500kW at 16:00, predicted peak of 800kW at 17:00, and predicted peak of 600kW at 18:00, etc., intuitively reflecting the time and intensity of future peak electricity consumption.

[0083] In step S300 of this embodiment, the peak power curve of the candidate node within a preset time window is obtained, and based on the node's design parameters and the peak power curve, the node redundancy power curve is obtained. Then, based on the transmission loss coefficient, the node redundancy power curve is corrected to obtain the node's usable power curve, including:

[0084] Historical load data and historical time period information of candidate nodes are obtained, load patterns of candidate nodes are analyzed based on machine learning, and multiple predicted peak powers of candidate nodes within a preset time window are predicted based on the current time period information. Based on the multiple predicted peak powers, the peak power curve of the node is plotted.

[0085] Based on the node design parameters, the maximum design capacity of the candidate nodes is obtained, and the peak power curve of the nodes is converted based on the maximum design capacity to obtain the node redundancy power curve.

[0086] Based on the aforementioned transmission loss coefficient, the node redundancy power curve is corrected to obtain the node available power curve.

[0087] In this embodiment of the application, the core purpose of step S300 is to quantitatively evaluate the actual effective power that the candidate node can provide to the target charging station within a future preset time window. By calculating the available power curve of the node, it is determined whether the node can meet the predicted peak power demand of the charging station, thus providing a power adaptability basis for the final selection of access nodes.

[0088] To achieve the above objectives, it is first necessary to obtain historical load data and historical time period information of the candidate nodes, analyze the load patterns of the candidate nodes based on machine learning, and predict multiple predicted peak powers of the candidate nodes within a preset time window based on the current time period information. Based on the multiple predicted peak powers, the peak power curve of the node is plotted.

[0089] Specifically, it is necessary to first obtain the "historical load data" of the candidate nodes, such as the actual power consumption during different time periods over a past period; and "historical time period information," that is, the historical time, date type, weather, etc., corresponding to the historical load data. Then, machine learning algorithms, such as time series models, are used to train the model based on the historical load data and historical time period information to discover the patterns of node load changes with time, date type, and weather. For example, candidate nodes in industrial areas may have high loads on weekdays and low loads on weekends. The training method for this time series model is similar to that of the aforementioned peak power predictor, and will not be elaborated here.

[0090] The current time period information is input into the trained time series model to predict the "predicted peak power" of the candidate node at multiple time points within a preset time window, that is, the maximum load at each time point. These power values ​​are arranged in chronological order to draw the "node peak power curve", which reflects the future peak electricity consumption trend of the candidate node itself.

[0091] Then, based on the node's design parameters, the maximum design capacity of the candidate nodes needs to be obtained. Based on the maximum design capacity, the peak power curve of the node is transformed to obtain the node redundancy power curve.

[0092] Specifically, the "maximum design capacity" needs to be extracted from the design parameters of the candidate nodes, which is the maximum total power that the candidate node's own power supply system can withstand, i.e., the node's power ceiling. Then, for each time point in the node's peak power curve, the predicted peak power at that time point is subtracted from the maximum design capacity to obtain the "node redundancy power" for that time point. The formula is: Node redundancy power at a certain time point = Maximum design capacity - Node predicted peak power at that time point. Finally, the node redundancy power at all time points is arranged in chronological order to form a "node redundancy power curve," reflecting the maximum power margin that the node can provide to the outside world in future time periods.

[0093] Finally, based on the transmission loss coefficient, the node redundancy power curve needs to be corrected to obtain the node available power curve. Specifically, the "transmission loss coefficient" of the candidate node calculated in step S100 needs to be retrieved. Then, for each time point in the node redundancy power curve, the redundant power at that time point is multiplied by (1 - transmission loss coefficient) to obtain the corrected "node available power". The formula is: node available power at a certain time point = node redundant power at that time point × (1 - transmission loss coefficient). Then, the node available power at all time points is arranged in chronological order to form the "node available power curve".

[0094] By following the steps above, the actual power supply capacity of each candidate node in future time periods can be clearly identified, providing a direct basis for comparing the available power curve of the node with the predicted peak power curve of the target charging station and determining whether the candidate node meets the power supply requirements.

[0095] In step S400 of this embodiment, the available power curve of the node is compared with the predicted peak power curve to obtain the time adaptation coefficient and the spatial adaptation coefficient, which are then weighted and fused to serve as the node power adaptation coefficient. This includes:

[0096] The predicted peak power curve and the node available power curve are grouped into the same two-dimensional coordinate system, where the horizontal axis is the preset time window.

[0097] The ratio of the length of time during which the available power curve of a computing node is higher than the predicted peak power curve to the preset time window is used as the time adaptation coefficient.

[0098] Based on the curve characteristics of the node available power curve and the predicted peak power curve in the two-dimensional coordinate system, the excess power supply area and the total demand area are calculated respectively, and the ratio of the excess power supply area to the total demand area is used as the spatial adaptation coefficient.

[0099] The node power adaptation coefficient is obtained by weighted fusion of the time adaptation coefficient and the spatial adaptation coefficient.

[0100] In this embodiment, the core objective of step S400 is to quantitatively evaluate the degree of compatibility between the available power of the candidate node and the predicted power demand of the target charging station from two dimensions: "time matching degree" (measured by the time adaptation coefficient) and "power supply sufficiency" (measured by the spatial adaptation coefficient). Finally, a comprehensive indicator, the "node power adaptation coefficient," is generated through weighted fusion. This indicator directly reflects whether the candidate node can cover peak charging times in terms of time and meet charging demands in terms of power, providing a crucial basis for subsequently calculating the total node adaptation coefficient and selecting the optimal access node.

[0101] To achieve the above objectives, it is first necessary to integrate the predicted peak power curve and the available power curve at the nodes into the same two-dimensional coordinate system. The aim is to unify the time and power dimensions of the two curves, establishing a common calculation benchmark for subsequent comparative analysis.

[0102] Specifically, the horizontal axis of the two-dimensional coordinate system is set to a "preset time window", such as the next 24 hours, and the scale is divided according to the unit time (hourly), for a total of 24 time points.

[0103] The vertical axis represents "power," with units such as kW. The scale range covers the maximum power values ​​of both curves to ensure that the curves are displayed completely.

[0104] The "predicted peak power curve" (reflecting charging station demand) obtained in step S200 and the "node available power curve" (reflecting node supply) obtained in step S300 are plotted simultaneously in this coordinate system, so that each time point of the two curves corresponds one-to-one with the power value, which facilitates subsequent calculations.

[0105] Next, it is necessary to calculate the ratio of the time length during which the available power curve of a node is higher than the predicted peak power curve to the preset time window, as the time adaptation coefficient. Specifically, it is necessary to traverse each time point within the preset time window in the coordinate system, such as the 24 time points corresponding to 24 hours; determine whether the power value of the available power curve of a node is higher than the power value of the predicted peak power curve at each time point. If it is higher, it is considered that the time point is "adapted" and meets the requirements; count the total time length corresponding to all "adapted" time points. For example, if there are 20 adaptations out of 24 time points, and each time point corresponds to 1 hour, then the total adaptation time length is 20 hours; then calculate the time adaptation coefficient according to the formula: Time adaptation coefficient = Adaptation time length / Total length of preset time window. For example, 20 hours / 24 hours ≈ 0.83. The closer the time adaptation coefficient is to 1, the higher the time matching degree.

[0106] Then, based on the curve characteristics of the node available power curve and the predicted peak power curve in the two-dimensional coordinate system, it is necessary to calculate the excess power supply area and the total demand area respectively, and use the ratio of the excess power supply area to the total demand area as the spatial adaptation coefficient.

[0107] In step S400 of this embodiment, based on the curve characteristics of the node available power curve and the predicted peak power curve in a two-dimensional coordinate system, the excess power supply area and the total demand area are calculated respectively, and the ratio of the excess power supply area to the total demand area is used as the spatial adaptation coefficient, including:

[0108] The numerical integration method is used to calculate the area enclosed by the portion of the node's available power curve that is higher than the predicted peak power curve within a preset time window, which is taken as the excess power supply area.

[0109] Calculate the total area enclosed by the predicted peak power curve and the time axis within a preset time window, and use it as the total demand area;

[0110] Calculate the ratio of the excess power supply area to the total required area to obtain the space adaptation coefficient.

[0111] First, a numerical integration method is needed to calculate the area enclosed by the portion of the node's available power curve that is higher than the predicted peak power curve within a preset time window. This area is defined as the excess power supply area. Specifically, the positional relationship between the node's available power curve and the predicted peak power curve needs to be determined using the same two-dimensional coordinate system, with the preset time window as the horizontal axis and power as the vertical axis. Then, each time interval within the preset time window is traversed. A time interval is a small segment divided by unit time. For each time interval, only the portion of the node's available power curve that is higher than the predicted peak power curve is selected for calculation. The total area of ​​this "excess portion" is calculated using a numerical integration method, which is the "excess power supply area," expressed in power × time, such as kW·h.

[0112] Next, it is necessary to calculate the total area enclosed by the predicted peak power curve and the time axis within the preset time window, which is taken as the total demand area. Similarly, using the numerical integration method, the total area enclosed by the predicted peak power curve and the horizontal axis (time axis) within the preset time window is calculated. This area represents the total power demand of the charging station in the future period, which is called the "total demand area". The unit is the same as the excess power supply area, such as kW·h.

[0113] Finally, the ratio of the excess power supply area to the total required area is calculated to obtain the spatial adaptation coefficient. That is, spatial adaptation coefficient = excess power supply area ÷ total required area. A higher spatial adaptation coefficient indicates that the available power at the node exceeds the demand, and the stronger the guarantee capability for the charging station's power demand.

[0114] Furthermore, the time adaptation coefficient and the spatial adaptation coefficient need to be weighted and fused to obtain the node power adaptation coefficient.

[0115] Specifically, the weights of the time adaptation coefficient and the spatial adaptation coefficient can be set, with a sum of 1. These weights can be adjusted according to actual needs; for example, if time coverage is more important, the time weight will be higher. The default weights for the time adaptation coefficient and spatial adaptation coefficient are 0.6 and 0.4, respectively. Then, the node power adaptation coefficient is calculated using the formula: Node power adaptation coefficient = (Time adaptation coefficient × Time weight) + (Spatial adaptation coefficient × Spatial weight). Assuming the time adaptation coefficient is 0.83 and the spatial adaptation coefficient is 0.5, then the node power adaptation coefficient = 0.83 × 0.6 + 0.5 × 0.4 = 0.698. The closer the final node power adaptation coefficient is to 1, the better the compatibility between the power supply of the candidate node and the power demand of the charging station.

[0116] Finally, based on the node loss coefficient and node power adaptation coefficient of multiple candidate nodes, the node adaptation coefficient of multiple candidate nodes is calculated, and the node with the largest node adaptation coefficient is selected as the access node, thus completing the multi-index fusion screening of the distribution network charging station access node. Specifically, the node adaptation coefficient can be calculated by weighting and fusing the node loss coefficient and the node power adaptation coefficient. For example, node adaptation coefficient = w1 × (1 - node loss coefficient) + w2 × node power adaptation coefficient, where w1 + w2 = 1, and the weights can be adjusted according to actual needs.

[0117] Example 2, as Figure 2 As shown, based on the same inventive concept as the multi-index fusion distribution network charging station access node screening method provided in Embodiment 1, this embodiment of the invention also provides a multi-index fusion distribution network charging station access node screening system, including:

[0118] The node loss coefficient acquisition module 11 is used to calculate the line cost coefficient and transmission loss coefficient between multiple candidate nodes and the target charging station, and to calculate the node loss coefficient of each candidate node based on the line cost coefficient and transmission loss coefficient.

[0119] The peak power curve prediction module 12 is used to obtain the predicted peak power curve within a preset time window based on the traffic flow and current time information of the area where the target charging station is located, through a preset peak power predictor.

[0120] The available power curve acquisition module 13 is used to acquire the node peak power curve of the candidate node within a preset time window based on the historical load data of the candidate node, and to obtain the node redundancy power curve based on the node design parameters and the node peak power curve. Based on the transmission loss coefficient, the node redundancy power curve is corrected to obtain the node available power curve.

[0121] The node selection module 14 is used to compare the available power curve of a node with the predicted peak power curve to obtain the time adaptation coefficient and the spatial adaptation coefficient, and then use the weighted fusion as the node power adaptation coefficient; based on the node loss coefficient and the node power adaptation coefficient of multiple candidate nodes, the node adaptation coefficient of multiple candidate nodes is calculated, and the node with the largest node adaptation coefficient is selected as the access node.

[0122] Furthermore, the node loss coefficient acquisition module 11 includes the following execution steps:

[0123] Establish a transmission line cost model based on geographic information systems;

[0124] Based on the transmission line cost model and the locations of the target charging station and multiple candidate nodes, multiple optimal paths for constructing transmission lines from the target charging station to the multiple candidate nodes are obtained;

[0125] Calculate the cost of constructing multiple optimal routes for each route, and obtain multiple route cost coefficients based on the costs of these routes.

[0126] Based on the lengths of multiple optimal paths, the transmission loss coefficient of each candidate node is calculated.

[0127] The node loss coefficient for each candidate node is calculated based on the line cost coefficient and transmission loss coefficient.

[0128] Furthermore, the peak power curve prediction module 12 includes the following execution steps:

[0129] Obtain real-time traffic flow data for the area where the target charging station is located, as well as the current time period information, which includes at least time, date type, and weather.

[0130] Historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations within a historical period are obtained as historical sample information. Based on the historical sample information, a preset peak power predictor is trained.

[0131] The real-time traffic flow data and current time period information are input into the peak power predictor to obtain the predicted peak power curve within a preset time window. The predicted peak power curve is composed of the predicted peak power at multiple time points.

[0132] This includes acquiring historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations over a historical period, as historical sample information, including:

[0133] Using vehicle counting equipment, the electric vehicle traffic per unit time within a preset range of the target charging station is obtained as potential passenger traffic;

[0134] Multiple existing charging stations were identified, their potential passenger flow was obtained for each, and compared with the potential passenger flow of the target charging station to obtain multiple passenger flow deviation coefficients.

[0135] A potential passenger flow deviation threshold is preset, and existing charging stations that meet the passenger flow deviation threshold are identified as similar charging stations.

[0136] Historical traffic flow, historical time period information, and corresponding historical peak power of similar charging stations are obtained as historical sample information.

[0137] Furthermore, the available power curve acquisition module 13 includes the following execution steps:

[0138] Historical load data and historical time period information of candidate nodes are obtained, load patterns of candidate nodes are analyzed based on machine learning, and multiple predicted peak powers of candidate nodes within a preset time window are predicted based on the current time period information. Based on the multiple predicted peak powers, the peak power curve of the node is plotted.

[0139] Based on the node design parameters, the maximum design capacity of the candidate nodes is obtained, and the peak power curve of the nodes is converted based on the maximum design capacity to obtain the node redundancy power curve.

[0140] Based on the aforementioned transmission loss coefficient, the node redundancy power curve is corrected to obtain the node available power curve.

[0141] Furthermore, the adaptation node selection module 14 includes the following execution steps:

[0142] The predicted peak power curve and the node available power curve are grouped into the same two-dimensional coordinate system, where the horizontal axis is the preset time window.

[0143] The ratio of the length of time during which the available power curve of a computing node is higher than the predicted peak power curve to the preset time window is used as the time adaptation coefficient.

[0144] Based on the curve characteristics of the node available power curve and the predicted peak power curve in the two-dimensional coordinate system, the excess power supply area and the total demand area are calculated respectively, and the ratio of the excess power supply area to the total demand area is used as the spatial adaptation coefficient.

[0145] The node power adaptation coefficient is obtained by weighted fusion of the time adaptation coefficient and the spatial adaptation coefficient.

[0146] Specifically, based on the curve characteristics of the node's available power curve and the predicted peak power curve in a two-dimensional coordinate system, the excess power supply area and the total demand area are calculated respectively. The ratio of the excess power supply area to the total demand area is used as the spatial adaptation coefficient, including:

[0147] The numerical integration method is used to calculate the area enclosed by the portion of the node's available power curve that is higher than the predicted peak power curve within a preset time window, which is taken as the excess power supply area.

[0148] Calculate the total area enclosed by the predicted peak power curve and the time axis within a preset time window, and use it as the total demand area;

[0149] Calculate the ratio of the excess power supply area to the total required area to obtain the space adaptation coefficient.

[0150] Example 3, as Figure 3 As shown, based on the same inventive concept as the multi-index fusion distribution network charging station access node screening method provided in Embodiment 1, this embodiment of the invention also provides an electronic device 300, including:

[0151] Memory 310 is used to store the first computer program 311;

[0152] The processor 320 is used to read and execute the first computer program 311, thereby implementing the multi-index fusion distribution network charging station access node screening method described in Embodiment 1.

[0153] The memory refers to a device in a computer used to temporarily store running programs and data, including but not limited to random access memory (RAM), read-only memory (ROM), virtual memory, etc.

[0154] The processor is a component in a computer responsible for executing program instructions, processing data, and controlling the operation of the computer. It includes, but is not limited to, general-purpose processors, graphics processing units (GPUs), and embedded processors.

[0155] Example 4, as Figure 4 As shown, based on the same inventive concept as the multi-index fusion distribution network charging station access node screening method provided in Embodiment 1, this embodiment of the invention also provides a storage medium 400. For example, the storage medium can be a non-transitory computer-readable storage medium, and the storage medium stores a second computer program 410. When the second computer program 410 is executed by a processor, it implements the multi-index fusion distribution network charging station access node screening method as described in Embodiment 1.

[0156] The non-transitory storage medium refers to a storage medium that can still retain data persistently after power failure, including but not limited to SSDs (solid-state drives), HDDs (hard disk drives), and flash memory devices (USB flash drives, memory cards), etc.

[0157] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0158] Those skilled in the art will understand that embodiments of the present invention can provide methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0159] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0160] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0161] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0162] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0163] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A multi-index fused power distribution network charging station access node screening method, characterized in that, The method comprises: calculating line cost coefficients and power transmission loss coefficients between the multiple candidate nodes and the target charging station, and calculating node loss coefficients of each candidate node based on the line cost coefficients and the power transmission loss coefficients; obtaining a predicted peak power curve in a preset time window through a preset peak power predictor based on traffic flow and current time period information of a region where the target charging station is located; obtaining a node peak power curve of the candidate node in the preset time window based on historical load data of the candidate node, and obtaining a node available power curve based on a design parameter of the node and the node peak power curve, and correcting the node available power curve based on the power transmission loss coefficients to obtain the node available power curve; comparing the node available power curve with the predicted peak power curve to obtain a time adaptation coefficient and a space adaptation coefficient, and weighting and fusing the time adaptation coefficient and the space adaptation coefficient to obtain a node power adaptation coefficient, calculating node adaptation coefficients of the multiple candidate nodes based on the node loss coefficients and the node power adaptation coefficients of the multiple candidate nodes, and selecting a node with the largest node adaptation coefficient as an access node; wherein the comparing of the node available power curve with the predicted peak power curve to obtain the time adaptation coefficient and the space adaptation coefficient, and the weighting and fusing of the time adaptation coefficient and the space adaptation coefficient to obtain the node power adaptation coefficient comprise: classifying the predicted peak power curve and the node available power curve into the same two-dimensional coordinate system, wherein a horizontal axis of the coordinate system is the preset time window; calculating a proportion of a time length during which the node available power curve is higher than the predicted peak power curve to the preset time window as the time adaptation coefficient; calculating an excess power supply area and a total demand area based on curve characteristics of the node available power curve and the predicted peak power curve in the two-dimensional coordinate system, and taking a ratio of the excess power supply area to the total demand area as the space adaptation coefficient; and weighting and fusing the time adaptation coefficient and the space adaptation coefficient to obtain the node power adaptation coefficient; wherein the calculating of the excess power supply area and the total demand area based on the curve characteristics of the node available power curve and the predicted peak power curve in the two-dimensional coordinate system, and the taking of the ratio of the excess power supply area to the total demand area as the space adaptation coefficient comprise: calculating, by using a numerical integration method, an area surrounded by a part of the node available power curve that is higher than the predicted peak power curve in the preset time window as the excess power supply area; calculating a total area surrounded by the predicted peak power curve and the time axis in the preset time window as the total demand area; and calculating a ratio of the excess power supply area to the total demand area to obtain the space adaptation coefficient.

2. The multi-metric fused distribution grid charging station access node screening method of claim 1, wherein, calculating line cost coefficients and power transmission loss coefficients between the multiple candidate nodes and the target charging station, and calculating node loss coefficients based on the line cost coefficients and the power transmission loss coefficients, comprises: establishing a power transmission line cost model based on a geographic information system; obtaining multiple best paths for constructing power transmission lines from the target charging station to the multiple candidate nodes according to the power transmission line cost model and positions of the target charging station and the multiple candidate nodes; calculating multiple line costs for constructing the multiple best paths, and obtaining multiple line cost coefficients based on the multiple line costs; The power transmission loss coefficient of each candidate node is calculated based on the length of the plurality of optimal paths; The node loss coefficient of each candidate node is calculated based on the line cost coefficient and the power transmission loss coefficient.

3. The multi-metric fused distribution grid charging station access node screening method of claim 1, wherein, Based on the traffic flow and the current time period information of the area where the target charging station is located, a predicted peak power curve within a preset time window is obtained through a preset peak power predictor, including: Obtain real-time traffic flow data and current time period information of the area where the target charging station is located, and the time period information at least includes time, date type, weather; Obtain historical traffic flow, historical time period information and corresponding historical peak power of similar charging stations in historical time as historical sample information, and train a preset peak power predictor based on the historical sample information; Input the real-time traffic flow data and the current time period information into the peak power predictor to obtain a predicted peak power curve within a preset time window, and the predicted peak power curve is composed of predicted peak power at multiple time points.

4. The multi-metric fused distribution grid charging station access node screening method of claim 3, wherein, Obtain historical traffic flow, historical time period information and corresponding historical peak power of similar charging stations in historical time as historical sample information, including: Use a vehicle counting device to obtain the electric vehicle flow per unit time within a preset range of the target charging station as potential passenger flow; Obtain the potential passenger flow of a plurality of already built charging stations respectively, and compare it with the potential passenger flow of the target charging station to obtain a plurality of passenger flow deviation coefficients; A potential passenger flow deviation threshold is preset, and the already built charging stations whose passenger flow deviation coefficients meet the passenger flow deviation threshold are obtained as similar charging stations; Obtain historical traffic flow, historical time period information and corresponding historical peak power of similar charging stations in historical time as historical sample information.

5. The multi-metric fused distribution grid charging station access node screening method of claim 1, wherein, Obtain the node peak power curve of the candidate node within a preset time window, and obtain the node available power curve based on the node design parameters and the node peak power curve, and correct the node redundant power curve based on the power transmission loss coefficient, including: Obtain historical load data and historical time period information of the candidate node, analyze the load law of the candidate node based on machine learning, and predict a plurality of predicted peak powers of the candidate node within a preset time window based on the current time period information, and draw the node peak power curve based on the plurality of predicted peak powers; Based on the design parameters of the node, obtain the maximum design capacity of the candidate node, and convert the node peak power curve based on the maximum design capacity to obtain the node redundant power curve; Correct the node redundant power curve based on the power transmission loss coefficient to obtain the node available power curve.

6. A multi-criteria fused distribution network charging station access node screening system, characterized in that, The system is used to realize the multi-index fusion power distribution network charging station access node screening method of any one of claims 1-5, including: A node loss coefficient acquisition module is configured to calculate the line cost coefficient and the power transmission loss coefficient between a plurality of candidate nodes and a target charging station, and to calculate the node loss coefficient of each candidate node based on the line cost coefficient and the power transmission loss coefficient; The peak power curve prediction module is configured to obtain a predicted peak power curve in a preset time window based on traffic flow and current period information of an area where the target charging station is located, by using a preset peak power predictor. The available power curve acquisition module is configured to obtain a node peak power curve of the candidate node in the preset time window based on historical load data of the candidate node, obtain a node redundant power curve based on a design parameter of the node and the node peak power curve, correct the node redundant power curve based on the power transmission loss coefficient, and obtain a node available power curve. The adaptive node selection module is configured to compare the node available power curve with the predicted peak power curve, obtain a time adaptation coefficient and a space adaptation coefficient, and fuse the time adaptation coefficient and the space adaptation coefficient as a node power adaptation coefficient; calculate node adaptation coefficients of the plurality of candidate nodes based on the node loss coefficients and the node power adaptation coefficients of the plurality of candidate nodes, and select a node with the largest node adaptation coefficient as the access node.

7. An electronic device, comprising: The memory is configured to store a first computer program. The processor is configured to read and execute the first computer program, thereby implementing the multi-index fusion power distribution network charging station access node screening method of any one of claims 1-5. The storage medium stores a second computer program, and the second computer program is executed by the processor to implement the multi-index fusion power distribution network charging station access node screening method of any one of claims 1-5.

8. A storage medium, characterized by ​

Images