Charging distribution data determination method and device of charging scene and electronic equipment

Abstract

The invention discloses a charging distribution data determination method and device of a charging scene and electronic equipment. The method comprises the steps that target scene data are acquired, and the target scene data are scene data representing a target charging scene; a target space grid corresponding to the charging scene is determined, the target space grid comprises a plurality of nodes, and the nodes have a topological relation; extracting target scene features of the target scene data; determining feature distances between the target scene features and node corresponding features in the target space grid; determining a target scene category corresponding to the target charging scene according to the plurality of feature distances; and determining charging distribution data corresponding to the target charging scene according to the target scene category. The technical problem that it is difficult to accurately determine the charging distribution data of the charging scene to provide the charging strategy in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of data processing, and in particular to a method, device and electronic device for determining charging distribution data of a charging scenario. Background Art

[0002] In recent years, with the rapid growth of new energy vehicle ownership and the rapid development of charging infrastructure, the imbalance between the service capacity of electric vehicle charging facilities and the charging needs of users has become increasingly prominent. In order to study the distribution and prediction technology of electric vehicle user charging needs in multiple scenarios, understand the future charging needs in different regions, provide an accurate basis for charging facility planning, avoid excess or insufficient charging resources, guide the reasonable allocation of charging resources, assist the power department in planning grid upgrades, and help charging facility operators select regions, determine scales, and formulate charging strategies, it is urgent to propose a method for accurately predicting the distribution data of charging scenarios.

[0003] However, there is a technical problem in the related art that it is difficult to accurately determine the charging distribution data of the charging scenario in order to provide a charging strategy.

[0004] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0005] Embodiments of the present invention provide a method, device and electronic device for determining charging distribution data of a charging scenario, so as to at least solve the technical problem in the related art that it is difficult to accurately determine the charging distribution data of a charging scenario to provide a charging strategy.

[0006] According to one aspect of an embodiment of the present invention, a method for determining charging distribution data of a charging scenario is provided, comprising: acquiring target scenario data, wherein the target scenario data is scenario data representing a target charging scenario; determining a target spatial grid corresponding to the charging scenario, wherein the target spatial grid includes a plurality of nodes, and the plurality of nodes have a topological relationship; extracting target scenario features of the target scenario data, wherein the target scenario features include at least: geographic location features, functional attribute features, and behavioral pattern features; determining feature distances between the target scenario features and corresponding features of nodes in the target spatial grid; determining a target scenario category corresponding to the target charging scenario based on a plurality of feature distances; determining charging distribution data corresponding to the target charging scenario based on the target scenario category, wherein the charging distribution data includes at least one of the following: charging time distribution data and charging demand distribution data.

[0007] Optionally, before determining the target space grid corresponding to the charging scenario, it also includes: selecting multiple typical scenes as seed nodes to construct an initial space grid; determining the vector space distances between the features of the sample scene and the features of the multiple typical scenes; dividing the sample scenes into sample scene categories based on the multiple vector space distances; and adding the sample scenes as new nodes to the initial space grid based on the sample scene categories to obtain the target space grid.

[0008] Optionally, based on the sample scene category, the sample scene is added as a new node to the initial spatial grid to obtain a target spatial grid, including at least one of the following: when the sample scene category is a target typical scene, the sample scene is added as a new node to the initial spatial grid, and the sample scene node is connected to the target typical scene to obtain the target spatial grid; when the sample scene category is not a target typical scene, the sample scene is added as a new node to the initial spatial grid, and the sample scene node is used as a seed node to obtain the target spatial grid.

[0009] Optionally, before determining the charging distribution data corresponding to the target charging scene based on the target scene category, it also includes: determining the coupling coordination degree corresponding to the target spatial grid; determining the overall autocorrelation and local charging scene correlation corresponding to the target spatial grid based on the coupling coordination degree; determining the charging distribution data corresponding to multiple scene categories based on the overall autocorrelation and the local charging scene correlation.

[0010] Optionally, based on the overall autocorrelation and the local charging scene correlation, determining charging distribution data corresponding to multiple scene categories includes: determining a distribution function corresponding to the target space grid, wherein the distribution function includes at least one of the following: a Gaussian mixture distribution function, a multidimensional normal distribution probability density function; based on the distribution function, the overall autocorrelation and the local charging scene correlation, determining the charging distribution data corresponding to the multiple scene categories.

[0011] Optionally, based on the target scenario category, determining charging distribution data corresponding to the target charging scenario includes: determining the charging distribution data corresponding to the target scenario category; determining distinguishing features between the target charging scenario and the target scenario category; and adjusting the charging distribution data based on the distinguishing features to obtain charging distribution data corresponding to the target charging scenario.

[0012] Optionally, according to the sample scene category, the sample scene is added as a new node to the initial spatial grid to obtain a target spatial grid, including: according to the sample scene category, the sample scene is added as a new node to the initial spatial grid to obtain an intermediate spatial grid; and multiple nodes in the intermediate spatial grid are clustered to obtain the target spatial grid.

[0013] According to one aspect of an embodiment of the present invention, a device for determining charging distribution data of a charging scenario is provided, comprising: an acquisition module, for acquiring target scenario data, wherein the target scenario data is scenario data representing a target charging scenario; a first determination module, for determining a target spatial grid corresponding to the charging scenario, wherein the target spatial grid includes a plurality of nodes, and the plurality of nodes have a topological relationship; an extraction module, for extracting target scenario features of the target scenario data, wherein the target scenario features include at least: geographic location features, functional attribute features, and behavioral pattern features; a second determination module, for determining feature distances between the target scenario features and corresponding features of nodes in the target spatial grid; a third determination module, for determining a target scenario category corresponding to the target charging scenario based on a plurality of feature distances; a fourth determination module, for determining the charging distribution data corresponding to the target charging scenario based on the target scenario category, wherein the charging distribution data includes at least one of the following: charging time distribution data and charging demand distribution data.

[0014] According to one aspect of an embodiment of the present invention, an electronic device is provided, comprising: a processor; and a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement a method for determining charging distribution data of a charging scenario as described in any one of the above.

[0015] According to one aspect of an embodiment of the present invention, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device can execute a method for determining charging distribution data of a charging scenario as described in any one of the above items.

[0016] In an embodiment of the present invention, target scene data is obtained, and the target scene data is scene data representing a target charging scene. A target spatial grid corresponding to the charging scene is determined, wherein the target spatial grid includes multiple nodes, and the multiple nodes have a topological relationship. The target scene features of the target scene data are extracted, wherein the target scene features include at least: geographic location features, functional attribute features, and behavior pattern features. The characteristic distances between the target scene features and the corresponding features of the nodes in the target spatial grid are determined. Based on multiple characteristic distances, the target scene category corresponding to the target charging scene is determined. Based on the target scene category, the charging distribution data corresponding to the target charging scene is determined, wherein the charging distribution data includes at least one of the following: charging time distribution data and charging demand distribution data. It can be seen that by calculating the distance from the corresponding features of the nodes in the target spatial grid, the scene category can be divided, and the charging distribution data corresponding to the target charging scene can be directly determined based on the corresponding scene category, so as to achieve the effect of accurately determining the charging distribution data of the charging scene, thereby solving the technical problem that it is difficult to accurately determine the charging distribution data of the charging scene in the related art to provide a charging strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0018] Figure 1 is a flow chart of a method for determining charging distribution data of a charging scenario according to an embodiment of the present invention;

[0019] Figure 2 is a schematic diagram of the characteristics of the number of parked vehicles in different scenarios provided by an optional embodiment of the present invention;

[0020] Figure 3 is a schematic diagram of the number of vehicles traveling in different scenarios provided by an optional embodiment of the present invention;

[0021] Figure 4 is a curve fitting diagram at a single peak of a mixed distribution provided by an optional embodiment of the present invention;

[0022] Figure 5 is a charging start time distribution fitting curve provided by an optional embodiment of the present invention;

[0023] Figure 6 It is a schematic diagram of the density function form of the log-normal distribution of electric vehicle charging provided by an optional embodiment of the present invention;

[0024] Figure 7 is a schematic diagram of a charging time distribution curve fitting result provided in an optional embodiment of the present invention;

[0025] Figure 8 A scatter plot of geographical coordinates of parking locations provided for an optional embodiment of the present invention;

[0026] Fig. 9 is a histogram of average charging power after data cleaning provided in an optional embodiment of the present invention;

[0027] Fig.10 A schematic diagram of the number of users who are charging simultaneously for an electric vehicle in a simulated 7-day period provided by an optional embodiment of the present invention;

[0028] Fig.11 A schematic diagram of a simulated daily power curve of an electric vehicle provided in an optional embodiment of the present invention;

[0029] Fig.12 A schematic diagram of a simulated 7-day power curve of an electric vehicle provided in an optional embodiment of the present invention;

[0030] Fig.13 4 is a structural block diagram of a device for determining charging distribution data of a charging scenario according to an embodiment of the present invention. DETAILED DESCRIPTION

[0031] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0033] Example 1

[0034] According to an embodiment of the present invention, an embodiment of a method for determining charging distribution data of a charging scenario is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0035] Figure 1 is a flow chart of a method for determining charging distribution data for a charging scenario according to an embodiment of the present invention. Figure 1 As shown, the method comprises the following steps:

[0036] Step S102, acquiring target scene data, wherein the target scene data is scene data representing a target charging scene;

[0037] In step S102 provided in the present application, target scene data is obtained, that is, detailed data of a specific charging scene (such as a charging pile area in a shopping mall parking lot, a charging station in a highway service area, etc.) collected. These data may come from a variety of channels, such as sensor networks, user behavior records, geographic information systems (GIS), etc. The target scene data may include a variety of data that can reflect the physical characteristics of the charging scene (such as location, area, number of facilities), environmental conditions (such as weather, traffic conditions), and user activity patterns, etc., to provide a basis for subsequent analysis.

[0038] Step S104, determining a target space grid corresponding to the charging scenario, wherein the target space grid includes a plurality of nodes, and a topological relationship exists between the plurality of nodes;

[0039] In step S104 provided in the present application, a target spatial grid corresponding to the charging scenario is determined, and the target spatial grid includes multiple nodes, and multiple nodes have a topological relationship. That is, this step spatializes the charging scenario and constructs a spatial grid by dividing it into a series of interrelated nodes (nodes represent different types of charging stations / charging scenarios). These nodes also reflect the spatial structure and similarity between different charging scenarios through topological relationships (such as adjacent, contained, etc.).

[0040] It should be noted that the fineness of the spatial grid should be determined according to the analysis requirements. It should be detailed enough to capture key features, but not too complex to cause computational inefficiency.

[0041] Step S106, extracting target scene features of the target scene data, wherein the target scene features at least include: geographic location features, functional attribute features, and behavior pattern features;

[0042] In step S106 provided in the present application, the target scene features of the target scene data are extracted. That is, key features are extracted from the collected target scene data, and these features are divided into at least three categories: geographical location (such as longitude and latitude, distribution of surrounding facilities), functional attributes (such as charging pile type, power, cost), and user or vehicle behavior patterns (such as charging frequency, peak hours). Feature extraction is a key step in machine learning or data analysis, which can ensure the accuracy and efficiency of subsequent analysis.

[0043] It should be noted that in addition to the basic characteristics mentioned above, it may also include environmental characteristics (such as temperature and humidity), policy influences (such as subsidies and changes in electricity prices), etc.

[0044] Step S108, determining the feature distances between the target scene features and the corresponding features of the nodes in the target space grid;

[0045] In step S108 provided in the present application, the characteristic distances between the target scene features and the corresponding features of the nodes in the target spatial grid are determined. That is, the distance between each node and the target scene feature is calculated, where the distance can be a numerical difference or a result based on a certain similarity measure. The purpose is to quantify the distribution of the target scene features in the spatial grid. Determine which node in the target spatial grid is more similar to the target scene feature.

[0046] The method for calculating the feature distance depends on the type of feature and the target of the analysis, and may use Euclidean distance, Manhattan distance, cosine similarity, etc. This step provides the basis for mapping the target scene features to the spatial grid.

[0047] Step S110, determining a target scene category corresponding to the target charging scene based on a plurality of feature distances;

[0048] In step S110 provided in the present application, the target scene category corresponding to the target charging scene is determined based on multiple feature distances. That is, based on the calculated feature distance, a classification algorithm (such as a decision tree, a support vector machine, a neural network, etc.) is used to classify the target charging scene into predefined scene categories. These categories may be divided based on factors such as the similarity of charging requirements and the similarity of geographical locations. The determination of the target scene category helps to identify the commonalities and differences between different charging scenarios, and provides a basis for formulating targeted charging strategies. The accuracy of the classification depends on the rationality of the feature selection, the richness of the data, and the choice of the classification algorithm.

[0049] Step S112, determining charging distribution data corresponding to the target charging scenario according to the target scenario category, wherein the charging distribution data includes at least one of the following: charging time distribution data, charging demand distribution data.

[0050] In step S112 provided in the present application, the charging distribution data corresponding to the target charging scenario is determined according to the target scenario category, wherein the charging distribution data includes at least one of the following: charging time distribution data, charging demand distribution data. That is, according to the target scenario category, specific charging distribution data is generated using historical data or prediction models or distribution data corresponding to typical scenarios. These data describe the distribution and changing trends of charging demand at different times and under different conditions. Charging distribution data is of great significance for optimizing the construction of charging facilities, formulating charging strategies, and predicting power demand.

[0051] Through the above steps S102-S112, the target scene data is obtained, and the target scene data is the scene data representing the target charging scene. The target spatial grid corresponding to the charging scene is determined, wherein the target spatial grid includes multiple nodes, and the multiple nodes have a topological relationship. The target scene features of the target scene data are extracted, wherein the target scene features at least include: geographic location features, functional attribute features, and behavior pattern features. The characteristic distances between the target scene features and the corresponding features of the nodes in the target spatial grid are determined. According to multiple characteristic distances, the target scene category corresponding to the target charging scene is determined. According to the target scene category, the charging distribution data corresponding to the target charging scene is determined, wherein the charging distribution data includes at least one of the following: charging time distribution data and charging demand distribution data. It can be seen that by calculating the distance from the corresponding features of the nodes in the target spatial grid, the scene category can be divided, and the charging distribution data corresponding to the target charging scene can be directly determined according to the corresponding scene category, so as to achieve the effect of accurately determining the charging distribution data of the charging scene, thereby solving the technical problem that it is difficult to accurately determine the charging distribution data of the charging scene in the related technology to provide a charging strategy.

[0052] As an optional embodiment, before determining the target space grid corresponding to the charging scene, it also includes: selecting multiple typical scenes as seed nodes to construct an initial space grid; determining the vector space distances between the characteristics of the sample scene and the characteristics of the multiple typical scenes; dividing the sample scenes into sample scene categories based on the multiple vector space distances; and adding the sample scenes as new nodes to the initial space grid based on the sample scene categories to obtain the target space grid.

[0053] In this embodiment, multiple typical scenes are selected as seed nodes to construct the initial spatial grid. At this stage, it is first necessary to screen out several representative typical scenes from a large number of charging scenes. These typical scenes should be able to cover the diversity of different geographical locations, functional attributes and user behavior patterns. These selected typical scenes will serve as seed nodes for constructing the initial spatial grid. The selection of seed nodes can be determined by expert evaluation, data analysis, or a combination of the two. When constructing the initial spatial grid, these seed nodes will serve as the starting point of the grid, laying the foundation for the subsequent addition of more nodes.

[0054] Then, the vector space distances between the features of the sample scene and the features of multiple typical scenes are determined. That is, the feature distances between other sample scenes that are not selected as seed nodes and multiple typical scenes are calculated. Vector space distance is used as a metric here, that is, by representing the features of each scene as a vector, and then calculating the Euclidean distance, cosine similarity or other suitable distance metrics between these vectors. The purpose of this step is to quantify the similarity between the sample scene and the typical scene, and provide a basis for the subsequent classification of the sample scene categories.

[0055] In order to divide the sample scene categories of the sample scene according to multiple vector space distances. Based on the calculated vector space distance, a clustering algorithm (such as K-means, hierarchical clustering, etc.) can be used to divide the sample scene into different categories, or the category with the shortest distance can be directly determined as the target scene category. These categories reflect the distribution of sample scenes in the feature space and their similarity with typical scenes.

[0056] Then, according to the sample scene category, the sample scene is added as a new node to the initial space grid to obtain the target space grid. According to the category of the sample scene, they are added as new nodes to the initial space grid. These newly added nodes and the original seed nodes together constitute the target space grid. The target space grid not only contains multiple typical scenes as seed nodes, but also covers more sample scenes with different characteristics as new nodes, thus more comprehensively reflecting the diversity and complexity of charging scenarios.

[0057] By selecting multiple typical scenarios as seed nodes and adding more sample scenarios with different characteristics as new nodes, the target space grid can more comprehensively cover different types of charging scenarios, improving the representativeness of the grid. Since the target space grid contains more diverse nodes, it can more accurately reflect the actual situation and improve the accuracy of the analysis when analyzing the characteristics of charging scenarios, determining scenario categories, and predicting charging distribution data. By constructing the target space grid and dividing the scenario categories, we can more clearly understand the needs and characteristics of different charging scenarios, thereby providing strong support for optimizing the construction of charging facilities. For example, the type, quantity, and layout of charging piles can be reasonably configured according to the needs of different scenarios.

[0058] As an optional embodiment, according to the sample scene category, the sample scene is added as a new node to the initial spatial grid to obtain a target spatial grid, including at least one of the following: when the sample scene category is a target typical scene, the sample scene is added as a new node to the initial spatial grid, and the sample scene node is connected to the target typical scene to obtain the target spatial grid; when the sample scene category is not a target typical scene, the sample scene is added as a new node to the initial spatial grid, and the sample scene node is used as a seed node to obtain the target spatial grid.

[0059] In this embodiment, in the process of constructing the target spatial grid, adding the sample scene as a new node to the initial spatial grid according to the sample scene category is a key step. This process involves the judgment of the sample scene category and the corresponding processing strategy. It is divided into two cases for processing. When the sample scene category is a target typical scene, it means that the scene has similar features or attributes to the seed node (i.e., the selected typical scene) in the initial spatial grid. At this time, the sample scene is directly added to the initial spatial grid as a new node, and a connection is established between the sample scene node and the corresponding target typical scene node. Through such processing, the target spatial grid not only expands the number of nodes, but also enhances the relevance and representativeness between nodes.

[0060] When the sample scene category is not a target typical scene, that is, when the sample scene is classified as a non-target typical scene, it means that the scene has different characteristics or attributes from the seed nodes in the initial spatial grid. At this time, although the sample scene does not meet the criteria of the target typical scene, it still has important information value. Therefore, such a sample scene is added to the initial spatial grid as a new node and is regarded as a potential seed node. This means that in the future analysis and optimization process, these non-target typical scene nodes may be re-evaluated and upgraded to typical scene nodes. At the same time, such a processing strategy also enhances the flexibility and scalability of the target spatial grid.

[0061] In practical applications, in order to more accurately judge the sample scene category and process different types of sample scenes, as the charging scenes continue to change and data accumulates, the structure and node relationship of the target space grid are dynamically adjusted to adapt to new scene needs and analysis requirements.

[0062] As an optional embodiment, before determining the charging distribution data corresponding to the target charging scene based on the target scene category, it also includes: determining the coupling coordination degree corresponding to the target space grid; determining the overall autocorrelation and local charging scene correlation corresponding to the target space grid based on the coupling coordination degree; determining the charging distribution data corresponding to multiple scene categories based on the overall autocorrelation and local charging scene correlation.

[0063] In this embodiment, before determining the charging distribution data corresponding to the target charging scenario according to the target scenario category, a series of pre-steps are required to ensure the accuracy and reliability of the charging distribution data.

[0064] The coupling coordination degree corresponding to the target space grid can be determined first. Coupling coordination is an indicator that measures the degree of correlation and interaction between different systems or elements and the coordination and consistency between them. In charging scenario analysis, coupling coordination can be used to measure the correlation and coordination between each node (i.e., charging scenario) in the target space grid. This usually involves a comprehensive consideration of multiple dimensions such as the distance between nodes, the distribution of charging facilities, and user behavior patterns. By calculating the coupling coordination degree, the overall structure of the target space grid and the interaction relationship between nodes can be understood, providing a basis for subsequent analysis.

[0065] In order to determine the overall autocorrelation and local charging scene correlation corresponding to the target space grid based on the coupling coordination degree. That is, after determining the coupling coordination degree, it is necessary to further analyze the overall autocorrelation and local charging scene correlation of the target space grid. The overall autocorrelation refers to the overall trend and law of the distribution of charging scenes in the entire target space grid, which reflects the degree of agglomeration or dispersion of the charging scenes in space. The local charging scene correlation refers to the correlation between charging scenes in a specific area, which reveals the mutual influence and dependency between different charging scenes. By analyzing the overall autocorrelation and local charging scene correlation, we can have a deeper understanding of the spatial distribution characteristics of charging scenes.

[0066] Finally, based on the overall autocorrelation and the local charging scene correlation, the charging distribution data corresponding to multiple scene categories are determined. That is, after mastering the overall autocorrelation and local charging scene correlation of the target spatial grid, the charging distribution data corresponding to multiple scene categories can be determined based on this information. This usually involves classifying the charging scenes (such as by geographic location, user type, charging facility type, etc.), and then calculating the charging distribution data for each category separately. The charging distribution data may include multiple dimensions such as charging amount, charging time, and charging frequency. In this way, the distribution data corresponding to the scene type can be determined in advance, which improves the efficiency of subsequent determination of the distribution data.

[0067] As an optional embodiment, charging distribution data corresponding to multiple scene categories are determined based on overall autocorrelation and local charging scene correlation, including: determining a distribution function corresponding to a target space grid, wherein the distribution function includes at least one of the following: a Gaussian mixture distribution function, a multidimensional normal distribution probability density function; based on the distribution function, overall autocorrelation and local charging scene correlation, charging distribution data corresponding to multiple scene categories are determined.

[0068] In this embodiment, when determining the charging distribution data corresponding to multiple scene categories, the appropriate distribution function can be selected based on the overall autocorrelation and the local charging scene correlation, combined with the characteristics of the target space grid. The distribution function corresponding to the target space grid can be determined first. The distribution function is a mathematical tool that describes the distribution characteristics of data, which can help us understand and predict the distribution law of data. In the charging scene analysis, the appropriate distribution function can be selected according to the characteristics of the target space grid and the characteristics of the charging data.

[0069] Among them, the Gaussian mixture distribution function is involved. The Gaussian mixture distribution is a composite distribution composed of multiple Gaussian distributions (normal distributions). It can flexibly capture the complex distribution characteristics of the data, especially when the data has multiple peaks or different distribution forms. In the charging scenario analysis, if the charging data shows obvious multi-peak distribution or heterogeneity under different scenario categories, then the Gaussian mixture distribution function will be a suitable choice.

[0070] Among them, the multidimensional normal distribution probability density function is also involved. The multidimensional normal distribution probability density function is used to describe the joint distribution of multiple variables in multidimensional space. In the charging scenario analysis, if the charging data involves multiple related variables (such as charging amount, charging time, charging frequency, etc.), and there is a certain correlation between these variables, then the multidimensional normal distribution probability density function will be able to more accurately describe the joint distribution characteristics of these variables.

[0071] Then, based on the distribution function, the overall autocorrelation and the local charging scene correlation, the charging distribution data corresponding to multiple scene categories are determined. After determining the appropriate distribution function, the overall autocorrelation and the local charging scene correlation are combined to further determine the charging distribution data corresponding to multiple scene categories. Because the overall autocorrelation reflects the overall trend and law of the distribution of charging scenes in the entire target space grid. By calculating the overall autocorrelation, the degree of agglomeration or dispersion of the charging scenes in space can be understood, and the overall distribution characteristics of the charging data can be accurately described in combination with the distribution function. Finally, according to the characteristics of multiple scene categories, the charging distribution data for each scene category is calculated separately. In this process, the parameters of the distribution function can be estimated and adjusted to ensure that the distribution function can accurately describe the distribution characteristics of the charging data under each scene category.

[0072] By selecting a suitable distribution function and combining the overall autocorrelation and local charging scenario correlation for analysis, the charging distribution data corresponding to multiple scenario categories can be determined more accurately, improving the accuracy and reliability of the data. By refining the scenario categories and considering local correlations, we can have a deeper understanding of the charging demand and distribution characteristics under different scenarios, providing more refined support for charging facility planning and operation strategy formulation. Based on accurate charging distribution data, more scientific and reasonable charging facility layout plans and optimization strategies can be formulated to improve the utilization rate and service level of charging facilities.

[0073] As an optional embodiment, charging distribution data corresponding to the target charging scene is determined based on the target scene category, including: determining the charging distribution data corresponding to the target scene category; determining the distinguishing features between the target charging scene and the target scene category; adjusting the charging distribution data based on the distinguishing features to obtain the charging distribution data corresponding to the target charging scene.

[0074] In this embodiment, when determining the charging distribution data corresponding to the target charging scenario, the target scenario category, the specific characteristics of the target charging scenario and the differences between them can be comprehensively considered to ensure that the obtained charging distribution data can accurately reflect the actual needs of the target charging scenario.

[0075] Determine the charging distribution data corresponding to the target scenario category, that is, determine the corresponding charging distribution data according to the characteristics of the target scenario category. This usually involves analyzing the existing charging data to understand the general laws and trends of charging activities under the scenario category. These laws may include the distribution of charging amount, preference for charging time, changes in charging frequency, etc. Through statistical analysis and data mining technology, the main characteristics of charging distribution under the target scenario category can be extracted, and the initial charging distribution data model can be constructed accordingly. Then determine the distinguishing features between the target charging scenario and the target scenario category. These differences may come from multiple aspects such as geographical location, user group, type of charging facility, regulations, etc. For example, a specific charging station may be located in a commercial center, and its user group is mainly office workers and shoppers, while the target scenario category in the area may be an urban public charging scenario. In this case, the target charging scenario may have higher charging demand and more concentrated charging time, which may be different from the general urban public charging scenario. At this time, the charging distribution data can be adjusted according to the distinguishing features. This may involve multiple aspects such as the redistribution of charging amount, adjustment of charging time, and change of charging frequency. So that the charging distribution data can more accurately reflect the actual needs and characteristics of the target charging scenario. By comprehensively considering the specific characteristics of the target scenario category and the target charging scenario, as well as the differences between them, the charging distribution data corresponding to the target charging scenario can be determined more accurately, thereby improving the accuracy and practicality of the data.

[0076] As an optional embodiment, according to the sample scene category, the sample scene is added as a new node to the initial space grid to obtain a target space grid, including: according to the sample scene category, the sample scene is added as a new node to the initial space grid to obtain an intermediate space grid; and multiple nodes in the intermediate space grid are clustered to obtain a target space grid.

[0077] In this embodiment, in the process of constructing the target space grid, based on the existing initial space grid and according to the new sample scene category, the sample scene can be added to the grid as a new node. In order to ensure the rationality and effectiveness of the grid, it is also necessary to further optimize the grid after adding the new node, such as clustering. That is, according to the sample scene category, the sample scene is added to the initial space grid as a new node to obtain the intermediate space grid. The sample scene is analyzed by category to clarify the scene category to which each sample scene belongs. After determining the scene category of the sample scene, these scenes are added to the initial space grid as new nodes. The addition of new nodes may change the structure and shape of the grid, and the connectivity and integrity of the grid need to be ensured. After adding the new node, an intermediate space grid containing the initial node and the new node is obtained. This grid may contain some redundant or repeated nodes, and there may also be some unreasonable grid divisions. Therefore, multiple nodes in the intermediate space grid can be clustered to obtain the target space grid. In order to optimize the intermediate space grid, the nodes in the grid can be clustered. The selection of clustering algorithm depends on the characteristics of the grid and the distribution of nodes. The selected clustering algorithm is applied to cluster the nodes in the intermediate space grid. The purpose of clustering is to classify similar nodes into one category to reduce redundancy and duplication in the grid while maintaining reasonable division of the grid. According to the clustering results, the grid is re-divided to obtain the target space grid. The target space grid has clear node division, reasonable grid shape and efficient connectivity.

[0078] Based on the above embodiments and optional embodiments, an optional implementation is provided, which is described in detail below.

[0079] Optional implementation methods of the present invention The purpose of the present invention is to provide a method for predicting the charging demand of electric vehicles based on a complex coupling network, to build a more complete regional spatial charging demand simulation system, to interact and influence each other through two or more charging piles, to use the coupling degree to measure the degree of coordination between charging piles, to effectively improve the scientificity and feasibility of charging facility planning, and to make full and effective use of charging pile resources. By studying the random variables of the charging start time of each electric vehicle, the distribution characteristics of the charging start time and the charging duration are considered. At the same time, using the geographic longitude and latitude as the coordinate system, the distribution of electric vehicle parking locations and charging locations is analyzed, with special attention paid to the charging demand in areas with dense parking locations. The power regulation capacity of the charging pile is analyzed, and the distribution characteristics of the average charging power are explored. Based on the distribution of the number of charging piles in the parking number, the relationship between the number of parking electric vehicles and the number of charging is studied, as well as the distribution of the number of charging piles under the condition of a certain number of parking piles.

[0080] In the related art, there are the following disadvantages when determining charging distribution data:

[0081] Limitations of the time distribution of charging demand: Current research mainly focuses on the distribution of charging start time and charging duration, but less consideration is given to factors such as the changing trend of charging demand in different time periods and seasonal changes, which leads to certain limitations in the accurate prediction of charging demand.

[0082] Incomplete distribution of charging demand locations: Although the distribution of electric vehicle parking locations and charging locations has been analyzed, the research on charging demand and its changing patterns in different regions is not in-depth enough, and there is a lack of detailed analysis of spatial distribution characteristics.

[0083] Simplification of charging power distribution: Current research mainly focuses on the distribution of average charging power, but there is little research on how charging piles with power regulation capabilities can optimize the control of charging power to adapt to different charging needs. This leads to the need to further improve the energy efficiency and efficiency of the charging process.

[0084] In view of the deficiencies of the above-mentioned prior art, an optional implementation manner of the present invention proposes a method for predicting the charging demand of electric vehicles based on a complex coupling network. By establishing a complex coupling network model, an optional implementation manner of the present invention can better describe the correlation and complexity between the various charging piles in the electric vehicle charging system. This network model can not only reveal the spatial distribution characteristics between the charging piles, but also analyze the temporal dynamic characteristics of the charging behavior, providing a richer information basis for the prediction model. At the same time, the method has certain flexibility and adaptability, and can be dynamically adjusted according to the changes in charging demand in different regions and different time periods. Compared with the prior art, it has obvious advantages in the comprehensiveness of data analysis, the accuracy of the prediction model, multi-dimensional data analysis, the optimization of charging facility planning, and the optimization of efficiency and energy efficiency. It is introduced below:

[0085] An optional implementation of the present invention provides a method for predicting charging demand of electric vehicles based on a complex coupling network, wherein the improvement lies in that based on multi-dimensional information such as electric vehicle charging data, charging pile information (pile location information (latitude and longitude coordinates), charging pile type (DC or AC)), external environmental factors (weather data, temperature data, traffic flow data and holiday information), electric vehicle information, charging user information), a typical network structure is formed through the edges between specific nodes, wherein nodes are used to represent individuals of different charging stations, and edges are used to represent the relationship between nodes, and often there is an edge if there is a certain specific relationship between two nodes, otherwise there is no connection, and two nodes connected by an edge are regarded as adjacent in the network. For the convenience of calculation, we usually use an adjacency matrix to represent the network. According to the different types of edges of the network, the network can be divided into unweighted undirected network, directed network, and weighted network. Through the interaction between two or more charging piles and the influence of each other, the degree of coordination between the charging piles is measured by coupling degree, and the corresponding electric vehicle spatiotemporal charging demand model is constructed, and these parameters or digital features are calculated by observation or sampling test of the model or process, and finally the approximate value of the solution is given. The method comprises the following steps:

[0086] (1) Based on urban planning and geographic location information, select a small amount of typical target scene information to construct spatial grids, and extract the data features of these grids as known seed sample grids;

[0087] (2) Extract the features of other grids and calculate their vector space distance with the seed sample grid. If the vector space distance between a grid and a scene seed grid is less than the weight Q ij , then the grid is divided into the scene, otherwise it is divided into the undetermined scene;

[0088] (3) Through step (2), K new divisible samples can be obtained. The new divisible samples are added to the seed samples, and the vector space distance calculation in step (2) is repeated to further classify the remaining grids to be divided. After each iteration, if any of the following conditions is met, the iteration is stopped:

[0089] ① The number of samples that can be divided reaches the set threshold of the number of charging piles;

[0090] ②The vector space distance of each sample is greater than the minimum space distance between it and all other divided samples.

[0091] (4) Based on the actual problem characteristics of electric vehicle charging demand prediction, combined with multi-dimensional data such as charging pile distribution, user behavior, and external environment, the coupling relationship between nodes (charging stations) is quantified. By using methods such as probability distribution or mathematical expectation, the charging demand of each node is accurately estimated;

[0092] (5) The sampling methods of various random variables with different distributions in the model are given, the simulation results are statistically processed, the charging demand of each charging station (node) is predicted, and the statistical estimation value of each node is obtained. At the same time, the accuracy estimation value of the predicted value of each node is calculated to evaluate the accuracy of the prediction result.

[0093] According to the different distribution random variables obtained in the model, these variables are processed using sampling methods, and finally the charging demand forecast results of each charging station (node) in a specific time period are counted, that is, the statistical estimate of each node. At the same time, for the prediction results of each node, its accuracy estimate is calculated to evaluate the reliability and accuracy of the prediction.

[0094] This patent mainly uses complex coupling networks to divide electric vehicle users into different grid characteristics, which can be specifically reflected in the four attributes of name, address, coordinates, and category, and divides the area into four types: residential area, industrial area, commercial area, and public service area. According to different grid types, in the process of charging demand prediction, the charging demand of electric vehicles is constructed for prediction and analysis.

[0095] There are many different types of complex networks in real life. In the electric vehicle charging network, by analyzing the charging stations and user charging behaviors, potential user groups can be discovered and personalized charging services can be provided to them. At the same time, the charging network can also analyze the user's charging behavior, reveal the characteristics of charging demand in different regions and time periods, and provide data support for the planning and optimization of charging facilities. In addition to representing the relationship between nodes through topological structures, complex networks with attributes also use attribute information to represent each node. Complex networks with attributes are usually represented by a topological graph G = (V, E, W, F), where V represents the node (grid attribute characteristics), E represents the edge relationship between nodes (similarity between charging stations), W is the weight of the edge (the strength of the connection between two charging stations or user-charging stations), and F represents the attributes of each node (describing the characteristics of the charging station).

[0096] (1) Constructing complex coupled network models:

[0097] By normalizing different grids in the space, all charging station variables v i And the weight w caused by the entropy weight method i , calculate the score of each charging station, and its score is

[0098] The coupling degree is:

[0099]

[0100] Where: X i represents: the number of electric vehicles; Yi It represents: distribution of electric vehicle charging stations; Z i Represents: economy and other factors.

[0101] Complex coupling coordination:

[0102]

[0103] In the charging demand forecast, we believe that the spatial distribution of charging stations, charging data, and quantity factors are equally important, and the weights are set as follows:

[0104] (2) Spatial autocorrelation of complex coupled networks:

[0105] 1. Analyze the overall autocorrelation, which can be expressed as:

[0106]

[0107] Among them, z i is the deviation of charging station i from its mean n is the total number of charging stations, Q ij represents the spatial weights of charging stations i and j, and T0 is the aggregation of the spatial weights of all charging stations, that is:

[0108] 2. Analyze the spatial correlation of local charging stations, which can be expressed as:

[0109]

[0110] in, n is the total number of charging stations, Q ij Represents the spatial weight of charging station i and charging station j.

[0111] (3) Gaussian mixture distribution:

[0112] Gaussian mixture distribution is an extension of normal distribution, using a combination of multiple normal distributions to describe data distribution. The probability density function of Gaussian mixture distribution is:

[0113]

[0114] Its distribution is the sum of K normal distributions, each of which has its own μ and σ parameters, as well as a corresponding weight parameter α i , the weight value must be a positive number, and the sum of all weights must be equal to 1 to ensure that the value given by the formula is a reasonable probability density value, that is, the integral value of the probability density function over the domain is 1.

[0115] According to the probability density function of the Gaussian mixture distribution, the specific form of its distribution depends on three parameters: weight α, mean μ and standard deviation σ. Since the weight α cannot be estimated using the moment estimation method, the parameter estimation of the Gaussian mixture distribution is generally calculated using the maximum likelihood estimation method.

[0116] (4) Multidimensional normal distribution probability density function:

[0117] The distribution of electric vehicle charging demand locations can be viewed as a probability distribution on a two-dimensional plane. When considering multidimensional data, a multidimensional normal distribution probability density function is used.

[0118]

[0119] where μ = μ n×1 =(μ1,μ2,…,μ n ) is the mean vector, ∑=∑ n×n is the covariance matrix.

[0120]

[0121] The latitude and longitude data is two-dimensional data, and the mean and covariance of the two-dimensional normal distribution are defined as follows:

[0122]

[0123] The correlation coefficient for two-dimensional data is:

[0124]

[0125] Therefore, the probability density function of the two-dimensional normal distribution is:

[0126]

[0127] in:

[0128]

[0129]

[0130] Specifically, the number of new energy electric vehicles is the determining factor of charging demand. A quantitative relationship model between the number of electric vehicles and the charging capacity of charging piles is constructed to predict the charging demand of electric vehicles under the condition of a certain number of electric vehicles. At the same time, the charging demand is simulated based on the distribution characteristics of the charging demand of electric vehicles in time and space.

[0131] 1. Select typical old residential areas and business districts to calculate regional features, and compare the similarities and differences of features in the two types of spaces. Vehicles in old residential areas travel slower, and vehicles in old residential areas generally have more parking needs at night, while business districts have more parking needs during the day. Based on this difference between different scenarios, data features are designed for different spatial scenarios. Figure 2 Schematic diagram of the number of parked vehicles in different scenarios provided by an optional embodiment of the present invention. Figure 2 As shown, by comparing the number of parked vehicles, it can be found that the mean number of parked vehicles in old residential areas and business districts is not much different, but the old residential areas have more parked vehicles mainly in the four time periods of 0:00-8:00, 8:00-12:00, 21:00-24:00, and weekdays, while the business districts have more parked vehicles mainly in the time period of 12:00-21:00.

[0132] Figure 3 is a schematic diagram of the number of vehicles traveling in different scenarios provided by an optional embodiment of the present invention, such as Figure 3 As shown in the figure, by comparing the number of vehicles, it can be found that the mean number of vehicles in old residential areas and commercial areas is not much different, but there are more vehicles in old residential areas from 12:00 to 17:00, and more vehicles in commercial areas from 0:00 to 8:00 and 21:00 to 24:00. The following locations are selected as seed grids for different scenes, and the similarity between other spaces and seed grids is analyzed.

[0133] Table 1 is a schematic table of different scene seed spaces provided by an optional implementation manner of the present invention, as shown in Table 1.

[0134] Table 1

[0135] Scene Category Place name longitude latitude Community A 116.442875 39.873178 Community B 116.370129 39.855749 Community C 116.347838 39.872085 …… …… …… …… Community D 116.293062 39.894911 SUBMARKET E 116.301514 39.834463 SUBMARKET F 116.310587 39.830303 SUBMARKET G 116.298047 39.832603 SUBMARKET H 116.309886 39.828821 SUBMARKET I 116.289261 39.832132

[0136] After calculation by the grid scene recognition algorithm of the complex coupling network algorithm, Table 2 is a schematic table of scenes recognized by the grid scene recognition results provided by an optional embodiment of the present invention, as shown in Table 2:

[0137] Table 2

[0138]

[0139]

[0140] 2. According to the probability density function of the normal distribution, the specific form of its distribution depends on two parameters: mean μ and standard deviation σ. The estimation of the parameters μ and σ of the normal distribution generally adopts the moment estimation method. Assume that the random variable X obeys the normal distribution with parameters μ and σ. Randomly draw a set of samples X1, X2, X n , use the moments of the sample to estimate the parameters of the population.

[0141] Estimation of μ, use the mean (first-order moment) to estimate μ.

[0142]

[0143] σ 2 According to the variance calculation formula DX = EX 2 -E 2 X, where E 2 X can be obtained from the estimate of μ, and EX 2 To be estimated.

[0144]

[0145] According to the calculation formula of variance DX=EX 2 -E 2 X gets σ 2 The estimated amount.

[0146]

[0147] The charging start time has multiple peaks in the interval [0,24]. The charging time in each peak neighborhood interval is a normal distribution with the peak time as the mean. Figure 4 is a curve fitting diagram at a single peak of a mixed distribution provided by an optional embodiment of the present invention, such as Figure 4 As shown, taking the peak at 8 o'clock as an example, the number of electric vehicles that start charging at 8 o'clock is the largest, and the number of electric vehicles charged around 8 o'clock drops rapidly. The distribution of the number of electric vehicles charged at 8 o'clock can be fitted with a normal distribution curve, and the distribution parameters can be estimated.

[0148] 3. The multiple peaks in the interval [0,24] can be regarded as the superposition of normal distributions with different distribution parameters. The superimposed distribution is called Gaussian mixture distribution. Figure 5 is a charging start time distribution fitting curve provided by an optional embodiment of the present invention, such as Figure 5 As shown in the figure, according to the probability density function of Gaussian mixture distribution, the specific form of its distribution depends on three parameters: weight α, mean μ and standard deviation σ. Since the weight α cannot be estimated by moment method, the parameter estimation of Gaussian mixture distribution generally adopts maximum likelihood estimation method. By observation, the charging start time is a Gaussian mixture distribution with 8 mixed components, and the distribution parameters are fitted.

[0149] Based on the above, the distribution parameters of the charging start time can be obtained. Table 1 is the distribution parameters of the charging start time provided by an optional implementation manner of the present invention, as shown in Table 3.

[0150] Table 3

[0151] Element Weight Mean variance subcomponent one 0.075212 2.2900 1.15433 subcomponent two 0.132836 15.3829 1.23064 subcomponent three 0.135883 7.6313 1.05596 subcomponent four 0.142206 18.2854 1.49412 subcomponent five 0.169875 12.6139 0.96221 Sub-component six 0.129187 4.8912 0.98560 subcomponent seven 0.129167 10.3521 1.08450 Subcomponent 8 0.085635 21.0354 2.01930

[0152] 4. Lognormal distribution is often used to describe the long-tail distribution of service time. Assuming X follows a standard normal distribution, then μ+Xσ follows with an expectation of μ and a variance of σ 2 The normal distribution of

[0153] log α Y=μ+Xσ

[0154] Then log α Y also follows the expectation μ and the variance σ 2 The normal distribution is:

[0155]

[0156] By calculation, the probability density function of x in linear coordinates can be obtained as:

[0157]

[0158] Figure 6 is a schematic diagram of the density function form of the log-normal distribution of electric vehicle charging provided by an optional embodiment of the present invention, such as Figure 6 As shown in the figure. According to the actual data distribution of charging time, most of the charging time is concentrated in the range of 0-2, and the distribution has a long tail, which basically conforms to the characteristics of negative exponential distribution. However, in the range of 0-2, the closer to 0, the higher the data concentration. There are also small data concentration parts in the long tail part, which shows that the charging time does not completely obey the negative exponential distribution.

[0159] Figure 7 is a schematic diagram of the charging time distribution curve fitting result provided by an optional embodiment of the present invention, such as Figure 7 As shown in the figure, the charging time T is a superposition of a negative exponential distribution and two normal distributions. The means of the two normal distributions are around 0.75 hours and 9 hours, respectively, indicating that there are three types of behavior patterns mixed in the charging time: (1) Charging behaviors that obey the negative exponential distribution, the longer the charging time, the fewer the number; (2) short-term charging behaviors with a mean of 0.75 hours, which are generally temporary charging behaviors; (3) long-term charging behaviors with a mean of 9 hours, which are generally used for charging at night or during working hours.

[0160] From the above analysis, it can be seen that the distribution parameters of the charging time are as shown in the following table. Table 4 is a schematic table of the distribution parameters of the charging time provided by an optional embodiment of the present invention, as shown in Table 4:

[0161] Table 2

[0162] Weight Mean variance lambda subcomponent one 0.338437 - - 0.274506 subcomponent two 0.468368 1.13157 0.17823 - subcomponent three 0.193195 6.07999 14.68428 -

[0163] 5. Figure 8 A scatter plot of the geographical coordinates of the parking locations provided in an optional embodiment of the present invention, such as Figure 8 As shown, a scatter plot of electric vehicle parking locations in geographic space, with the horizontal and vertical axes representing longitude and latitude respectively, and the color and size of the points representing the number of electric vehicles parked at the location.

[0164] Through the scatter plot of electric vehicle parking locations, we can see that the parking locations are scattered in various areas. Due to different municipal planning, some locations have densely parked vehicles, while some locations have sparsely parked vehicles, showing an uneven distribution.

[0165] It can be seen from the longitude and latitude distribution diagram of electric vehicle parking locations that electric vehicles are densely distributed in urban areas and suburbs, while electric vehicles are sparsely distributed in remote suburbs, rural areas, and mountainous areas. By estimating the longitude and latitude distribution parameters of the parking locations, Table 5 is a table of longitude and latitude distribution parameters of electric vehicle parking locations provided by an optional embodiment of the present invention, and the results are shown in the following table:

[0166] Table 3

[0167]

[0168] 6. Fig. 9 is an average charging power histogram after data cleaning provided by an optional embodiment of the present invention, such as Fig. 9 As shown, we can see that the average charging power has two peaks, the first peak is around 5kW, and the second peak is around 30kW. That is to say, the charging process can be divided into slow charging and fast charging according to the average charging power. The average charging power is a Gaussian mixture distribution in which multiple normal distributions with different distribution parameters are superimposed.

[0169] In summary, Table 6 is a schematic table of distribution parameters of average charging power provided by an optional implementation manner of the present invention. The distribution parameters of average charging power can be calculated as follows:

[0170] Table 4

[0171]

[0172]

[0173] In order to simulate the charging power curve, assuming that there are 100 electric vehicles in a certain area, a simulation is carried out through the charging start time distribution parameters, charging duration distribution parameters, and charging power distribution parameters to simulate the charging behavior of electric vehicles in the grid for 7 days.

[0174] Fig.10 This is a schematic diagram of the number of users who are charging simultaneously for an electric vehicle in a simulated 7-day period provided by an optional embodiment of the present invention. Fig.11 A schematic diagram of a simulated daily power curve of an electric vehicle provided in an optional embodiment of the present invention, Fig.12A schematic diagram of a simulated 7-day power curve of an electric vehicle provided in an optional embodiment of the present invention, such as Fig.10 , 11 As shown in Figures 1 and 12, it is the number of electric vehicles charging at the same time. The maximum number of users charging simultaneously per day is 15-20, and the average number of users charging simultaneously is 7. The number of users charging simultaneously from 21:00 pm to 3:00 am is relatively large, and the number of users charging simultaneously from 9:00 am to 3:00 pm is relatively small.

[0175] From the charging demand power curve, we can see that the maximum daily charging power is around 500kW and the average charging power is 220kW.

[0176] The charging demand of all grids is simulated. The following table shows the grids with larger charging demand, and gives the maximum charging power and the maximum number of users charging at the same time of these grids. Table 7 is a schematic table of simulation prediction of the maximum power and number of charging demand provided by an optional implementation manner of the present invention, as shown in Table 7.

[0177] Table 5

[0178] Grid Maximum charging power Maximum number of users charging simultaneously 61336******20511 606 21 61336******61023 303 16 61336******49631 290 13 61336******46783 486 19 61336******64191 177 8 61336******52799 363 15 …… …… ……

[0179] By simulating future development trends, the spatial and temporal distribution characteristics of charging demand are predicted. The simulation model is used to simulate the charging behavior of electric vehicles in the future and predict their distribution in space and time. Through the simulation and prediction of electric vehicle charging demand, the construction and layout of charging facilities can be planned more effectively, charging efficiency can be improved, resource utilization can be optimized, and the sustainable development of the electric vehicle industry can be promoted.

[0180] Through the above optional implementation, at least the following beneficial effects can be achieved:

[0181] (1) Complex coupling networks can comprehensively consider the impact of multiple factors on charging demand, such as the connection relationship between charging piles, the mutual influence between vehicles, and external environmental factors such as weather and traffic flow. This comprehensive analysis enables the charging demand prediction model based on complex coupling networks to more accurately describe the correlation and complexity between charging piles in the electric vehicle charging system, and reveal the spatial distribution characteristics of charging piles and the temporal dynamic characteristics of charging behavior, providing a richer information basis for the prediction model;

[0182] (2) The complex coupled network model is flexible and adaptable, and can be dynamically adjusted according to the changes in charging demand in different regions and at different times, thereby improving the accuracy and practicality of the prediction model. Accurate charging demand prediction results provide support for energy companies and charging facility operators to formulate more scientific charging strategies and plans, which helps to optimize resource allocation and charging facility layout, thereby improving resource utilization efficiency;

[0183] (3) This method comprehensively considers the impact of social factors on charging demand, such as user behavior, preferences, economic factors, etc., and provides a reference for formulating more targeted charging strategies. At the same time, the complex coupled network model has certain real-time and responsiveness, which can capture the changes in charging demand in a timely manner and adjust the prediction model in a timely manner, providing effective decision-making support for the operation and management of charging facilities.

[0184] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity, they are all described as a series of action combinations, but those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

[0185] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0186] Example 2

[0187] According to an embodiment of the present invention, a device for implementing the above-mentioned charging distribution data determination method for the charging scenario is also provided. Fig.13 is a structural block diagram of a device for determining charging distribution data for a charging scenario according to an embodiment of the present invention, such as Fig.13 As shown, the device includes: an acquisition module 1302, a first determination module 1304, an extraction module 1306, a second determination module 1308, a third determination module 1310 and a fourth determination module 1312. The device is described in detail below.

[0188] The acquisition module 1302 is used to acquire target scene data, wherein the target scene data is scene data representing the target charging scene; the first determination module 1304 is connected to the acquisition module 1302, and is used to determine the target space grid corresponding to the charging scene, wherein the target space grid includes multiple nodes, and the multiple nodes have a topological relationship; the extraction module 1306 is connected to the first determination module 1304, and is used to extract the target scene features of the target scene data, wherein the target scene features at least include: geographic location features, functional attribute features, and behavior pattern features; the second determination module 1308 is connected to the extraction module 1306, and is used to determine the feature distances between the target scene features and the corresponding features of the nodes in the target space grid; the third determination module 1310 is connected to the second determination module 1308, and is used to determine the target scene category corresponding to the target charging scene based on multiple feature distances; the fourth determination module 1312 is connected to the third determination module 1310, and is used to determine the charging distribution data corresponding to the target charging scene based on the target scene category, wherein the charging distribution data includes at least one of the following: charging time distribution data, charging demand distribution data.

[0189] It should be noted here that the above-mentioned acquisition module 1302, the first determination module 1304, the extraction module 1306, the second determination module 1308, the third determination module 1310 and the fourth determination module 1312 correspond to steps S102 to S112 in the method for determining the charging distribution data for implementing the charging scenario, and the instances and application scenarios implemented by the multiple modules are the same as the corresponding steps, but are not limited to the contents disclosed in the above-mentioned embodiment 1.

[0190] Example 3

[0191] According to another aspect of an embodiment of the present invention, there is also provided an electronic device, comprising: a processor; and a memory for storing instructions executable by the processor, wherein the processor is configured to execute instructions to implement any of the above-mentioned methods for determining charging distribution data for a charging scenario.

[0192] Example 4

[0193] According to another aspect of an embodiment of the present invention, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device can execute any of the above-mentioned methods for determining charging distribution data for a charging scenario.

[0194] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0195] In the above embodiments of the present invention, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0196] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of units or modules, which can be electrical or other forms.

[0197] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0198] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0199] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program codes.

[0200] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for determining charging distribution data of a charging scenario, characterized in that: include: Acquiring target scene data, wherein the target scene data is scene data representing a target charging scene; Determine a target space grid corresponding to the charging scenario, wherein the target space grid includes a plurality of nodes, and the plurality of nodes have a topological relationship; Extracting target scene features of the target scene data, wherein the target scene features at least include: geographic location features, functional attribute features, and behavior pattern features; Determining feature distances between the target scene features and corresponding features of nodes in the target spatial grid; Determining a target scene category corresponding to the target charging scene based on a plurality of feature distances; According to the target scenario category, charging distribution data corresponding to the target charging scenario is determined, wherein the charging distribution data includes at least one of the following: charging time distribution data, charging demand distribution data.

2. The method according to claim 1, characterized in that Before determining the target space grid corresponding to the charging scenario, the method further includes: Select multiple typical scenes as seed nodes to construct the initial spatial grid; Determining vector space distances between features of the sample scene and features of the plurality of typical scenes respectively; Classifying the sample scenes into sample scene categories according to a plurality of vector space distances; According to the sample scene category, the sample scene is added as a new node to the initial spatial grid to obtain a target spatial grid.

3. The method according to claim 2, characterized in that According to the sample scene category, the sample scene is added as a new node to the initial spatial grid to obtain a target spatial grid, including at least one of the following: In the case where the sample scene category is a target typical scene, adding the sample scene as a new node to the initial spatial grid, connecting the sample scene node with the target typical scene, and obtaining the target spatial grid; In the case that the sample scene category is not a target typical scene, the sample scene is added to the initial spatial grid as a new node, and the sample scene node is used as a seed node to obtain the target spatial grid.

4. The method according to claim 1, characterized in that: Before determining the charging distribution data corresponding to the target charging scenario according to the target scenario category, the method further includes: Determining a coupling coordination degree corresponding to the target spatial grid; Determining, according to the coupling coordination degree, the overall autocorrelation and the local charging scene correlation corresponding to the target spatial grid; Charging distribution data corresponding to a plurality of scene categories is determined according to the overall autocorrelation and the local charging scene correlation.

5. The method according to claim 4, characterized in that Determining charging distribution data corresponding to a plurality of scene categories according to the overall autocorrelation and the local charging scene correlation includes: Determine a distribution function corresponding to the target space grid, wherein the distribution function includes at least one of the following: a Gaussian mixture distribution function, a multidimensional normal distribution probability density function; Based on the distribution function, the overall autocorrelation and the local charging scene correlation, charging distribution data corresponding to the multiple scene categories are determined.

6. The method according to claim 4, characterized in that Determining charging distribution data corresponding to the target charging scenario according to the target scenario category includes: Determining charging distribution data corresponding to the target scenario category; Determining distinguishing features between the target charging scenario and the target scenario category; The charging distribution data is adjusted according to the distinguishing features to obtain charging distribution data corresponding to the target charging scenario.

7. The method according to claim 2, characterized in that According to the sample scene category, the sample scene is added as a new node to the initial spatial grid to obtain a target spatial grid, including: According to the sample scene category, adding the sample scene as a new node to the initial spatial grid to obtain an intermediate spatial grid; Clustering is performed on multiple nodes in the intermediate space grid to obtain the target space grid.

8. A device for determining charging distribution data of a charging scenario, characterized in that: include: An acquisition module, used to acquire target scene data, wherein the target scene data is scene data representing a target charging scene; A first determination module is used to determine a target space grid corresponding to the charging scenario, wherein the target space grid includes a plurality of nodes, and the plurality of nodes have a topological relationship; An extraction module, used for extracting target scene features of the target scene data, wherein the target scene features at least include: geographic location features, functional attribute features, and behavior pattern features; A second determination module is used to determine the feature distances between the target scene features and the corresponding features of the nodes in the target space grid; A third determination module, configured to determine a target scene category corresponding to the target charging scene based on a plurality of feature distances; The fourth determination module is used to determine the charging distribution data corresponding to the target charging scenario according to the target scenario category, wherein the charging distribution data includes at least one of the following: charging time distribution data, charging demand distribution data.

9. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the method for determining charging distribution data for a charging scenario as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the method for determining charging distribution data of a charging scenario as described in any one of claims 1 to 7.

Images