Methods, devices, equipment, and storage media for delineating autonomous charging zones for electric vehicles

Abstract

This invention discloses a method, apparatus, device, and storage medium for delineating autonomous charging zones for electric vehicles. By constructing a behavioral spatiotemporal Laplace matrix and an electrical spatiotemporal Laplace matrix, spectral embedding and fusion are performed on these matrices to obtain a joint feature space containing behavioral and electrical embedding features. Spatiotemporal points are clustered within this joint feature space to obtain the autonomous charging zone delineation result corresponding to the target region. The method and apparatus provided by this invention achieve automatic delineation of autonomous charging zones for electric vehicles without manual parameter tuning, improving the adaptability and generalizability of the partitioning, and providing basic partitioning unit support for subsequent regional pricing, demand response, and refined control of the distribution network for charging zones.

Description

Technical Field

[0001] This invention relates to the field of intelligent control technology for electric vehicle charging, and in particular to a method, apparatus, equipment, and storage medium for dividing electric vehicle charging zones in a power distribution network. Background Technology

[0002] With the increasing popularity of electric vehicles, charging loads have become one of the more important loads in power distribution networks. Static zoning based solely on geographical administrative regions or electrical topology may lead to inconsistencies between zoning boundaries and actual load coupling boundaries. Manual division of autonomous regions, due to urban environments, different seasons, and varying penetration rates, makes it impossible to generalize the methods to other areas. Therefore, neither static nor manual zoning in existing technologies can achieve stable autonomous control for each autonomous unit within a given region.

[0003] Therefore, the existing technology needs further improvement. Summary of the Invention

[0004] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method, apparatus, device and storage medium for dividing autonomous charging areas for electric vehicles in power distribution networks, overcoming the defect that the autonomous charging areas for electric vehicles cannot be adaptively divided in the prior art.

[0005] The technical solution adopted by this invention to solve the technical problem is as follows:

[0006] In a first aspect, the present invention provides a method for dividing electric vehicle charging autonomous zones for power distribution networks, comprising:

[0007] Acquire electric vehicle user behavior data within the target area, construct a spatiotemporal graph of electric vehicle user behavior, and construct a spatiotemporal Laplace matrix of behavior using spatiotemporal points as graph nodes of the behavior spatiotemporal graph; wherein, the spatiotemporal points are constructed from the distribution network nodes and discrete time slices within the target area;

[0008] Obtain the electrical configuration and operation information within the target area, construct the electrical Laplace matrix, and perform spatiotemporal extension on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix;

[0009] The behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix are spectrally embedded and fused to obtain a joint feature space containing behavioral embedding features and electrical embedding features;

[0010] Clustering of spatiotemporal points in the joint feature space yields the autonomous charging region partitioning result corresponding to the target region.

[0011] Optionally, the step of acquiring electric vehicle user behavior data within the target area and constructing a spatiotemporal map of electric vehicle user behavior includes:

[0012] Obtain a set of distribution network nodes consisting of multiple distribution network nodes and a set of discrete time slices consisting of multiple discrete time slices. Use the state of each distribution network node in each discrete time slice as a spatiotemporal point to construct a set of spatiotemporal points.

[0013] Obtain migration and dwell data of electric vehicle users within the target area;

[0014] Using the migration data as migration edges, the dwell data as dwell edges, and each spatiotemporal point in the spatiotemporal point set as a node, a spatiotemporal graph of electric vehicle behavior is constructed.

[0015] Optionally, the step of constructing the spatiotemporal Laplace matrix by using the spatiotemporal points constructed from the distribution network nodes and discrete time slices within the target area as graph nodes of the behavioral spatiotemporal graph includes:

[0016] The migration edges and dwell edges are assigned values ​​according to the preset edge weights, and the weighted adjacency matrix of the behavior spatiotemporal graph is calculated.

[0017] The weighted adjacency matrix of the behavioral spatiotemporal graph is subjected to self-loop removal and symmetry processing to construct the behavioral spatiotemporal Laplace matrix.

[0018] Optionally, the steps of obtaining the electrical configuration and operation information within the target area, constructing the electrical Laplace matrix, and performing spatiotemporal extension on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix include:

[0019] Obtain information on power distribution network lines and transformer connections within the target area, and construct an electrical topology diagram based on the power distribution network lines and transformer connections.

[0020] The electrical edge weights are determined according to the reciprocal of the line impedance magnitude, the reciprocal of the line resistance, or a preset equivalent admittance, and a weighted adjacency matrix of the electrical topology is established.

[0021] The weighted adjacency matrix of the electrical topology graph is subjected to self-loop removal and symmetry processing to construct the electrical Laplace matrix of the distribution network;

[0022] The electrical Laplace matrix is ​​spatiotemporally extended along the time dimension to obtain the electrical spatiotemporal Laplace matrix.

[0023] Optionally, the step of spectrally embedding and fusing the behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features includes:

[0024] The behavior-spatiotemporal Laplacian matrix is ​​spectrally embedded to obtain the behavior embedding features;

[0025] The electrical spatiotemporal Laplace matrix is ​​spectrally embedded to obtain electrical embedding features;

[0026] The behavioral embedding features and electrical embedding features are concatenated to obtain a joint feature space.

[0027] Optionally, the step of concatenating the behavioral embedding features with the electrical embedding features to obtain the concatenated joint feature space includes:

[0028] The behavior embedding features are subjected to intra-block whitening and energy equalization scaling to obtain the processed behavior embedding features;

[0029] Furthermore, the electrical embedding feature is subjected to intra-block whitening and energy equalization scaling to obtain the processed electrical embedding feature;

[0030] The processed behavioral embedding features and the processed electrical embedding features are concatenated to obtain a joint feature space.

[0031] Optionally, the step of clustering spatiotemporal points in the joint feature space to obtain the autonomous charging region partitioning result corresponding to the target region includes:

[0032] A clustering algorithm is used to cluster spatiotemporal points in the joint feature space to obtain the region label of each spatiotemporal point;

[0033] Mapping the region label of each spatiotemporal point to the behavior spatiotemporal map yields a spatiotemporal distribution map of autonomous charging regions within the target region.

[0034] Secondly, this application provides a device for delineating electric vehicle charging zones in a power distribution network, comprising:

[0035] The behavior spatiotemporal matrix construction module is used to acquire electric vehicle user behavior data in the target area, construct an electric vehicle user behavior spatiotemporal graph, and use the spatiotemporal points constructed by the distribution network nodes and discrete time slices in the target area as the graph nodes of the behavior spatiotemporal graph to construct the behavior spatiotemporal Laplace matrix.

[0036] The electrical spatiotemporal matrix construction module is used to obtain the configuration and operation information of electrical systems within the target area, construct the electrical Laplace matrix, and perform spatiotemporal expansion on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix.

[0037] A joint space construction module is used to perform spectral embedding and fusion of the behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features;

[0038] The division output module is used to cluster spatiotemporal points in the joint feature space to obtain the autonomous charging region division result corresponding to the target region.

[0039] Thirdly, this application discloses a computer device, comprising: the computer device including a memory and a processor, wherein the memory stores a program, and when the program is executed by the processor, the processor is used to execute the electric vehicle charging autonomous region division method for power distribution networks.

[0040] Fourthly, this application discloses a computer storage medium, which is a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a computer, the computer is used to execute the method for dividing electric vehicle charging autonomous regions for distribution networks.

[0041] Beneficial effects:

[0042] This invention provides a method, apparatus, device, and storage medium for dividing autonomous charging zones for electric vehicles (EVs) in a power distribution network. It acquires EV user behavior data within a target area, constructs a spatiotemporal graph of EV user behavior, and uses spatiotemporal points constructed from power distribution network nodes and discrete time slices within the target area as graph nodes to construct a spatiotemporal Laplace matrix of behavior. It also acquires electrical configuration and operation information within the target area, constructs an electrical Laplace matrix, and spatiotemporally expands the electrical Laplace matrix to obtain an electrical spatiotemporal Laplace matrix. The invention performs spectral embedding and fusion on the behavior spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix to obtain a joint feature space containing behavior embedding features and electrical embedding features. Finally, it clusters spatiotemporal points in the joint feature space to obtain the autonomous charging zone division result corresponding to the target area. The method and apparatus provided by this invention can automatically divide EV charging zones without manual parameter tuning, improving the adaptability and generalizability of the partitioning, and providing technical support for subsequent autonomous zone control. Attached Figure Description

[0043] Figure 1 A flowchart illustrating the steps of the method for dividing electric vehicle charging autonomous regions for power distribution networks provided by the present invention;

[0044] Figure 2 The result principle block diagram of the electric vehicle charging autonomous region division device for power distribution network provided by the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0046] It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown in the flowchart. The terms "first," "second," etc., used in the specification, claims, and the foregoing drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0047] With the widespread adoption of electric vehicles, charging load has become one of the fastest-growing and most volatile new types of loads in the power distribution network. Currently, the charging behavior of electric vehicles exhibits significant spatiotemporal aggregation and migration coupling characteristics. That is, within the same time period, there is linkage between charging demand in different areas caused by commuting, operation, and travel chains. Within the same area, it is also affected by the capacity of charging infrastructure, price signals, and user preferences, resulting in load response with the characteristics of "strong correlation, strong non-uniformity, and strong time variation".

[0048] Existing technologies have proposed methods for achieving more intelligent power distribution control through vehicle-to-grid (V2G) interactive regulation and refined operation of the distribution network. V2G interactive regulation involves two-way energy and information interaction between electric vehicles and the power grid via charging equipment. For example, an electric vehicle, acting as a load, draws power from the grid for charging and simultaneously feeds power back to the grid. Refined operation of the distribution network, on the other hand, uses technologies such as communication control to monitor, analyze, and optimize the network's operational status in real time, thereby improving the reliability and economy of power supply.

[0049] In the context of interactive control of the power grid and refined operation of the distribution network, traditional control on the distribution network side is often based on electrical boundaries such as substations / feeders / transformer areas. However, the coupling relationship of electric vehicle charging demand does not propagate entirely along the electrical topology boundaries. This is because, on the one hand, user migration can cause highly synchronized load changes in some electrically distant nodes at the same time period; on the other hand, the electrical topology and impedance distribution determine the voltage / power flow sensitivity and risk propagation path, making load disturbances of adjacent nodes more likely to affect each other.

[0050] Therefore, when implementing zoned pricing, demand response guidance, and zoned autonomous control, static zoning based solely on geographic administrative regions or electrical topology often leads to the following problems:

[0051] (1) The partition boundary is inconsistent with the actual load coupling boundary, which means that the control signal needs to be linked across regions to be effective; partitioning by geographical / business rules alone is difficult to cover the load correlation caused by users migrating across regions, resulting in inconsistent load response within the same control unit and large fluctuations in control effect.

[0052] (2) The load response within the region is inconsistent, making it difficult to unify the control objectives within the same region, that is, the control granularity and electrical constraints are mismatched; partitioning only according to electrical topology may ignore the synchronous aggregation phenomenon on the behavioral side, resulting in an unclear effective range of control signals in the spatiotemporal dimension, making it difficult to form a stable autonomous management unit.

[0053] (3) The division of autonomous regions relies on manually set weights (e.g., “behavioral graph weights / electrical graph weights”). Due to insufficient robustness under different cities, seasons and penetration rates, it is difficult to form a generalizable methodology.

[0054] Therefore, there is a need for an automatic autonomous region partitioning method that can simultaneously characterize the behavioral coupling of electric vehicles and the electrical coupling of the power distribution network, without requiring manual weight adjustment, so as to provide stable and interpretable regional units for subsequent regional autonomous control.

[0055] To address the shortcomings of existing technologies, this invention proposes a method, device, equipment, and storage medium for dividing autonomous charging regions for electric vehicles in power distribution networks. By constructing a spatiotemporal map of electric vehicle user behavior and a spatiotemporal map of power distribution network electrical systems, spectral embedding is performed on both to form a joint feature space. Autonomous charging regions are obtained by clustering "node-time" spatiotemporal points in the joint feature space, achieving a unified characterization of behavioral coupling and electrical coupling, avoiding manual weight parameter tuning, and improving the adaptability and generalizability of the partitioning.

[0056] The following description, in conjunction with the accompanying drawings, provides a more detailed account of the method, apparatus, equipment, and storage medium for dividing autonomous charging zones for electric vehicles in a power distribution network as disclosed in this embodiment.

[0057] This invention provides a method for dividing electric vehicle charging autonomous regions for power distribution networks, such as... Figure 1 As shown, it includes:

[0058] Step S1: Obtain electric vehicle user behavior data within the target area, construct a spatiotemporal graph of electric vehicle user behavior, and construct a spatiotemporal Laplace matrix of behavior using spatiotemporal points as graph nodes of the behavior spatiotemporal graph; wherein, the spatiotemporal points are constructed from the distribution network nodes and discrete time slices within the target area;

[0059] This step first acquires electric vehicle (EV) user behavior data within the target area, and then uses this data to construct a spatiotemporal map of EV user behavior within that target area. Specifically, the EV user behavior data refers to the behavioral data generated by users when using EVs, primarily including user charging behavior data, vehicle operation preference data, and user driving habit data. Since this step divides the autonomous charging area based on the acquired EV user behavior data, it mainly utilizes migration and dwell-related data contained within the user charging behavior data and driving habit data.

[0060] In detail, user migration data refers to the spatial movement trajectories and relationships of users between different charging autonomous regions. It reflects the user's cross-regional charging selection logic and mainly includes: the association between the starting point and the destination, migration frequency and preferences, and the spatiotemporal characteristics of migration. The association between the starting point and the destination refers to the relationship between the user's previous charging location and their current charging location, where the current charging location is considered the destination and the previous charging location the starting point. Migration frequency and preferences refer to the number of times a specific user group travels between charging autonomous regions, such as the migration frequency from charging autonomous region A to charging autonomous region B. The spatiotemporal characteristics of migration refer to the time period of migration and the distance range of migration.

[0061] Charging dwell time data refers to the time-dimensional characteristics of users within a single charging zone, reflecting charging efficiency and user dwell time habits. It mainly includes: charging duration and dwell time distribution. Charging duration is the total time from vehicle connection to disconnection from the charging pile, including effective charging time and ineffective dwell time. Ineffective dwell time includes the time the vehicle is not moved promptly after charging is complete. Dwell time distribution refers to the peak time periods for user dwell time within the charging zone. For example, evening charging dwell time is 6-8 hours, while daytime dwell time is 1-2 hours.

[0062] Furthermore, after acquiring the aforementioned electric vehicle user behavior data, a spatiotemporal map of electric vehicle user behavior is constructed based on this data. This spatiotemporal map of electric vehicle user behavior is a chart that integrates the time and spatial dimensions of user charging behavior data for electric vehicles, visually presenting the behavioral characteristics of electric vehicle users across all scenarios. It maps user charging behavior onto two dimensions: a time axis and geographic spatial coordinates, to show the spatiotemporal distribution patterns of user charging behavior.

[0063] Furthermore, the step of acquiring electric vehicle user behavior data within the target area and constructing a spatiotemporal map of electric vehicle user behavior includes:

[0064] Step S11: Obtain a set of distribution network nodes consisting of multiple distribution network nodes and a set of discrete time slices consisting of multiple discrete time slices. Use the state of each distribution network node in each discrete time slice as a spatiotemporal point to construct a set of spatiotemporal points.

[0065] In order to construct a spatiotemporal map of user behavior, this step first obtains a set of distribution network nodes consisting of multiple distribution network nodes in the target area and a set of discrete time slices consisting of multiple discrete time slices. Based on the set of distribution network nodes and the set of discrete time slices, a set of spatiotemporal points is constructed.

[0066] In detail, the distribution network nodes are set up. The set of discrete time slices is Define the spacetime point as And by using a unified set of all distribution network nodes and discrete time slices, we obtain:

[0067] (1)

[0068] Therefore, the total number of spacetime points is ;

[0069] Where n is a single distribution network node and t is a single discrete time slice.

[0070] Step S12: Obtain migration and stay data of electric vehicle users within the target area.

[0071] The migration and dwell data in this step can come from charging orders, vehicle trajectory, trip origin and destination (OD), station entry and exit records, etc., or from simulation data; this invention does not limit the data source.

[0072] Step S13: Using the migration data as migration edges, the dwell data as dwell edges, and each spatiotemporal point in the spatiotemporal point set as a node, construct a spatiotemporal graph of the electric vehicle's behavior.

[0073] Once the migration data of electric vehicles within the target area is obtained, for any electric vehicle / user Let it be in discrete time slices The node where it is located is The value -1 indicates that the electric vehicle is in a travel state at this time, rather than staying at any node.

[0074] Once the migration and stay data are obtained, then for Two types of edges are constructed: dwell edges and migration edges. Based on the dwell edges and migration edges connecting each node, a spatiotemporal graph of electric vehicle behavior is constructed.

[0075] Furthermore, the step of constructing the spatiotemporal Laplace matrix by using the distribution network nodes within the target area and the spatiotemporal points constructed from discrete time slices as graph nodes of the behavioral spatiotemporal graph includes:

[0076] Step S14: Assign values ​​to the migration edges and dwell edges according to the preset edge weights, and calculate the weighted adjacency matrix of the behavior spatiotemporal graph.

[0077] In this step, weights are set for the dwell edges and migration edges constructed in step S13 above, and values ​​are assigned to the migration edges and dwell edges according to the preset edge weights, thereby calculating the weighted adjacency matrix.

[0078] Staying edge: If Then in Establish edges and accumulate weights;

[0079] Migrate edge: If And arrive after being in a travel state for several consecutive time periods. Then in Establish edges and accumulate weights, where i and j are arbitrary nodes.

[0080] Step S15: Perform self-loop removal and symmetry processing on the weighted adjacency matrix of the behavior spatiotemporal graph to construct the behavior spatiotemporal Laplace matrix.

[0081] Let the weighted adjacency matrix of the behavioral spatiotemporal graph be... Its elements can be obtained by accumulating the number of vehicles / migration frequency:

[0082] (2)

[0083] right After removing self-loops, nonnegating, and symmetricizing, an undirected behavioral graph is obtained:

[0084] (3)

[0085] Construct degree matrix Thus, we obtain the behavior-spacetime Laplace:

[0086] (4)

[0087] Symmetric normalization is used to improve spectral embedding stability:

[0088] (5)

[0089] in, It is the identity matrix. For degree matrix, It is an undirected behavioral graph.

[0090] Step S2: Obtain the electrical configuration and operation information within the target area, construct the electrical Laplace matrix, and perform spatiotemporal expansion on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix.

[0091] An electrical topology graph is established based on the topological connections of power distribution lines, transformers, etc. This includes constructing the electrical topology graph and electrical edge weights, constructing and normalizing the electrical Laplace matrix, spatiotemporal expansion and cross-graph node alignment, and characterizing the unified representation of the electrical coupling structure in the spatiotemporal coordinate system.

[0092] In detail, the steps of acquiring the electrical configuration and operation information within the target area, constructing the electrical Laplace matrix, and performing spatiotemporal extension on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix include:

[0093] Step S21: Obtain the distribution network line information and transformer connection relationship in the target area, and construct an electrical topology diagram based on the distribution network line information and transformer connection relationship.

[0094] Obtain information on line parameters (e.g., line length, type, impedance), transformer parameters (e.g., transformer capacity, turns ratio, connection group), and switchgear status (e.g., circuit breaker and disconnector location) within the target area's distribution network. Use a Geographic Information System (GIS) to obtain the line spatial coordinates, tower locations, and equipment attributes. Based on the connection relationships between transformers in the distribution network lines, construct an electrical topology map using substation outgoing terminals, transformer low-voltage sides, critical loads, circuit breakers, disconnectors, fuses, and other equipment as nodes, and the distribution lines connecting two nodes as edges.

[0095] Step S22: Determine the electrical edge weights according to the reciprocal of the line impedance magnitude, the reciprocal of the line resistance, or the preset equivalent admittance, and establish the weighted adjacency matrix of the electrical topology.

[0096] Establish an electrical adjacency matrix based on the connection relationships of power distribution network lines and transformers. For any line Its edge weights are defined according to the reciprocal of the line impedance modulus:

[0097] (6)

[0098] in , To prevent numerical divergence, a minimum impedance threshold is set. For transformer branches, a pre-set large weight can be used to reflect strong coupling.

[0099] Step S23: Perform self-loop removal and symmetry processing on the weighted adjacency matrix of the electrical topology graph to construct the electrical Laplace matrix of the distribution network.

[0100] Establish electrical network degree matrix The Laplace matrix of the electrical diagram is:

[0101] (7)

[0102] And construct a symmetric normalized electrical Laplace:

[0103] (8)

[0104] in, It is the identity matrix. For the degree matrix of the electrical network, It is an electrical adjacency matrix.

[0105] Step S24: Spatiotemporally extend the electrical Laplace matrix along the time dimension to obtain the electrical-spatial Laplace matrix.

[0106] Extending the electric Laplace to the spacetime dimension via the Kronecker product, we obtain the electric spacetime Laplace:

[0107] (9)

[0108] in, for The identity matrix is ​​indexed by using equation (1). ,make The node order and the spatiotemporal relationship of electric vehicles Figure 1 This ensures strict alignment of the two types of spatiotemporal graphs.

[0109] Step S3: Perform spectral embedding and fusion on the behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features.

[0110] In this step, the behavioral spatiotemporal Laplacian matrix and the electrical spatiotemporal Laplacian matrix are spectral embedded to obtain behavioral embedding features and electrical embedding features. This includes spectral embedding of the two types of spatiotemporal Laplacian matrices, intra-block whitening and energy equalization scaling, joint feature splicing and normalization, and obtaining a joint feature space representation for clustering.

[0111] The step of concatenating the behavioral embedding features with the electrical embedding features to obtain the concatenated joint feature space includes:

[0112] The behavioral embedding features are subjected to intra-block whitening and energy equalization scaling to obtain processed behavioral embedding features; and the electrical embedding features are subjected to intra-block whitening and energy equalization scaling to obtain processed electrical embedding features; the processed behavioral embedding features and the processed electrical embedding features are concatenated to obtain a concatenated joint feature space.

[0113] In detail, the steps of spectral embedding and fusing the behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features include:

[0114] The behavioral spatiotemporal Laplace matrix is ​​spectrally embedded to obtain behavioral embedding features, and the electrical spatiotemporal Laplace matrix is ​​spectrally embedded to obtain electrical embedding features. The behavioral embedding features and the electrical embedding features are concatenated to obtain a joint feature space.

[0115] In detail, this step first involves spectral embedding of the two types of spatiotemporal Laplaces.

[0116] right Take before Each minimum eigenvalue corresponds to an eigenvector, resulting in the behavior embedding matrix. ;right Take before The smallest eigenvalue corresponds to an eigenvector, resulting in the electrical embedding matrix. To eliminate scale differences, the embedded row vectors can be normalized:

[0117] (10)

[0118] in, For embedding row vectors, This is a preset value.

[0119] Secondly, intra-block whitening and energy equalization scaling are performed on the behavioral embedding matrix and the electrical embedding matrix, respectively.

[0120] By separately analyzing the behavior embedding matrix Electrical Embedded Matrix Zero-mean and ZCA whitening were performed to obtain , Then, adaptive scaling is performed according to its Frobenius norm to ensure that the energy of the two feature blocks is consistent, thus avoiding subjective manual weight assignment.

[0121] (11)

[0122] in, The average of the two energy sources can be obtained. This process does not rely on manually setting fixed weights. This is a preset value.

[0123] Next, the joint features are concatenated and normalized.

[0124] The behavioral embedding features and electrical embedding features are concatenated to obtain the joint feature matrix:

[0125] (12)

[0126] Among them, the behavior embedding matrix Electrical Embedded Matrix Zero-mean and ZCA whitening were performed to obtain , . To pair according to its Frobenius norm The first coefficient is obtained by adaptive scaling; According to their Frobenius norm pairs The second coefficient is obtained by adaptive scaling.

[0127] And on Row vectors are normalized to form a joint feature space, which is used to uniformly represent the spatiotemporal point locations of "behavioral coupling + electrical coupling".

[0128] Step S4: Cluster the spatiotemporal points in the joint feature space to obtain the autonomous charging region division result corresponding to the target region.

[0129] Clustering of spatiotemporal points in the joint feature space, including clustering of joint features, outputting autonomous region division results and mapping them back to the node-time grid; outputting the autonomous region label to which each spatiotemporal point belongs and mapping it back to the node-time grid, forming the spatiotemporal partitioning results of the autonomous charging region; the partitioning results can be used as the basis for dividing the management units of partition pricing, partition demand response and partition autonomous control.

[0130] In detail, the step of clustering spatiotemporal points in the joint feature space to obtain the autonomous charging region partitioning result corresponding to the target region includes:

[0131] Clustering algorithms are used to cluster the spatiotemporal points in the joint feature space to obtain the region label of each spatiotemporal point.

[0132] In practical implementation, K-means, spectral clustering, or density clustering can be used. Clustering is performed to obtain the region label for each spatiotemporal point. Taking K-means as an example, its optimization objective is to minimize the intra-cluster squared error:

[0133] (13)

[0134] in, For each joint feature point in the joint feature space, the k-th family of the joint feature points are defined. It is the center of the k-th cluster.

[0135] In detail, the silhouette coefficient of each joint specific point under different k values ​​is calculated. The closer the silhouette coefficient is to 1, the better the clustering effect. By comparing the average silhouette coefficient under different k values, the k value with the largest average silhouette coefficient is determined as the number of clusters.

[0136] Furthermore, k time points are selected from the joint feature space as initial cluster centers. For each spatiotemporal point, the distance between it and each cluster center is calculated, and each time point is assigned to the cluster corresponding to the nearest cluster center.

[0137] For each cluster, calculate the mean of all spatiotemporal points within it across all features, and use this mean as the new cluster center. Repeat the steps of assigning data points to clusters and updating cluster centers until a stopping condition is met. The stopping condition could be that the cluster centers no longer change significantly, or that a preset maximum number of iterations has been reached, yielding the final clustering result.

[0138] Mapping the region label of each spatiotemporal point to the behavior spatiotemporal map yields a spatiotemporal distribution map of autonomous charging regions within the target region.

[0139] According to formula (1), the label Mapping back to spacetime point This yields a spatiotemporal partition map of the autonomous charging region. Further, the spatiotemporal point set for each autonomous region can be output:

[0140] (14)

[0141] In one embodiment, the number of clusters It can be determined by the eigenvalue gap (eigengap), profile coefficient, or business-side constraints (e.g., the desired number of autonomous management units); this invention is not limited to this. The determination method for the eigenvalue gap is as follows: The eigenvalue gap is based on the theory of spectral clustering, and the optimal number of clusters is determined by analyzing the eigenvalue distribution of the Laplacian matrix of the data. The silhouette coefficient evaluates the clustering quality by quantifying the intra-cluster tightness and inter-cluster separation of each data point.

[0142] This embodiment discloses a method for dividing autonomous regions for electric vehicle charging in a distribution network. Under a unified "distribution network node-time slice" spatiotemporal coordinate system, it constructs a behavioral spatiotemporal graph of the travel-stay-migration relationship of electric vehicle users and forms a behavioral spatiotemporal Laplace matrix. Simultaneously, it constructs an electrical topology graph of the distribution network and forms an electrical Laplace matrix. Then, through spatiotemporal expansion, it obtains an electrical spatiotemporal Laplace matrix aligned with the nodes of the behavioral spatiotemporal graph. Spectral embedding is performed on both the behavioral and electrical spatiotemporal Laplace matrices to obtain a joint feature space reflecting behavioral and electrical coupling. Clustering of node-time spatiotemporal points within this joint feature space yields the spatiotemporal partitioning results for autonomous regions of electric vehicle charging. By applying intra-block whitening and energy adaptive equalization to the two types of embedded features, automatic division of autonomous regions can be achieved without manually setting behavioral and electrical weights, providing basic partitioning unit support for zoned pricing, demand response, and refined control of the distribution network.

[0143] Secondly, this application provides a device for dividing electric vehicle charging zones in a power distribution network, such as... Figure 2 As shown, it includes:

[0144] The behavior spatiotemporal matrix construction module 210 is used to acquire electric vehicle user behavior data within the target area, construct an electric vehicle user behavior spatiotemporal graph, and construct a behavior spatiotemporal Laplace matrix using spatiotemporal points as graph nodes of the behavior spatiotemporal graph; wherein, the spatiotemporal points are constructed by the distribution network nodes and discrete time slices within the target area; its function is as described in step S1.

[0145] On a unified set of distribution network nodes and discrete time slices, a spatiotemporal point set of "node-time" is constructed; electric vehicle migration and dwell data (including but not limited to order records, trajectory OD, arrival and departure events or simulation sequences) are acquired, and dwell edges and migration edges are established between spatiotemporal points to form a behavioral spatiotemporal graph that reflects the user's cross-node and cross-time period relationship; a behavioral spatiotemporal adjacency matrix is ​​constructed based on the behavioral spatiotemporal graph, and a behavioral spatiotemporal Laplace matrix is ​​further constructed as the spectral representation of the behavioral coupling structure.

[0146] The electrical spatiotemporal matrix construction module 220 is used to obtain the configuration and operation information of electrical systems within the target area, construct the electrical Laplace matrix, and perform spatiotemporal expansion on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix; its function is as described in step S2.

[0147] An electrical topology diagram is established based on the topological connections of distribution network lines, transformers, etc. Electrical edge weights are determined according to the reciprocal of impedance (or resistance) to construct the distribution network electrical Laplace matrix. The electrical Laplace matrix is ​​then spatiotemporally extended along the time dimension to form an electrical spatiotemporal Laplace matrix that is strictly aligned with the node index of the behavioral spatiotemporal diagram, which is used to characterize the unified representation of the electrical coupling structure in the spatiotemporal coordinate system.

[0148] The joint space construction module 230 is used to perform spectral embedding and fusion of the behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features; its function is as described in step S3.

[0149] The behavior-space-time Laplacian matrix and the electrical-space-time Laplacian matrix are spectral embedded to obtain behavior embedding features and electrical embedding features, respectively. The two embedding features are subjected to intra-block normalization / whitening and adaptive equalization scaling based on feature energy to avoid manually setting fixed fusion weights. The processed behavior embedding features and electrical embedding features are concatenated and normalized to obtain a joint feature space representation for clustering.

[0150] The division output module 240 is used to cluster spatiotemporal points in the joint feature space to obtain the autonomous charging region division result corresponding to the target region, and its function is as described in step S4.

[0151] Clustering of spatiotemporal points in the joint feature space of "node-time" is performed, and the label of the autonomous region to which each spatiotemporal point belongs is output and mapped back to the "node-time" grid to form the spatiotemporal partitioning result of the autonomous charging region; the partitioning result can be used as the basis for dividing the management units of partition pricing, partition demand response and partition autonomous control.

[0152] Thirdly, embodiments of this disclosure also provide a computer device, including: at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that are executed by the at least one processor to cause the at least one processor to perform the method as described in any of the embodiments of the first aspect of this application when executing the instructions.

[0153] The computer device includes: processor, memory, input / output interface, communication interface, and bus.

[0154] The processor can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this disclosure.

[0155] The memory can be implemented in the form of ROM (Read Only Memory), static storage device, dynamic storage device, or RAM (Random Access Memory). The memory can store the operating system and other applications. When the technical solutions provided in the embodiments of this specification are implemented through software or firmware, the relevant program code is stored in the memory and called by the processor to execute the energy scheduling method of the embodiments of this disclosure.

[0156] Input / output interfaces are used to implement information input and output;

[0157] The communication interface is used to enable communication and interaction between this device and other devices. Communication can be achieved through wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.); and the bus is used to transmit information between various components of the device (such as processor, memory, input / output interface and communication interface); among which the processor, memory, input / output interface and communication interface are connected to each other within the device through the bus.

[0158] This disclosure also provides a storage medium, which is a computer-readable storage medium storing computer-executable instructions for causing a computer to perform the methods of this disclosure.

[0159] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0160] This invention provides a method, apparatus, computer equipment, and computer storage medium for delineating autonomous charging zones for electric vehicle (EV) charging systems in power distribution networks. It belongs to the field of power system automation and vehicle-grid interaction control technology. The method includes: constructing a spatiotemporal map of EV user behavior reflecting user migration and dwell time relationships; constructing a power distribution network electrical topology map and its electrical Laplace matrix, and forming an electrical-spatial-temporal Laplace matrix aligned with the nodes of the behavior spatiotemporal map through spatiotemporal expansion; performing spectral embedding on the behavior-spatial-temporal Laplace matrix and the electrical-spatial-temporal Laplace matrix respectively to obtain a joint feature space; and clustering nodes-time-spatial points in the joint feature space to form autonomous charging zones. Using this embodiment, automatic delineation and quantification of the control impact range of autonomous charging zones can be achieved without manual parameter tuning, providing support for zoned pricing, demand response guidance, and power distribution network risk management.

[0161] The embodiments described in this disclosure are for the purpose of more clearly illustrating the technical solutions of this disclosure and do not constitute a limitation on the technical solutions provided by this disclosure. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by this disclosure are also applicable to similar technical problems.

[0162] It will be understood by those skilled in the art that Figure 1 The technical solutions shown in the figures do not constitute a limitation on the embodiments of this disclosure. They may include more or fewer steps than those shown in the figures, or combine certain steps, or different steps.

[0163] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0164] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0165] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus 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 apparatus.

[0166] It should be understood that in this application, "at least one (item)" means one or more, and "more than" means two or more. "And / or" is used to describe the relationship between related objects, indicating that three relationships can exist. For example, "A and / or B" can represent three cases: only A exists, only B exists, and both A and B exist simultaneously, where A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one (item) of a, b, or c can represent: a, b, c, "a and b", "a and c", "b and c", or "a and b and c", where a, b, and c can be single or multiple.

[0167] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0168] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0169] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0170] 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 this application, 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 multiple 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 this application. The aforementioned storage medium includes various media capable of storing programs, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0171] The preferred embodiments of the present disclosure have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present disclosure. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and spirit of the present disclosure shall be within the scope of the claims of the present disclosure.

Claims

1. A power distribution network-oriented autonomous division method of an electric vehicle charging area, characterized by, include: Acquire electric vehicle user behavior data within the target area, construct a spatiotemporal graph of electric vehicle user behavior, and construct a spatiotemporal Laplace matrix of behavior using spatiotemporal points as graph nodes of the behavior spatiotemporal graph; wherein, the spatiotemporal points are constructed from the distribution network nodes and discrete time slices within the target area; Obtain the electrical configuration and operation information within the target area, construct the electrical Laplace matrix, and perform spatiotemporal expansion on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix; The behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix are spectrally embedded and fused to obtain a joint feature space containing behavioral embedding features and electrical embedding features; Clustering of the spatiotemporal points in the joint feature space yields the autonomous charging region partitioning result corresponding to the target region. The step of spectrally embedding and fusing the behavioral spatiotemporal Laplacian matrix and the electrical spatiotemporal Laplacian matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features includes: The behavior-spatiotemporal Laplacian matrix is ​​spectrally embedded to obtain the behavior embedding features; The electrical spatiotemporal Laplace matrix is ​​spectrally embedded to obtain electrical embedding features; The behavior embedding features are subjected to intra-block whitening and energy equalization scaling to obtain the processed behavior embedding features; Furthermore, the electrical embedding feature is subjected to intra-block whitening and energy equalization scaling to obtain the processed electrical embedding feature; wherein, the energy equalization scaling is performed by automatically calculating the scaling factor based on the Frobenius norm to make the energy of the behavioral embedding feature and the electrical embedding feature consistent. The processed behavioral embedding features and the processed electrical embedding features are concatenated to obtain a joint feature space.

2. The power distribution grid-oriented autonomous division of electric vehicle charging zones method according to claim 1, characterized in that, The steps of acquiring electric vehicle user behavior data within the target area and constructing a spatiotemporal map of electric vehicle user behavior include: Obtain a set of distribution network nodes consisting of multiple distribution network nodes and a set of discrete time slices consisting of multiple discrete time slices. Use the state of each distribution network node in each discrete time slice as a spatiotemporal point to construct a set of spatiotemporal points. Obtain migration and dwell data of electric vehicle users within the target area; Using the migration data as migration edges, the dwell data as dwell edges, and each spatiotemporal point in the spatiotemporal point set as a node, a spatiotemporal graph of electric vehicle behavior is constructed.

3. The power distribution grid oriented electric vehicle charging autonomous zone division method according to claim 2, characterized by, The step of constructing the behavioral spatiotemporal Laplace matrix by using spatiotemporal points as graph nodes of the behavioral spatiotemporal graph includes: The migration edges and dwell edges are assigned values ​​according to the preset edge weights, and the weighted adjacency matrix of the behavior spatiotemporal graph is calculated. The weighted adjacency matrix of the behavioral spatiotemporal graph is subjected to self-loop removal and symmetry processing to construct the behavioral spatiotemporal Laplace matrix.

4. The method for dividing electric vehicle charging autonomous regions for power distribution networks according to claim 1, characterized in that, The steps of acquiring the electrical configuration and operation information within the target area, constructing the electrical Laplace matrix, and performing spatiotemporal extension on the electrical Laplace matrix to obtain the electrical spatiotemporal Laplace matrix include: Obtain information on power distribution network lines and transformer connections within the target area, and construct an electrical topology diagram based on the power distribution network lines and transformer connections. The electrical edge weights are determined according to the reciprocal of the line impedance magnitude, the reciprocal of the line resistance, or a preset equivalent admittance, and a weighted adjacency matrix of the electrical topology is established. The weighted adjacency matrix of the electrical topology graph is subjected to self-loop removal and symmetry processing to construct the electrical Laplace matrix of the distribution network; The electrical Laplace matrix is ​​spatiotemporally extended along the time dimension to obtain the electrical spatiotemporal Laplace matrix.

5. The method for dividing electric vehicle charging autonomous regions for power distribution networks according to claim 1, characterized in that, The step of clustering the spatiotemporal points in the joint feature space to obtain the autonomous charging region partitioning result corresponding to the target region includes: A clustering algorithm is used to cluster spatiotemporal points in the joint feature space to obtain the region label of each spatiotemporal point; Mapping the region label of each spatiotemporal point to the behavior spatiotemporal map yields a spatiotemporal distribution map of autonomous charging regions within the target region.

6. A device for delineating autonomous charging zones for electric vehicles in a power distribution network, characterized in that, include: The behavior spatiotemporal matrix construction module is used to acquire electric vehicle user behavior data in the target area, construct an electric vehicle user behavior spatiotemporal graph, and use the spatiotemporal points constructed by the distribution network nodes and discrete time slices in the target area as the graph nodes of the behavior spatiotemporal graph to construct the behavior spatiotemporal Laplace matrix. An electrical spatiotemporal matrix construction module is used to obtain the configuration and operation information of electrical systems within the target area, construct an electrical Laplace matrix, and perform spatiotemporal expansion on the electrical Laplace matrix to obtain an electrical spatiotemporal Laplace matrix. A joint space construction module is used to perform spectral embedding and fusion of the behavioral spatiotemporal Laplace matrix and the electrical spatiotemporal Laplace matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features; The partitioning output module is used to cluster the spatiotemporal points in the joint feature space to obtain the autonomous charging region partitioning result corresponding to the target region. The step of spectrally embedding and fusing the behavioral spatiotemporal Laplacian matrix and the electrical spatiotemporal Laplacian matrix to obtain a joint feature space containing behavioral embedding features and electrical embedding features includes: The behavior-spatiotemporal Laplacian matrix is ​​spectrally embedded to obtain the behavior embedding features; The electrical spatiotemporal Laplace matrix is ​​spectrally embedded to obtain electrical embedding features; The behavior embedding features are subjected to intra-block whitening and energy equalization scaling to obtain the processed behavior embedding features; Furthermore, the electrical embedding feature is subjected to intra-block whitening and energy equalization scaling to obtain the processed electrical embedding feature; wherein, the energy equalization scaling is performed by automatically calculating the scaling factor based on the Frobenius norm to make the energy of the behavioral embedding feature and the electrical embedding feature consistent. The processed behavioral embedding features and the processed electrical embedding features are concatenated to obtain a joint feature space.

7. A computer device, characterized in that, include: The computer device includes a memory and a processor, wherein the memory stores a program that, when executed by the processor, causes the processor to perform the electric vehicle charging autonomous region division method for a power distribution network as described in any one of claims 1 to 5.

8. A computer storage medium, wherein the storage medium is a computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a computer, is used by the computer to perform the electric vehicle charging autonomous region division method for a power distribution network as described in any one of claims 1 to 5.

Images