Multi-gun low-power charging pile system and charging method

Abstract

The present invention relates to the technical field of charging piles, and proposes a multi-gun low-power charging pile system and method. The system includes: a power grid load prediction module, which predicts power grid load prediction data for a target period; a charging pile status classification module, which classifies the status of the charging piles and obtains the charging pile status of each group of charging piles; a user demand acquisition module, which obtains the user's charging gun usage requirements; and a charging strategy generation module, which generates a charging strategy for controlling the charging guns of each group of charging piles to perform charging actions based on the power grid load prediction data, the charging pile status of each group of charging piles, and the user's charging gun usage requirements. By generating a charging strategy for multiple charging guns for each group of charging piles, the present invention provides more charging locations to solve the problem of imbalanced supply and demand of charging piles, while taking into account the impact of peak hours on the power grid, and generates a charging strategy for each group of charging piles that adapts to the user's charging needs, thereby solving the problem of unbalanced power grid load.

Description

Technical Field

[0001] The present invention relates to the technical field of charging piles, and in particular to a multi-gun low-power charging pile system and method. Background Art

[0002] Electric vehicles are gaining widespread attention and promotion as an environmentally friendly and efficient means of transportation. However, as electric vehicles become more prevalent and their number of users increases, charging infrastructure presents some real challenges and shortcomings. During peak demand periods, queues often occur at charging stations. This results in users spending more time waiting for charging, which is not only inconvenient but can also prolong the charging process and waste time. Some users of charging stations will continue to occupy the station after charging is complete without moving their vehicles, limiting access for other users. Even in some areas where charging stations are available, their number is often insufficient to meet user demand. This can lead to difficulties for users in charging while traveling, particularly in urban and residential areas.

[0003] As the number of electric vehicles increases, an imbalance in the supply and demand of charging stations is gradually emerging. In some regions, congestion and scarcity of charging stations have become commonplace, impacting the user charging experience and the promotion of electric vehicles. Large-scale electric vehicle charging can place enormous loads on the power grid during peak hours, leading to grid instability and power supply issues. Grid load balancing remains a key issue that needs to be addressed. Improving charging efficiency and speed is a key goal. Currently, some electric vehicles still require a relatively long charging time, which limits their use and promotion. Furthermore, private charging involves significant time waste, with charging time far less than parking time or the time the charging station is occupied.

[0004] To sum up, how to solve the imbalance between supply and demand of charging piles and the imbalance of power grid load during peak hours is an urgent problem that needs to be solved. Summary of the Invention

[0005] In order to solve the above-mentioned problems in the prior art, the present invention provides a multi-gun low-power charging pile system and method, which aims to solve the problems of imbalance in supply and demand of charging piles and imbalance in power grid load during peak hours in the prior art.

[0006] A first aspect of the present invention provides a multi-gun low-power charging pile system, comprising:

[0007] A power grid load forecasting module, configured to obtain historical power grid load data within a region and forecast power grid load forecast data for a target period based on the historical power grid load data;

[0008] A charging pile status classification module, wherein the charging pile status determination module is configured to obtain historical status data of each group of charging piles, and classify the charging piles according to the historical status data to obtain the charging pile status of each group of charging piles;

[0009] A user demand acquisition module, configured to acquire the user's demand for using a charging gun;

[0010] A charging strategy generation module is configured to generate a charging strategy for controlling the charging guns of each group of charging piles to perform charging actions based on the grid load forecast data, the charging pile status of each group of charging piles, and the user's charging gun usage requirements.

[0011] Optionally, the power grid load forecasting module specifically includes:

[0012] A prediction model establishment unit, wherein the prediction model establishment unit is configured to extract historical time window data from historical power grid load data, and input the historical time window data into an initial model for training to obtain a trained power grid load prediction model;

[0013] A power grid load prediction unit is configured to input future time window data into the power grid load prediction model to obtain power grid load prediction data output by the power grid load prediction model.

[0014] Optionally, the prediction model building unit specifically includes:

[0015] a historical time window data extraction subunit, wherein the historical time window data extraction subunit is configured to divide the historical power grid load data into time windows of fixed length, and regard the power grid load value in each time window as the pixel value of the image to form the historical time window data;

[0016] The model training subunit is configured to input each data point in the historical time window data into the CNN model for training to obtain a trained power grid load forecasting model.

[0017] Optionally, in the model training subunit, the historical time window data includes N input data points X at time steps i and feature dimensions d. i,d,1 ; The CNN model includes:

[0018] Convolutional layer, the expression of the convolutional layer is:

[0019]

[0020] Among them, F is the convolution kernel size, Z (1) is the output of the convolutional layer, W(1) is the weight of the convolution kernel, k is the index of the convolution kernel;

[0021] Pooling layer, the expression of the pooling layer is:

[0022]

[0023]

[0024] Among them, P is the pooling window size, is the activation function ReLU;

[0025] Fully connected layer, the expression of the fully connected layer is:

[0026]

[0027] Among them, Z (2) is the output of the fully connected layer, W (2) is the weight of the fully connected layer, h is the index of the neuron;

[0028] Output layer, the expression of the output layer is:

[0029]

[0030] in, Represents the predicted value of the power grid load at time step i.

[0031] Optionally, the charging pile status classification module specifically includes:

[0032] a historical status data processing unit configured to obtain historical status data of each group of charging piles, extract status features of the historical status data, and calculate the confidence of the charging piles based on the status features;

[0033] The charging pile state classification unit is configured to calculate the confidence of the charging pile and the fuzzy membership degree of the threshold value using the fuzzy C-means method, and classify the state of each group of charging piles according to the fuzzy membership degree to obtain the charging pile state of each group of charging piles.

[0034] Optionally, the charging pile status classification unit specifically includes:

[0035] The classification definition subunit is configured to define a feature vector of a data point as X=(x1, x2, ..., x N ), define the eigenvector of the cluster center as V=(v1,v2,…,v C ); where N is the number of data points, C is the number of cluster centers, and each cluster center represents the status of a charging pile;

[0036] The fuzzy membership calculation subunit is configured to calculate the fuzzy membership matrix U: [U=[u ij ] N×C ], using the Euclidean distance d ij :[d ij =||X i -V j || 2 ], calculate the fuzzy membership value Among them, X i is the eigenvector of the i-th data point, V j is the eigenvector of the jth cluster center, m is the fuzziness parameter, which is usually an integer greater than 1, and t represents the number of iterations;

[0037] The charging pile state classification subunit is configured to take the cluster center with the largest fuzzy membership value as the charging pile state of the charging pile.

[0038] Optionally, the system further includes: a charging pile status update module, wherein the charging pile status update module specifically includes:

[0039] A charging pile real-time status acquisition unit, wherein the charging pile real-time status acquisition module is configured to obtain the real-time status of each group of charging piles;

[0040] A cluster center updating unit is configured to update the fuzzy membership value according to the real-time status of the charging pile, and update the cluster center according to the updated fuzzy membership value; wherein the cluster center update expression is specifically:

[0041]

[0042] Among them, V j represents the new location of the jth cluster center, u ij is the fuzzy membership value of data point i to cluster center j, X i is the eigenvector of data point i, and m is the fuzziness parameter;

[0043] A charging pile state updating unit is configured to update the charging pile state of the charging pile in the target time period according to the updated cluster center.

[0044] Optionally, the charging gun usage requirements include a charging time period requirement and a charging power requirement of the charging gun.

[0045] Optionally, the charging strategy generation module specifically includes:

[0046] A charging power distribution unit for the charging piles, the charging power distribution unit for the charging piles being configured to distribute corresponding charging power to each group of charging piles based on grid load forecast data and the status of the charging piles of each group of charging piles, so as to balance the grid load;

[0047] A charging strategy generation unit for the charging gun is configured to generate a charging strategy for executing charging actions for each charging gun in each group of charging piles based on the charging power allocated to each group of charging piles and the charging period and charging power requirements of the user's charging gun, so that each user's charging gun usage requirements match the corresponding charging gun.

[0048] The second aspect of the present invention provides a multi-gun low-power charging pile method, comprising:

[0049] Obtaining historical grid load data within a region, and predicting grid load forecast data for a target period based on the historical grid load data;

[0050] Obtaining historical status data of each group of charging piles, and classifying the charging piles according to the historical status data to obtain the charging pile status of each group of charging piles;

[0051] Obtain the user's demand for charging guns;

[0052] A charging strategy for controlling the charging guns of each group of charging piles to perform charging actions is generated according to the grid load forecast data, the charging pile status of each group of charging piles, and the user's charging gun usage requirements.

[0053] The beneficial effects of the present invention are: a multi-gun low-power charging pile system and method are proposed, which generates charging strategies for multiple charging guns in each group of charging piles by predicting grid load data and utilizing the status of the charging piles and the user's charging gun usage needs. While providing more charging locations to solve the problem of imbalance in supply and demand of charging piles, it can also take into account the load imbalance impact on the power grid caused by peak hours, and generate charging strategies for each group of charging piles that adapt to the user's charging needs, thereby solving the problem of grid load imbalance. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a structural diagram of the multi-gun low-power charging pile system provided by the present invention;

[0055] Figure 2 This is a flow chart of the multi-gun low-power charging pile method provided by the present invention.

[0056] Reference numerals:

[0057] 10-Grid load forecasting module; 20-Charging pile status classification module; 30-User demand acquisition module; 40-Charging strategy generation module. DETAILED DESCRIPTION

[0058] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0059] Example 1:

[0060] Reference Figure 1 , Figure 1 A schematic structural diagram of a multi-gun low-power charging pile system provided in an embodiment of the present invention.

[0061] like Figure 1 As shown, a multi-gun low-power charging pile system includes: a power grid load forecasting module 10, the power grid load forecasting module 10 is configured to obtain historical power grid load data within a regional range, and predict the power grid load forecast data of the target period based on the historical power grid load data; a charging pile state classification module 20, the charging pile state determination module is configured to obtain historical state data of each group of charging piles, and classify the charging piles according to the historical state data to obtain the charging pile state of each group of charging piles; a user demand acquisition module 30, the user demand acquisition module 30 is configured to obtain the user's charging gun usage requirements; a charging strategy generation module 40, the charging strategy generation module 40 is configured to generate a charging strategy for controlling the charging guns of each group of charging piles to perform charging actions based on the power grid load forecasting data, the charging pile state of each group of charging piles and the user's charging gun usage requirements.

[0062] It should be noted that due to the continuous increase in the number of electric vehicles, the supply and demand imbalance of charging piles has gradually emerged. As a result, in certain areas, especially urban and residential areas, there will be long waiting times for charging piles, which affects the user experience of electric vehicles. In some cases, charging pile users will continue to occupy the charging piles after charging is completed without removing their vehicles, which limits the use of other users. At the same time, large-scale electric vehicle charging may also place huge load pressure on the power grid during peak hours, leading to power grid instability and power supply problems. To address this issue, this embodiment proposes a multi-gun low-power charging pile system. By predicting power grid load data and utilizing the status of charging piles and the user's charging gun usage needs, a charging strategy for multiple charging guns for each group of charging piles is generated. While providing more charging locations to solve the problem of imbalance in the supply and demand of charging piles, it can also take into account the load imbalance impact on the power grid caused by peak hours and generate a charging strategy for each group of charging piles that adapts to the user's charging needs, thereby solving the problem of unbalanced power grid load. Therefore, through data-driven intelligent charging control and the use of multiple guns and low-power design of a charging pile group, it is possible to provide charging locations for as many users as possible, and by generating more adaptable charging strategies, it is possible to provide more users with high-efficiency and high-flexibility charging solutions, thereby improving the overall charging efficiency within the region, reducing the overall charging time, making it easier for electric vehicles to charge, reducing the load on the power grid, reducing peak loads, and improving the stability of the power grid.

[0063] In a preferred embodiment, the power grid load forecasting module 10 specifically includes: a forecasting model establishment unit, which is configured to extract historical time window data from historical power grid load data, and input the historical time window data into the initial model for training to obtain a trained power grid load forecasting model; a power grid load forecasting unit, which is configured to input future time window data into the power grid load forecasting model to obtain power grid load forecasting data output by the power grid load forecasting model.

[0064] Among them, the prediction model establishment unit specifically includes: a historical time window data extraction subunit, which is configured to divide the historical power grid load data into time windows of fixed length, and regard the power grid load value in each time window as the pixel value of the image to form historical time window data; a model training subunit, which is configured to input each data point in the historical time window data into the CNN model for training to obtain a trained power grid load prediction model.

[0065] In this embodiment, in order to achieve grid load forecasting for the target period, a grid load forecasting algorithm based on a convolutional neural network is proposed. By capturing the temporal characteristics of historical time series data, the grid load in the future period is forecasted, thereby providing data support for generating a charging strategy for each charging pile. Specifically, the grid load forecasting algorithm based on a convolutional neural network includes the following implementation steps:

[0066] 1. Data Preparation: First, prepare historical grid load data. This data is a time series, including date and time stamps and corresponding grid load values. Divide this data into training and test sets.

[0067] 2. Feature Engineering: To use CNN for grid load forecasting, time series data must be converted into a form suitable for CNN input. The time series data is divided into fixed-length time windows, and the load values ​​within each time window are treated as image pixel values. This combines the time dimension and the load value dimension to form a two-dimensional input.

[0068] 3. Build a CNN model: Design a CNN model suitable for time series data. CNNs typically consist of convolutional layers, pooling layers, and fully connected layers. Convolutional layers extract time series features, pooling layers reduce data dimensionality, and fully connected layers output prediction results.

[0069] When using a convolutional neural network (CNN) for grid load forecasting, the following forward propagation process of the model is used:

[0070] (1) Input data representation:

[0071] There is a time series containing historical grid load data, which includes N time steps. We use X to represent the input data, whose dimensions are (N, D, 1), where N is the number of time steps and D is the feature dimension of each time step, which is 1 here (because we only consider the grid load value). i,d,1 represents a data point at time step i and feature dimension d.

[0072] (2) Convolutional layer operation:

[0073] Use a convolution layer for feature extraction, assuming that the convolution layer has K convolution kernels and the convolution kernel size is F. The convolution operation is expressed as:

[0074]

[0075] Among them, Z (1) is the output of the convolutional layer, W (1) is the weight of the convolution kernel, and k is the index of the convolution kernel. This operation extracts features by sliding the convolution kernel over the input data.

[0076] (3) Activation function:

[0077] To the output of each convolution kernel, we apply an activation function ReLU:

[0078]

[0079] (4) Pooling layer operation:

[0080] Next, we use the pooling layer to downsample the feature map. Assuming that we use the maximum pooling operation and the pooling window size is P, the pooling operation is expressed as:

[0081]

[0082] Pooling layers help reduce the dimensionality of the data and retain important features.

[0083] (5) Fully connected layer operation:

[0084] Finally, we connect the output of the pooling layer to a fully connected layer with H neurons. The fully connected operation is expressed as:

[0085]

[0086] Among them, Z (2) is the output of the fully connected layer, W (2) is the weight of the fully connected layer, and h is the index of the neuron.

[0087] (6) Output layer:

[0088] Finally, we can apply the output of the fully connected layer to an appropriate activation function, such as a linear activation function, to obtain the predicted value of the grid load:

[0089]

[0090] Predicted value Represents the predicted value of the power grid load at time step i.

[0091] 4. Loss Function and Optimizer: Choose an appropriate loss function to measure the difference between the predicted value and the true value. For regression problems, mean squared error is often used as the loss function. Choose an optimizer to minimize the loss function.

[0092] 5. Model training: Use the training set to train the model. Input the time window data into the CNN model and update the weights according to the loss function until the stopping condition is met (such as a fixed training cycle or loss function convergence).

[0093] The following is a specific example of an algorithm formula for loss function and weight update:

[0094] Loss function:

[0095] Among them, L(θ) represents the loss function, θ represents the parameters of the CNN model, N represents the number of samples, and y i Indicates the actual grid load value, Represents the predicted value of the CNN model.

[0096] Weight update uses the gradient descent method to update the model weights according to the gradient of the loss function:

[0097]

[0098] in, represents the gradient of the loss function with respect to the parameter θ, Represents the gradient of the CNN model output with respect to the parameter θ.

[0099] 6. Model evaluation: Use the test set to evaluate the performance of the model. In this embodiment, the accuracy of the model is measured by calculating the mean square error (MSE) or other related indicators.

[0100] 7. Predicting future grid load: The trained CNN model can be used to predict future grid load. By inputting data from a future time window into the model, the model will output the predicted grid load value.

[0101] Therefore, this embodiment uses convolutional neural networks to perform regional power grid load forecasting, which can automatically capture complex time series features in time series data and realize power grid load forecasting for future time periods, thereby improving the accuracy of power grid load forecasting.

[0102] In a preferred embodiment, the charging pile state classification module 20 specifically includes: a historical state data processing unit, which is configured to obtain the historical state data of each group of charging piles, extract the state characteristics of the historical state data, and calculate the confidence of the charging piles based on the state characteristics; a charging pile state classification unit, which is configured to calculate the confidence of the charging piles and the fuzzy membership degree of the threshold value using the fuzzy C-means method, and classify the state of each group of charging piles according to the fuzzy membership degree to obtain the charging pile state of each group of charging piles.

[0103] In this embodiment, after obtaining the historical status data, the data is first preprocessed, including data cleaning, outlier detection and missing value filling, to ensure the accuracy and completeness of the data, and then the historical status data processing unit is used to extract the state characteristics of the collected historical status data (charging power and charging period), for example: using Fourier transform or wavelet transform, extract frequency domain characteristics, such as spectrum, frequency components, power spectrum density and other state characteristics, and analyze the state characteristics to obtain the confidence of the charging pile. After that, the charging pile state classification unit uses the fuzzy C-means algorithm (FCM) to calculate the confidence of the charging pile and the fuzzy membership degree of the threshold, so as to classify the charging pile state. Among them, when the charging pile state classification unit uses the fuzzy C-means algorithm (FCM) to calculate the confidence of the charging pile and the fuzzy membership degree of the threshold, it specifically includes the following implementation steps:

[0104] (1) Classification definition:

[0105] 1. Assume there are N data points and C cluster centers, each cluster center represents the status of a charging pile.

[0106] 2. Define the eigenvector of the data point: X=(x1,x2,…,x N )

[0107] 3. Define the characteristic vector V of the cluster center = (v1, v2, ..., v C )

[0108] (2) Fuzzy membership calculation:

[0109] 4. Fuzzy membership matrix U: [U=[u ij ] N×C ], where (u ij ) represents the membership value of the i-th data point to the j-th cluster center.

[0110] 5. Distance measurement formula, usually using Euclidean distance: [d ij =||X i -V j || 2 ], where (X i ) is the eigenvector of the i-th data point, (V j ) is the eigenvector of the j-th cluster center.

[0111] 6. By calculating the U matrix, the affiliation between each charging pile state category and each data point in the charging pile historical data can be determined. This means that each data point is assigned to a different state category, and its membership value describes the degree to which it belongs to each state. By analyzing the U matrix, it is possible to determine which state category each data point is most likely to be in. Specifically, for each data point, the cluster center with the largest membership value in the U matrix can be found, and this cluster center is the most likely state category. Furthermore, it can help identify the current state of the charging pile, such as idle state, charging state, fault state, and deactivated state, so that corresponding measures and decisions can be taken. Therefore, the fuzzy membership update formula is:

[0112]

[0113] Where (m) is the ambiguity parameter (usually an integer greater than 1), and (t) represents the number of iterations.

[0114] (3) Charging pile status classification:

[0115] 7. Cluster centers usually represent different state categories in the charging pile management system, such as idle state, charging state, fault state and disabled state. Each cluster center of each state category represents a charging pile state, and the membership value u ij Indicates the degree to which data point i belongs to charging pile state j.

[0116] Therefore, this embodiment establishes a fuzzy membership matrix to determine the most likely charging pile state of the charging pile corresponding to each data point, thereby classifying each charging pile, and then helping to identify the current state of the charging pile, and then providing data support for the generation of a charging strategy for each charging pile, realizing a data-driven intelligent charging method, optimizing grid power, and improving power system stability.

[0117] In a preferred embodiment, the system also includes: a charging pile status update module, which specifically includes: a charging pile real-time status acquisition unit, which is configured to obtain the real-time status of the charging piles of each group of charging piles; a cluster center update unit, which is configured to update the fuzzy membership value according to the real-time status of the charging piles, and update the cluster center according to the updated fuzzy membership value; a charging pile status update unit, which is configured to update the charging pile status of the charging piles in the target time period according to the updated cluster center.

[0118] In this embodiment, considering the potential for charging pile failures, communication interruptions, or other abnormalities within the charging pile network, which can lead to instability across the entire charging pile network, the system updates the fuzzy membership values ​​by obtaining the real-time status of each group of charging piles, thereby updating the cluster centers. By continuously iteratively updating the positions of the cluster centers, the system can gradually adjust the boundaries and characteristics of each state to better match the actual data distribution. Ultimately, the state represented by the cluster center will more accurately reflect the status of the charging pile, thereby helping the system determine the status of the charging piles during the target time period.

[0119] The cluster center position will be updated according to the membership value of the data point during the algorithm iteration process. The cluster center update formula is:

[0120]

[0121] The cluster center update formula is to re-determine the position of each cluster center by considering the degree of membership of the data point and the cluster center to better fit the data distribution. j represents the new location of the jth cluster center, u ij is the membership value of data point i to cluster center j, X i is the eigenvector of data point i, and m is the fuzziness parameter. The cluster center update formula is a key step for adjusting the position of the charging pile status category. By updating the cluster center, the system can more accurately determine the status of the charging pile. Therefore, the charging pile status update module provided in this embodiment continuously monitors the behavior of the charging pile for dynamic adjustment, and then updates the charging pile status of the charging pile based on the updated cluster center. When a charging pile fails, it can provide timely warning and identification, thereby achieving the effect of optimizing grid power utilization, improving power system stability, and reducing costs.

[0122] In a preferred embodiment, the charging gun usage requirements include the charging time period and charging power requirements of the charging gun. Based on this, the charging strategy generation module 40 specifically includes: a charging power allocation unit for charging piles, configured to allocate corresponding charging power to each group of charging piles based on grid load forecast data and the status of each group of charging piles to balance the grid load; and a charging strategy generation unit for charging guns, configured to generate a charging strategy for each charging gun in each group of charging piles based on the charging power allocated to each group of charging piles and the charging time period and charging power requirements of the user's charging gun, so as to match each user's charging gun usage requirements with the corresponding charging gun.

[0123] In this embodiment, first, the system allocates corresponding charging power to each group of charging piles based on the grid load forecast data and the charging pile status of each group of charging piles to balance the grid load. That is, the system considers the status of the charging piles of each group of charging piles and allocates different charging powers to different charging piles to achieve regional grid load balance. Under such conditions, even if there is a charging peak, it will not affect the grid balance.

[0124] On this basis, taking into account the user's charging period requirements and charging power requirements, a charging strategy for executing charging actions is generated for each charging gun of each group of charging piles, so that each user's charging gun usage requirements match the corresponding charging gun. That is, by obtaining the charging needs of each user, the corresponding charging control instructions are allocated to the charging gun of the corresponding charging pile to adapt to the user's charging needs. Under such conditions, charging locations can be provided for as many users as possible. At the same time, since the charging strategy (charging time and charging power) is precisely allocated according to time and charging guns, even if a user does not move the vehicle in time after charging is completed, the low-power design of multiple charging guns will not affect the use of other charging guns, which, to a certain extent, improves the overall charging efficiency.

[0125] Therefore, the charging power allocation unit of the charging pile allocates corresponding charging power to each group of charging piles according to the grid load forecast data and the charging pile status of each group of charging piles, and the charging strategy generation unit of the charging gun generates a charging strategy for each charging gun of each group of charging piles according to the charging power allocated to each group of charging piles and the charging period requirement and charging power requirement of the user's charging gun. The charging pile multi-gun low-power charging intelligent control solution can achieve balanced grid load, provide users with as many charging locations as possible, and improve overall charging efficiency, thereby improving system benefits.

[0126] Reference Figure 2 , Figure 2 A schematic flow chart of a multi-gun low-power charging pile method provided in an embodiment of the present invention.

[0127] like Figure 2 As shown, a multi-gun low-power charging pile method includes the following steps:

[0128] S1: Acquire historical grid load data within a region, and predict grid load forecast data for a target period based on the historical grid load data;

[0129] S2: Obtain historical status data of each group of charging piles, and classify the charging piles according to the historical status data to obtain the charging pile status of each group of charging piles;

[0130] S3: Obtain the user's charging gun usage requirements;

[0131] S4: generating a charging strategy for controlling the charging guns of each group of charging piles to perform charging actions according to the grid load forecast data, the charging pile status of each group of charging piles, and the user's charging gun usage requirements.

[0132] In this embodiment, by predicting the grid load data and utilizing the status of the charging piles and the user's charging gun usage needs, a charging strategy for multiple charging guns of each group of charging piles is generated. This can provide more charging locations to solve the problem of imbalance in supply and demand of charging piles while taking into account the load imbalance impact on the grid caused by peak hours. A charging strategy that adapts to the user's charging needs is generated for each group of charging piles, thereby solving the problem of imbalance in grid load.

[0133] The specific implementation of the multi-gun low-power charging pile method of the present application is basically the same as the above-mentioned embodiments of the multi-gun low-power charging pile system, and will not be repeated here.

[0134] In the description of the embodiments of the present invention, it should be understood that the terms "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "center", "top", "bottom", "top", "bottom", "inside", "outside", "inner side", "outer side" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. Among them, "inside" refers to an internal or enclosed area or space. "Periphery" refers to the area surrounding a specific component or specific area.

[0135] In the description of the embodiments of the present invention, the terms "first," "second," "third," and "fourth" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly specifying the number of the technical features indicated. Therefore, a feature specified as "first," "second," "third," or "fourth" may explicitly or implicitly include one or more of the features. In the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0136] In the description of the embodiments of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "installed," "connected," "connected," and "assembled" should be understood in a broad sense. For example, they may refer to fixed connections, detachable connections, or integral connections; they may refer to direct connections, indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention in specific circumstances.

[0137] In the description of the embodiments of the present invention, specific features, structures, materials or characteristics may be combined in an appropriate manner in any one or more embodiments or examples.

[0138] In describing the embodiments of the present invention, it should be understood that "-" and "~" represent a range between two values, and the range includes the endpoints. For example, "AB" represents a range greater than or equal to A and less than or equal to B. "A~B" represents a range greater than or equal to A and less than or equal to B.

[0139] In describing the embodiments of the present invention, the term "and / or" is used herein to describe a relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. Furthermore, the character " / " is generally used herein to indicate that the associated objects are in an "or" relationship.

[0140] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A multi-gun low-power charging pile system, characterized in that: include: A power grid load forecasting module, configured to obtain historical power grid load data within a region and forecast power grid load forecast data for a target period based on the historical power grid load data; A charging pile status classification module, wherein the charging pile status determination module is configured to obtain historical status data of each group of charging piles and classify the charging piles according to the historical status data to obtain the charging pile status of each group of charging piles; the charging pile status includes: idle state, charging state, fault state and disabled state; A user demand acquisition module, configured to acquire the user's demand for using a charging gun; a charging strategy generation module, the charging strategy generation module being configured to generate a charging strategy for controlling the charging guns of each group of charging piles to perform charging actions based on the grid load forecast data, the charging pile status of each group of charging piles, and the user's charging gun usage requirements; The charging pile state classification module specifically includes: a historical state data processing unit, the historical state data processing unit is configured to obtain historical state data of each group of charging piles, extract state features of the historical state data, and calculate the confidence of the charging piles based on the state features; a charging pile state classification unit, the charging pile state classification unit is configured to calculate the confidence of the charging piles and the fuzzy membership degree of the threshold value using the fuzzy C-means method, and classify the state of each group of charging piles according to the fuzzy membership degree to obtain the charging pile state of each group of charging piles; The charging pile state classification unit specifically includes: a classification definition subunit, wherein the classification definition subunit is configured to define the feature vector of the data point as , define the eigenvector of the cluster center as Wherein, N is the number of data points, C is the number of cluster centers, and each cluster center represents the state of a charging pile; the fuzzy membership calculation subunit is configured to calculate the fuzzy membership matrix according to the fuzzy membership matrix. , using Euclidean distance , calculate the fuzzy membership value : ;in, is the eigenvector of the i-th data point, It is The eigenvectors of cluster centers, is the fuzziness parameter, Take an integer greater than 1, represents the number of iterations; a charging pile state classification subunit, wherein the charging pile state classification subunit is configured to take the cluster center with the largest fuzzy membership value as the charging pile state of the charging pile.

2. The multi-gun low-power charging pile system according to claim 1 is characterized in that: The power grid load prediction module specifically includes: A prediction model establishment unit, wherein the prediction model establishment unit is configured to extract historical time window data from historical power grid load data, and input the historical time window data into an initial model for training to obtain a trained power grid load prediction model; A power grid load prediction unit is configured to input future time window data into the power grid load prediction model to obtain power grid load prediction data output by the power grid load prediction model.

3. The multi-gun low-power charging pile system according to claim 2, characterized in that: The prediction model building unit specifically includes: a historical time window data extraction subunit, wherein the historical time window data extraction subunit is configured to divide the historical power grid load data into time windows of fixed length, and regard the power grid load value in each time window as the pixel value of the image to form the historical time window data; The model training subunit is configured to input each data point in the historical time window data into the CNN model for training to obtain a trained power grid load forecasting model.

4. The multi-gun low-power charging pile system according to claim 3, characterized in that: In the model training subunit, the historical time window data includes N input data points at time step i and feature dimension d. ; The CNN model includes: Convolutional layer, the expression of the convolutional layer is: ； Among them, F is the convolution kernel size, is the output of the convolutional layer, is the weight of the convolution kernel is the index of the convolution kernel; Pooling layer, the expression of the pooling layer is: ； ； Among them, P is the pooling window size, is the activation function ReLU; Fully connected layer, the expression of the fully connected layer is: ； in, is the output of the fully connected layer, is the weight of the fully connected layer, is the index of the neuron; Output layer, the expression of the output layer is: ； in, Indicates that at time step The predicted value of grid load.

5. The multi-gun low-power charging pile system according to claim 1, characterized in that: The system further includes a charging pile status update module, which specifically includes: A charging pile real-time status acquisition unit, wherein the charging pile real-time status acquisition module is configured to obtain the real-time status of each group of charging piles; A cluster center updating unit is configured to update the fuzzy membership value according to the real-time status of the charging pile, and update the cluster center according to the updated fuzzy membership value; wherein the cluster center update expression is specifically: ； in, Representative The new locations of the cluster centers, is a data point For cluster centers The fuzzy membership value of is a data point The eigenvector of is the fuzziness parameter; A charging pile state updating unit is configured to update the charging pile state of the charging pile in the target time period according to the updated cluster center.

6. The multi-gun low-power charging pile system according to claim 1, characterized in that: The charging gun usage requirements include the charging time period requirements and charging power requirements of the charging gun.

7. The multi-gun low-power charging pile system according to claim 6, characterized in that: The charging strategy generation module specifically includes: A charging power distribution unit for the charging piles, the charging power distribution unit for the charging piles being configured to distribute corresponding charging power to each group of charging piles based on grid load forecast data and the status of the charging piles of each group of charging piles, so as to balance the grid load; A charging strategy generation unit for the charging gun is configured to generate a charging strategy for executing charging actions for each charging gun in each group of charging piles based on the charging power allocated to each group of charging piles and the charging period and charging power requirements of the user's charging gun, so that each user's charging gun usage requirements match the corresponding charging gun.

8. A multi-gun low-power charging pile method, characterized in that: include: Obtaining historical grid load data within a region, and predicting grid load forecast data for a target period based on the historical grid load data; Obtaining historical status data of each group of charging piles, and classifying the charging piles according to the historical status data to obtain the charging pile status of each group of charging piles; The charging pile status includes: idle state, charging state, fault state and disabled state; Obtain the user's demand for charging guns; generating a charging strategy for controlling the charging guns of each group of charging piles to perform charging actions based on the grid load forecast data, the charging pile status of each group of charging piles, and the user's charging gun usage requirements; The method further includes: obtaining historical status data of each group of charging piles, extracting status features of the historical status data, and calculating the confidence of the charging piles based on the status features; calculating the confidence of the charging piles and the fuzzy membership degree of the threshold using the fuzzy C-means method, and classifying the status of each group of charging piles based on the fuzzy membership degree to obtain the charging pile status of each group of charging piles; The eigenvector of a data point is defined as , define the eigenvector of the cluster center as Wherein, N is the number of data points, C is the number of cluster centers, and each cluster center represents the state of a charging pile; the fuzzy membership calculation subunit is configured to calculate the fuzzy membership matrix according to the fuzzy membership matrix. , using Euclidean distance , calculate the fuzzy membership value : ;in, is the eigenvector of the i-th data point, It is The eigenvectors of cluster centers, is the fuzziness parameter, Take an integer greater than 1, Indicates the number of iterations; the charging pile status classification subunit takes the cluster center with the largest fuzzy membership value as the charging pile status of the charging pile.

Images