Electric vehicle charging load prediction method and system considering influence of electric power spot market
Through the prediction method based on graph attention network, combined with the K-means mean clustering method and the traffic flow equalization model, the accuracy problem of electric vehicle charging load prediction is solved, fast and accurate prediction is achieved, and the stability of the power system is improved.
Patent Information
- Application Number
- CN202510038246.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-10
- Publication Date
- 2025-04-25
Smart Images
Figure CN119886885A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of power load forecasting, and in particular relates to a method and system for forecasting electric vehicle charging load taking into account the impact of the power spot market. Background Art
[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.
[0003] According to relevant data, the proportion of electric vehicle charging is increasing rapidly, and the instability of the power system has increased significantly. In order to ensure the smooth operation of the power system, the interaction between electric vehicles and the power system needs to be closer and more reasonable. The interaction between electric vehicles and the power system refers to the transfer of electric energy between electric vehicles as distributed charging and discharging mobile devices and the power system. However, the charging behavior of electric vehicles is affected by the electricity spot market, and the spatiotemporal changes in the prices of the electricity spot market will directly affect the spatiotemporal distribution of the charging behavior of electric vehicles. To this end, an accurate and effective method for predicting the charging load of electric vehicles considering the impact of the electricity spot market needs to be proposed. Summary of the invention
[0004] To solve the above problems, the present invention proposes a method and system for predicting electric vehicle charging load considering the impact of the electricity spot market. The electric vehicle charging load is predicted based on the graph attention network, and the impact of the actual electricity spot market is considered. This can achieve rapid prediction of the electric vehicle charging load, providing a prerequisite for meeting the increasingly close and reasonable interaction between electric vehicles and power systems.
[0005] According to some embodiments, a first solution of the present invention provides a method for predicting electric vehicle charging load considering the impact of the electricity spot market, using the following technical solutions: A method for predicting electric vehicle charging load considering the impact of the electricity spot market, comprising: Obtain a database of spatiotemporal electricity price distribution in the electricity spot market; based on K-means The mean clustering method is used to extract typical spatiotemporal electricity price scenarios and their occurrence probabilities from the acquired spatiotemporal electricity price distribution database; Taking the lowest cost as the optimization goal, an electric vehicle travel model that satisfies traffic flow balance is constructed; According to the obtained typical space-time electricity price scenarios and their occurrence probabilities, combined with the constructed electric vehicle travel model, the electric vehicle charging load data under different typical space-time electricity price scenarios are obtained; Based on the graph attention network and the obtained electric vehicle charging load data, an electric vehicle charging load prediction model is constructed; Based on the constructed prediction model, the electric vehicle charging load forecast value under typical time and space electricity price scenarios is generated, and the electric vehicle charging load forecast considering the impact of the electricity spot market is completed.
[0006] As a further technical limitation, the constructed electric vehicle travel model that satisfies traffic flow equilibrium at least satisfies: (1) Fuel-powered vehicles seek the shortest route, while electric vehicles seek the route with the lowest economic cost; (2) The queuing time of electric vehicles at fast charging pile nodes increases as the number of electric vehicles increases; (3) The transit time of a road increases as the number of cars on the road increases; (4) There are fast charging stations between nodes in the transportation network, so users can queue up for fast charging or pass through directly on the bypass; (5) The initial charging load of the electric vehicle and the amount of charge each time remain unchanged.
[0007] As a further technical limitation, a typical space-time electricity price scenario probability matrix is constructed based on the obtained typical space-time electricity price scenarios and their occurrence probabilities. ;in, is in the The probability of a typical space-time electricity price scenario is calculated by Monte Carlo simulation method, that is, ;in, yes The random numbers generated by the uniform distribution probability model will Defined as 0, when the randomly generated In When the interval is within the range, it represents the Monte Carlo simulation method sampling to obtain the first Typical time-space electricity price scenarios; repeated sampling based on Monte Carlo simulation method times, get Typical space-time electricity price scenarios; based on the obtained S A typical space-time electricity price scenario is constructed, and the electric vehicle charging load data under different typical space-time electricity price scenarios are obtained by combining the constructed electric vehicle travel model.
[0008] Furthermore, based on Monte Carlo sampling An electric vehicle load sample set is formed by the electric vehicle charging load data under different typical spatiotemporal electricity price scenarios. According to the obtained sample set, an electric vehicle charging load forecasting model based on graph attention network is trained to obtain an electric vehicle load forecasting model based on model and data hybrid drive and considering the impact of electricity spot market.
[0009] Furthermore, the adjacency matrix and node feature matrix are used to characterize the topological characteristics and spatiotemporal electricity price characteristics of the transportation network. The graph attention network is used to construct the mapping relationship between the adjacency matrix, node feature matrix and electric vehicle charging load data under typical spatiotemporal electricity price scenarios. The corresponding relationship between the transportation network, spatiotemporal electricity prices and electric vehicle charging load is obtained, that is, an electric vehicle charging load prediction model considering the impact of the electricity spot market is obtained.
[0010] As a further technical limitation, in the process of extracting typical spatiotemporal electricity price scenarios and their occurrence probabilities from the acquired spatiotemporal electricity price distribution database, continuous sampling is performed in the acquired spatiotemporal electricity price distribution database with the day as the minimum sampling point. Influenced by the electricity spot market The electricity price distribution of each node is used to construct the electricity price feature matrix ; In the constructed electricity price characteristic matrix Select The characteristic matrix is used as the initial cluster center to calculate other electricity price characteristic matrices and The distance between cluster centers; according to the electricity price feature matrix The distance from the cluster center is used to calculate the electricity price feature matrix Classification, constant reselection The feature matrix is used as the cluster center until the cluster center no longer changes. K-means Mean clustering, we get Typical spatiotemporal electricity price distribution scenarios and their occurrence probabilities; among them, for K-means Mean number of cluster centers.
[0011] According to some embodiments, a second solution of the present invention provides a system for predicting electric vehicle charging load considering the impact of the electricity spot market, using the following technical solutions: A system for predicting electric vehicle charging load considering the impact of the electricity spot market, comprising: An acquisition module configured to acquire a temporal and spatial electricity price distribution database in a power spot market; The clustering module is configured based on K-means The mean clustering method is used to extract typical spatiotemporal electricity price scenarios and their occurrence probabilities from the acquired spatiotemporal electricity price distribution database; A construction module, which is configured to take the lowest cost as the optimization goal and construct an electric vehicle travel model that satisfies the traffic flow balance; A calculation module is configured to obtain electric vehicle charging load data under different typical space-time electricity price scenarios based on the obtained typical space-time electricity price scenarios and their occurrence probabilities, combined with the constructed electric vehicle travel model; The prediction module is configured to construct an electric vehicle charging load prediction model based on the graph attention network and the obtained electric vehicle charging load data; based on the constructed prediction model, generate electric vehicle charging load prediction values under typical spatiotemporal electricity price scenarios, and complete the electric vehicle charging load prediction considering the impact of the electricity spot market.
[0012] According to some embodiments, a third solution of the present invention provides a computer-readable storage medium, which adopts the following technical solution: A computer-readable storage medium stores a program, which, when executed by a processor, implements the steps in the method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in the first solution of the present invention.
[0013] According to some embodiments, a fourth solution of the present invention provides an electronic device, which adopts the following technical solution: An electronic device comprises a memory, a processor and a program stored in the memory and running on the processor, wherein when the processor executes the program, the steps in the method for predicting charging load of electric vehicles considering the impact of the electricity spot market as described in the first solution of the present invention are implemented.
[0014] According to some embodiments, a fifth solution of the present invention provides a computer program product, which adopts the following technical solution: A computer program product includes software codes, wherein the program in the software codes executes the steps in the method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in the first embodiment of the present invention.
[0015] Compared with the prior art, the present invention has the following beneficial effects: The present invention is based on the traffic flow equilibrium model and the graph attention network to realize the prediction of electric vehicle charging load based on the hybrid drive of model and data and considering the influence of the electricity spot market; K-means The mean clustering method extracts typical spatiotemporal electricity price scenarios from the historical spatiotemporal electricity price database, greatly reducing the spatiotemporal state of electricity prices and significantly reducing the amount of calculation; based on the traffic flow equilibrium model and typical spatiotemporal electricity price scenarios, an electric vehicle behavior model that considers the impact of the electricity spot market on spatiotemporal electricity prices is established, and the electricity spot market is included in the considerations of electric vehicle behavior, making the constructed electric vehicle behavior model more in line with actual needs; the adjacency matrix and node feature matrix are used to represent the topological characteristics and spatiotemporal electricity price characteristics of the transportation network, and the graph attention network is used to construct the mapping relationship between the adjacency matrix, the node feature matrix and the electric vehicle charging load, thereby indirectly obtaining the relationship between the transportation network, spatiotemporal electricity prices and the electric vehicle charging load, and establishing a charging load prediction model for electric vehicles that considers the impact of the electricity spot market, which greatly improves the efficiency of electric vehicle charging load prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] The drawings in the specification that constitute a part of this embodiment are used to provide a further understanding of this embodiment. The schematic embodiments of this embodiment and their descriptions are used to explain this embodiment and do not constitute improper limitations on this embodiment.
[0017] Figure 1 This is a flow chart of a method for predicting electric vehicle charging load considering the impact of the electricity spot market in Embodiment 1 of the present invention; Figure 2 Schematic diagram of the regression prediction model structure of the graph attention network in the first embodiment of the present invention; Figure 3 This is a structural block diagram of an electric vehicle charging load forecasting system considering the impact of the electricity spot market in Embodiment 2 of the present invention. DETAILED DESCRIPTION
[0018] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.
[0019] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.
[0020] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.
[0021] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.
[0022] Embodiment 1 Embodiment 1 of the present invention introduces a method for predicting charging load of electric vehicles taking into account the impact of the electricity spot market.
[0023] like Figure 1 A method for predicting electric vehicle charging load considering the impact of the electricity spot market is shown, comprising the following steps: Step S01: Based on the spatiotemporal distribution data of electricity prices in the existing electricity spot market, a database including spatiotemporal distribution characteristics of electricity prices is established; Step S02: Use k-means The mean clustering method extracts typical spatiotemporal electricity price scenarios and the probability of each scenario occurring; Step S03: constructing an electric vehicle travel model based on the traffic flow equilibrium model; Step S04: using Monte Carlo simulation method to sample and generate massive electric vehicle charging load data under different typical time and space electricity price scenarios; Step S05: Based on the electric vehicle charging load data, combined with the graph attention network, an electric vehicle load forecasting model based on a hybrid drive of model and data and considering the impact of the electricity spot market is obtained; Step S06: Based on the electric vehicle charging load prediction model, the electric vehicle charging load considering the impact of the electricity spot market is obtained.
[0024] As one or more implementation modes, in step S02, the day is used as the minimum sampling point, and the time-space electricity price database is continuously sampled to obtain Influenced by the electricity spot market The node electricity price distribution is recorded as a matrix ;use k-means Mean clustering method matrix Perform classification and feature extraction to obtain Typical electricity price distribution scenarios and the probability matrix of each scenario Specific: 1) Based on the input Influenced by the electricity spot market The electricity price distribution of each node is used to construct the electricity price feature matrix ; The definition of is: ; in, It is Nodes in the subsampled data The daily price matrix.
[0025] The definition of is: ; in, It is Nodes in the subsampled data No. Hourly electricity price.
[0026] 2) Set up the extraction required A typical electricity price distribution scenario; yes k-means Mean number of cluster centers; 3) Select in the electricity price feature matrix feature matrix as the initial cluster center; The cluster centers are denoted as ; 4) Calculate the remaining electricity price characteristic matrix and The distance between cluster centers; The distance calculation formula is: ; in, It is Cluster Centers The elements.
[0027] 5) Classification is performed based on the distance between the electricity price feature matrix and the cluster center. The category of the electricity price feature matrix is the category of the cluster center with the smallest distance from it. categorization; 6) Return to step 2) and select The electricity price characteristic matrix is k-means The cluster center of mean clustering is denoted as When the cluster center When there is no change, k-means Mean clustering ends; 7) Record the final result A typical electricity price characteristic matrix ; 8) Calculation Probability matrix of typical space-time electricity price scenarios ,Right now: ; in, It is The number of scenes in a cluster, Yes Input k-means The total number of scenes for mean clustering, is in the The probability of a typical space-time electricity price scenario.
[0028] As one or more implementation modes, in step S03, the electric vehicle travel model based on the traffic flow equilibrium model is as follows: wherein, fuel vehicles and electric vehicles travel should meet the following assumptions: 1) Fuel vehicles look for the shortest route, and electric vehicles look for the route with the lowest economic cost; 2) The queuing time of electric vehicles at fast charging pile nodes increases as the number of electric vehicles increases; 3) The transit time of a road increases as the number of cars on the road increases; 4) There are fast charging stations between nodes in the transportation network, so users can queue up for fast charging or pass directly through the bypass; 5) The initial charging load of the electric vehicle and the amount of charging each time remain unchanged.
[0029] The traffic network in this embodiment can be represented by a connectivity graph Indicates; among them, is the set of nodes in the transportation network; is a collection of roads connecting nodes.
[0030] The network topology can be used The matrix Representation; Matrix Each column of represents a road and has only two non-zero numbers, where 1 represents the starting point of the road and -1 represents the end point of the road.
[0031] Define the starting node Departure to destination node The number of electric vehicles ending is defined as the number of vehicles from the starting node Departure to destination node The electric vehicle traffic flow that ends is recorded as ; From the starting node Departure to destination node The final electric vehicle path (i.e., road set) is denoted as Similarly, the fuel car starts from the starting node Departure to destination node The end of the traffic flow is recorded as , the path set is denoted as .
[0032] Slave Node Departure to the node End of electric vehicle path and road The correlation matrix The elements in , the incidence matrix Each column represents the topological structure of a path. Represents the path of an electric vehicle With Road The relationship between When, the path Passing Road ,when When , the path does not pass through the road. Similarly, the association matrix between the fuel car path and the road is recorded as , the elements in the incidence matrix are recorded as .
[0033] The change in road travel time caused by traffic congestion is related to the size of traffic flow.
[0034] If there is only one road between two nodes in the traffic network and there is no fast charging station, the change in the travel time of the road can be expressed as the delayed travel time. express. for: ; in, It's the road Traffic flow on , It is a road without fast charging stations; It is free travel time not affected by traffic congestion; It's the road The capacity limit parameter.
[0035] Delayed travel time on a road with a fast charging station between two nodes in the traffic network for: ; in, is a road with fast charging stations; is the control parameter.
[0036] If there is a fast charging station between two nodes of the traffic network, there can also be a bypass. The travel time of the bypass is not affected by traffic congestion. It can be expressed as follows: ; in, , It is a bypass road.
[0037] The traffic flow of the traffic network is: ; ; ; in, When the starting node is , the destination node is When the fuel car is on the path Traffic flow on the When the starting node is , the destination node is When the electric vehicle is on the path Traffic flow on the .
[0038] When the departure node is , the destination node is When the fuel car passes through the path Time and costs for: ; ; in, is the discounted cost of travel time.
[0039] When the departure node is , the destination node is When the fuel car passes through the path Time and costs for: ; ; in, It's the road Connect the distribution network nodes The electricity price of fast charging stations; is the charge capacity of the electric car.
[0040] In this embodiment, Slave node electricity price matrix Select, every hour, you need to update to , and recalculate the traffic flow equilibrium model.
[0041] The prerequisite for achieving traffic flow balance is that the costs on all paths are the same, and the following constraints should be met: ; ; ; ; in, and Fuel vehicles and electric vehicles The minimum cost between two points.
[0042] The objective function and constraints of the traffic flow equilibrium model are as follows: Objective function: ; Constraints: ; ; ; Among them, the objective function is the minimum value of the discounted cost.
[0043] As one or more implementation methods, in step S04, a Monte Carlo simulation method is used to sample, based on typical spatiotemporal electricity price scenarios and their probabilities and an electric vehicle travel model considering traffic flow equilibrium, to generate a large amount of electric vehicle charging load data under different typical spatiotemporal electricity price scenarios, denoted as data; specific: According to the obtained typical space-time electricity price scenarios and their occurrence probabilities, a typical space-time electricity price scenario probability matrix is constructed. ;in, is in the The probability of a typical space-time electricity price scenario is calculated by Monte Carlo simulation method, that is, ;in, yes The random numbers generated by the uniform distribution probability model will Defined as 0, when the randomly generated In When the interval is within the range, it represents the Monte Carlo simulation method sampling to obtain the first Typical time-space electricity price scenarios; repeated sampling based on Monte Carlo simulation method times, get Typical space-time electricity price scenarios; based on the obtained S A typical space-time electricity price scenario is constructed, and the electric vehicle charging load data under different typical space-time electricity price scenarios are obtained by combining the constructed electric vehicle travel model.
[0044] As one or more implementation modes, in step S05, based on the Monte Carlo sampling generated The electric vehicle load sample set formed by the electric vehicle charging load data under different typical spatiotemporal electricity price scenarios is used to train the electric vehicle charging load forecasting model based on the graph attention network, and an electric vehicle load forecasting model based on model and data hybrid drive considering the impact of the electricity spot market is obtained; specifically: The input of the prediction model is the adjacency matrix of the traffic network and the node feature matrix ; That is, the adjacency matrix of the transportation network for: ; in, is the adjacency matrix No. Row, No. Elements of the column, representing nodes and nodes connection status.
[0045] Node feature matrix of transportation network for: ; in, is the node feature matrix No. List, Is with the node A collection of connected roads. It's the road The electricity price of fast charging piles.
[0046] The output of the prediction model is the charging load of electric vehicles ,Right now: ; in, It is Charging load of electric vehicles.
[0047] The regression prediction model of the graph attention network used in this embodiment is as follows Figure 2 As shown in the figure, it consists of three layers of graph attention layers and one layer of average pooling layer; specifically: The graph attention layer includes the graph attention coefficient The calculation and attention layer output are generated in two parts: ; in, K is the number of heads in the graph attention layer, and the parameters of each head are independent of each other; It is l Node feature matrix of layer graph attention input; It is k The linear transformation weight matrix of the head is used to change the node feature dimension; It is k The result of linear change of head weight.
[0048] The following formula is used to calculate the k Header Node n and nodes m The graph attention coefficient between , represents the strength of the connection between nodes; that is, ; ; in, and yes No. n Row and m row, indicating the k Header Node n and nodes m The symbol “||” indicates the gluing of tensors; It is kThe weight matrix of the head forms a single-layer neural network and transforms the output result into a scalar with a dimension of 1×1; is the activation function; is a custom masking function that aims to eliminate the graph attention coefficients between nodes that are not directly connected to each other; is the activation function; Is a node n Including its own adjacent nodes.
[0049] The calculation rules of the symbol “||” are: ; The activation function is: ; Mask function for: ; in, Represents negative infinity.
[0050] The activation function is: ; The attention layer output is generated as: ; in, It is k Header Node The output result of the feature matrix is, It is k The attention coefficient matrix of each head.
[0051] The attention layer has different outputs in different layers of the prediction model.
[0052] In the first and second graph attention layers, the output K The tensor glue of the output results of each head, that is: ; In the third graph attention layer, the output K The purpose of this is to extract the average feature, that is: ; The activation function is: ; The activation function maps the normalized output to Between, specifically: ; Sum function The output matrices of the first and second layers are added together as the input of the third layer in order to speed up the training of the neural network, namely: ; The purpose of the global average pooling layer is to reduce the dimension of the output layer and extract global feature information, specifically: ; in, is the average feature matrix No. OK. , is the number of nodes in the transportation network.
[0053] The error function of the model uses mean square error, which is: ; in, is the actual value of the electric vehicle charging load, is the predicted value of electric vehicle charging load.
[0054] As one or more implementation methods, in the process of predicting the charging load of the electric vehicle in step S06, based on the Monte Carlo simulation method considering the probability of the typical time-space electricity price scenario, a random sampling is performed once to obtain a typical time-space electricity price scenario, which is recorded as ; The expanded form is: ; in, It is The node in Hourly electricity price.
[0055] Based on the traffic network model and typical space-time electricity price scenarios , we can get Adjacency matrix for hours and the node feature matrix ; Based on the electric vehicle charging load prediction model, the electric vehicle charging load under typical time and space electricity price scenarios is generated, specifically: ; in, It is an electric vehicle charging load prediction model, which reflects the mapping relationship between the spatiotemporal distribution of electric vehicle charging load and the spatiotemporal electricity price distribution scenario; Electric cars are the hours of charging load.
[0056] This embodiment is based on the traffic flow equilibrium model and the graph attention network to realize the prediction of electric vehicle charging load based on the hybrid drive of model and data and considering the influence of the electricity spot market; K-means The mean clustering method extracts typical spatiotemporal electricity price scenarios from the historical spatiotemporal electricity price database, greatly reducing the spatiotemporal state of electricity prices and significantly reducing the amount of calculation; based on the traffic flow equilibrium model and typical spatiotemporal electricity price scenarios, an electric vehicle behavior model that considers the impact of the electricity spot market on spatiotemporal electricity prices is established, and the electricity spot market is included in the considerations of electric vehicle behavior, making the constructed electric vehicle behavior model more in line with actual needs; the adjacency matrix and node feature matrix are used to represent the topological characteristics and spatiotemporal electricity price characteristics of the transportation network, and the graph attention network is used to construct the mapping relationship between the adjacency matrix, the node feature matrix and the electric vehicle charging load, thereby indirectly obtaining the relationship between the transportation network, spatiotemporal electricity prices and the electric vehicle charging load, and establishing a charging load prediction model for electric vehicles that considers the impact of the electricity spot market, which greatly improves the efficiency of electric vehicle charging load prediction.
[0057] Embodiment 2 Embodiment 2 of the present invention introduces a system for predicting charging load of electric vehicles taking into account the impact of the electricity spot market.
[0058] like Figure 3 The electric vehicle charging load forecasting system considering the impact of the electricity spot market is shown, comprising: An acquisition module configured to acquire a temporal and spatial electricity price distribution database in a power spot market; The clustering module is configured based on K-means The mean clustering method is used to extract typical spatiotemporal electricity price scenarios and their occurrence probabilities from the acquired spatiotemporal electricity price distribution database; A construction module, which is configured to take the lowest cost as the optimization goal and construct an electric vehicle travel model that satisfies the traffic flow balance; A calculation module is configured to obtain electric vehicle charging load data under different typical space-time electricity price scenarios based on the obtained typical space-time electricity price scenarios and their occurrence probabilities, combined with the constructed electric vehicle travel model; The prediction module is configured to construct an electric vehicle charging load prediction model based on the graph attention network and the obtained electric vehicle charging load data; based on the constructed prediction model, generate electric vehicle charging load prediction values under typical spatiotemporal electricity price scenarios, and complete the electric vehicle charging load prediction considering the impact of the electricity spot market.
[0059] The detailed steps are the same as those of the method for predicting electric vehicle charging load considering the impact of the electricity spot market provided in Example 1, and will not be repeated here.
[0060] Embodiment 3 Embodiment 3 of the present invention provides a computer-readable storage medium.
[0061] A computer-readable storage medium stores a program thereon, which, when executed by a processor, implements the steps in the method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in the first embodiment of the present invention.
[0062] The detailed steps are the same as those of the method for predicting electric vehicle charging load considering the impact of the electricity spot market provided in Example 1, and will not be repeated here.
[0063] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The solutions in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and interpreted scripting language JavaScript, etc.
[0064] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0065] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.
[0066] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.
[0067] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.
[0068] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
[0069] Embodiment 4 A fourth embodiment of the present invention provides an electronic device.
[0070] An electronic device comprises a memory, a processor and a program stored in the memory and running on the processor, wherein when the processor executes the program, the steps in the method for predicting charging load of electric vehicles considering the impact of the electricity spot market as described in the first embodiment of the present invention are implemented.
[0071] The detailed steps are the same as those of the method for predicting electric vehicle charging load considering the impact of the electricity spot market provided in Example 1, and will not be repeated here.
[0072] Embodiment 5 Embodiment 5 of the present invention provides a computer program product.
[0073] A computer program product includes software codes, wherein the program in the software codes executes the steps in the method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in the first embodiment of the present invention.
[0074] The detailed steps are the same as those of the method for predicting electric vehicle charging load considering the impact of the electricity spot market provided in Example 1, and will not be repeated here.
[0075] The above description is only a preferred embodiment of the present embodiment and is not intended to limit the present embodiment. For those skilled in the art, the present embodiment may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present embodiment shall be included in the protection scope of the present embodiment.
Claims
1. A method for predicting electric vehicle charging load considering the impact of the electricity spot market, characterized in that: include: Obtain a database of spatiotemporal electricity price distribution in the electricity spot market; based on K-means The mean clustering method is used to extract typical spatiotemporal electricity price scenarios and their occurrence probabilities from the acquired spatiotemporal electricity price distribution database; Taking the lowest cost as the optimization goal, an electric vehicle travel model that satisfies traffic flow balance is constructed; According to the obtained typical space-time electricity price scenarios and their occurrence probabilities, combined with the constructed electric vehicle travel model, the electric vehicle charging load data under different typical space-time electricity price scenarios are obtained; Based on the graph attention network and the obtained electric vehicle charging load data, an electric vehicle charging load prediction model is constructed; Based on the constructed prediction model, the electric vehicle charging load forecast value under typical time and space electricity price scenarios is generated, and the electric vehicle charging load forecast considering the impact of the electricity spot market is completed.
2. A method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in claim 1, characterized in that: The constructed electric vehicle travel model that meets traffic flow equilibrium at least meets the following requirements: (1) Fuel-powered vehicles seek the shortest route, while electric vehicles seek the route with the lowest economic cost; (2) The queuing time of electric vehicles at fast charging pile nodes increases as the number of electric vehicles increases; (3) The transit time of a road increases as the number of cars on the road increases; (4) There are fast charging stations between nodes in the transportation network, so users can queue up for fast charging or pass through directly on the bypass; (5) The initial charging load of the electric vehicle and the amount of charge each time remain unchanged.
3. A method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in claim 1, characterized in that: According to the obtained typical space-time electricity price scenarios and their occurrence probabilities, a typical space-time electricity price scenario probability matrix is constructed. ;in, is in the The probability of a typical space-time electricity price scenario is calculated by Monte Carlo simulation method, that is, ;in, yes The random numbers generated by the uniform distribution probability model will Defined as 0, when the randomly generated In When the interval is within the range, it represents the Monte Carlo simulation method sampling to obtain the first Typical time-space electricity price scenarios; repeated sampling based on Monte Carlo simulation method times, get Typical space-time electricity price scenarios; based on the obtained S typical space-time electricity price scenarios and the constructed electric vehicle travel model, the electric vehicle charging load data under different typical space-time electricity price scenarios are obtained.
4. A method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in claim 3, characterized in that: Generated based on Monte Carlo sampling An electric vehicle load sample set is formed by the electric vehicle charging load data under different typical spatiotemporal electricity price scenarios. According to the obtained sample set, an electric vehicle charging load forecasting model based on graph attention network is trained to obtain an electric vehicle load forecasting model based on model and data hybrid drive and considering the impact of electricity spot market.
5. A method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in claim 4, characterized in that: The adjacency matrix and node feature matrix are used to represent the topological characteristics and spatiotemporal electricity price characteristics of the transportation network. The graph attention network is used to construct the mapping relationship between the adjacency matrix, node feature matrix and electric vehicle charging load data under typical spatiotemporal electricity price scenarios. The corresponding relationship between the transportation network, spatiotemporal electricity prices and electric vehicle charging load is obtained, that is, an electric vehicle charging load prediction model considering the impact of the electricity spot market is obtained.
6. A method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in claim 1, characterized in that: In the process of extracting typical spatiotemporal electricity price scenarios and their occurrence probabilities from the acquired spatiotemporal electricity price distribution database, continuous sampling is performed in the acquired spatiotemporal electricity price distribution database with the day as the minimum sampling point. Influenced by the electricity spot market The electricity price distribution of each node is used to construct the electricity price feature matrix ; In the constructed electricity price characteristic matrix Select The characteristic matrix is used as the initial cluster center to calculate other electricity price characteristic matrices and The distance between cluster centers; according to the electricity price feature matrix The distance from the cluster center is used to calculate the electricity price feature matrix Classification, constant reselection The feature matrix is used as the cluster center until the cluster center no longer changes. K- means Mean clustering, we get Typical spatiotemporal electricity price distribution scenarios and their occurrence probabilities; among them, for K-means Mean number of cluster centers.
7. A system for predicting electric vehicle charging load considering the impact of the electricity spot market, characterized in that: include: An acquisition module configured to acquire a temporal and spatial electricity price distribution database in a power spot market; The clustering module is configured based on K-means The mean clustering method is used to extract typical spatiotemporal electricity price scenarios and their occurrence probabilities from the acquired spatiotemporal electricity price distribution database; A construction module, which is configured to construct an electric vehicle travel model that satisfies traffic flow equilibrium by taking the lowest cost as an optimization goal; A calculation module is configured to obtain electric vehicle charging load data under different typical space-time electricity price scenarios based on the obtained typical space-time electricity price scenarios and their occurrence probabilities, combined with the constructed electric vehicle travel model; The prediction module is configured to construct an electric vehicle charging load prediction model based on the graph attention network and the obtained electric vehicle charging load data; based on the constructed prediction model, generate electric vehicle charging load prediction values under typical spatiotemporal electricity price scenarios, and complete the electric vehicle charging load prediction considering the impact of the electricity spot market.
8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in any one of claims 1-6 are implemented.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the program, the steps of the method for predicting electric vehicle charging load considering the impact of the electricity spot market are implemented as described in any one of claims 1-6.
10. A computer program product comprising software code, characterized in that The program in the software code executes the steps of the method for predicting electric vehicle charging load considering the impact of the electricity spot market as described in any one of claims 1-6.