Intelligent management and control method and system for orderly charging in transformer area, terminal and storage medium

Abstract

The invention relates to the technical field of transformer area charging, and particularly provides a transformer area ordered charging intelligent control method and system, a terminal and a storage medium, and the method comprises the steps: building bidirectional communication connection between a charging pile and a transformer area use and collection terminal; after the connection is established, acquiring charging information and real-time electricity price by the transformer area using and acquiring terminal; constructing an MLP fusion model, fusing multi-dimensional data in the charging information into a fusion feature vector based on the MLP fusion model, and training a CNN-LSTM load prediction model based on the fusion feature vector; a load prediction result is obtained, a charging mode is selected based on the predicted load result, the real-time electricity price and the charging information, a charging strategy is dynamically adjusted based on the charging model, and the charging mode comprises a standard charging mode, a timing mode and an intelligent charging mode. The beneficial effects of improving the system compatibility, enhancing the load prediction accuracy, improving the intelligent management level to optimize the user experience, reducing the communication cost, optimizing the charging load management and the like are achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of charging in the substation area, and particularly relates to an intelligent control method, system, terminal and storage medium for orderly charging in the substation area. Background Art

[0002] With the development of new energy vehicles, the occupancy rate of new energy vehicles is getting higher and higher, the demand for vehicle charging has increased sharply, and the installation volume of charging piles has increased significantly.

[0003] The existing electric vehicle charging management system mainly includes traditional charging piles, a centralized charging management system and some basic load forecasting and control methods. Traditional charging piles are distributed in multiple places such as residential areas, commercial areas and public areas; the centralized charging management system can centrally monitor and manage multiple charging piles; through load forecasting, when it is predicted that the charging load may exceed the grid carrying capacity, a variety of control methods are used to adjust.

[0004] Due to the variety of existing charging pile brands and models and the lack of industry standards, the compatibility between devices is poor; relying on simple statistical methods results in low load forecasting accuracy; the centralized charging management system often focuses on static scheduling and cannot dynamically adjust the charging strategy according to real-time data; it increases the complexity of system integration, has a negative impact on the overall operation efficiency of the power grid, resulting in low power utilization efficiency and reduced user experience. Summary of the Invention

[0005] In view of the above deficiencies of the prior art, the present invention provides an intelligent control method, system, terminal and storage medium for orderly charging in the substation area to solve the above technical problems.

[0006] In the first aspect, the present invention provides an intelligent control method for orderly charging in the substation area, including: S1, establishing a bidirectional communication connection between the charging pile and the power acquisition terminal in the substation area based on the power line carrier communication technology and protocol conversion technology; S2, after establishing the connection, the power acquisition terminal in the substation area obtains charging information, user requirements and real-time electricity prices, and the charging information includes charging trips, charging frequencies, SOC, charging habits and historical grid load data; S3, constructing an MLP fusion model and fusing multi-dimensional data in the charging information into a fusion feature vector based on the MLP fusion model, and training a CNN-LSTM load forecasting model based on the fusion feature vector; S4, obtaining a load forecasting result based on the CNN-LSTM forecasting model combined with real-time charging information, selecting a charging mode based on the predicted load result, user requirements and real-time electricity price, and dynamically adjusting the charging strategy based on the charging mode, and the charging mode includes an immediate charging mode, a timing mode and a smart charging mode.

[0007] In an alternative embodiment, step S1 specifically includes: Converting the data sent by the charging pile from the CAN protocol to the Ethernet protocol, or converting the data sent by the power consumption terminal in the substation area from the Ethernet protocol to the CAN protocol; Converting the data sent by the charging pile or the power consumption terminal in the substation area into a high-frequency signal suitable for power line transmission based on modulation technology; Amplifying the high-frequency signal based on a gain mechanism and transmitting the amplified high-frequency signal through the power line; Receiving the amplified high-frequency signal transmitted through the power line, and demodulating the high-frequency signal into the original low-frequency digital signal based on a demodulation algorithm; The power consumption terminal in the substation area or the charging pile receives the low-frequency digital signal, completing the two-way communication connection between the charging pile and the power consumption terminal in the substation area.

[0008] In an alternative embodiment, Converting the data sent by the charging pile from the CAN protocol to the Ethernet protocol specifically includes: When the charging pile sends data, parsing the data into a CAN data frame based on the CAN protocol specification, and the data frame includes a start bit, an arbitration field, a control field, a data field, a CRC check field, and an end bit; Extracting key information from the parsed CAN data field and temporarily storing this information in a memory buffer; Encapsulating the key information into an Ethernet protocol frame in the format of the Ethernet protocol, and the Ethernet protocol frame includes the destination address of the power consumption terminal in the substation area, the conversion source address, the Ethernet type, the data information, and the frame check sequence; Sending the encapsulated Ethernet frame through an Ethernet interface.

[0009] In an alternative embodiment, Converting the data sent by the power consumption terminal in the substation area from the Ethernet protocol to the CAN protocol specifically includes: When the power consumption terminal in the substation area sends data, parsing the Ethernet data frame based on the Ethernet protocol specification, and the Ethernet data frame includes the target charging pile address, the source address of the power consumption terminal in the substation area, the data type, and the data content; Parsing relevant parameters according to the control instructions in the data content and temporarily storing the relevant parameters in a memory buffer; Encapsulating the control instructions and relevant parameters into a CAN protocol frame according to the format of the CAN protocol, and the CAN protocol frame includes a start bit, an arbitration field, a control field, a data field, a CRC check field, and an end bit; Sending the encapsulated CAN protocol data to the CAN bus.

[0010] In an alternative embodiment, in step S3, constructing the MLP fusion model specifically includes: Construct a data set based on historical charging information and the corresponding historical fusion multi-dimensional vectors, and divide the data set into a training set, a validation set, and a test set; Determine the number of neurons in the input layer based on the number of multi-dimensional data features in the basic information, set the hidden layer, and determine the output layer based on the output fusion feature vector to construct an initial MLP fusion model. The input vector is the historical charging information, and the output vector is the historical fusion multi-dimensional vector; For each sample in the training set, input the input feature vector into the initial MLP fusion model, and calculate the predicted output vector of the model through forward propagation; Calculate the loss value between the predicted output vector and the actual output feature vector according to the loss function, and use the backpropagation algorithm to optimize the parameters of the initial MLP fusion model; Calculate the evaluation index based on the validation set combined with the root mean square error function every preset number of training times, and adjust the model structure based on the change trend of the evaluation index; Perform a final evaluation on the trained initial MLP fusion model based on the test set to obtain the final MLP fusion model.

[0011] In an alternative embodiment, in step S3, training the CNN-LSTM load prediction model specifically includes: Construct a data set based on the historical fusion feature vector and the historical load result, and divide the data set into a training set, a validation set, and a test set; Determine the parameters of the convolutional layer, including the number of convolutional layers, the size of the convolutional kernel, and the stride parameter; Configure the pooling layer based on the max pooling method, and reduce the dimension of the features output by the convolutional layer by setting the size of the pooling window; Determine the number of LSTM layers and the number of neurons in each LSTM layer; Configure the output layer based on the dimension of the load prediction result to obtain the initial CNN-LSTM load prediction model; the historical fusion feature vector is the input vector of the model, and the historical load result is the output vector of the model; For each sample in the training set, input the input feature vector into the CNN-LSTM load prediction model, perform forward propagation calculation. In the convolutional layer, perform a convolution operation by sliding the convolutional kernel on the historical fusion feature vector to extract feature data, and then input the feature data into the LSTM layer after passing through the pooling layer to learn the temporal pattern in the data, and finally output the load prediction result through the output layer; Calculate the loss value between the predicted output vector and the actual output feature vector according to the loss function, and use the backpropagation algorithm to optimize the parameters of the initial CNN-LSTM load prediction model; Calculate the evaluation index based on the root mean square error function in combination with the validation set every preset number of training times, and adjust the model structure based on the change trend of the evaluation index; Finally evaluate the trained initial CNN-LSTM load prediction model based on the test set to obtain the final CNN-LSTM load prediction model.

[0012] In an alternative embodiment, step S4 specifically includes: When it is determined that the load prediction result is abnormal, charging is not performed; When it is determined that the load prediction result is lower than the load threshold, there are no special time requirements in the user demand, and the stepped electricity price is lower than the electricity price threshold, select the immediate charging mode, and the charging pile immediately starts to charge the vehicle at a preset high power; When it is determined that the fluctuation of the load prediction result is less than the fluctuation threshold and there are time requirements in the user demand, select the timing mode. The user sets the charging time, configures the timed charging power based on the charging time, and the charging pile performs charging operations based on the charging time and the timed charging power; In other cases, select the intelligent mode, obtain the low-load period and the high-load period based on the load prediction result and combine the real-time electricity price, and automatically adjust the charging period or the charging power; During the charging process, obtain the load prediction result in real time, determine whether there is a sudden increase or abnormality in the load. If it is determined that there is, adjust the charging power of the device being charged or stop charging.

[0013] In a second aspect, the present invention provides a smart control system for orderly charging in a substation area. When the system is implemented, it executes the above-mentioned smart control method for orderly charging in a substation area. The system includes: A connection establishment module that establishes a two-way communication connection between the charging pile and the substation area usage and collection terminal based on power line carrier communication technology and protocol conversion technology; An information acquisition module. After establishing the connection, the substation area usage and collection terminal acquires charging information, user demand, and real-time electricity price. The charging information includes the number of charging times, charging frequency, SOC, charging habits, and historical power grid load data; A model construction module that constructs an MLP fusion model and fuses multi-dimensional data in the charging information into a fusion feature vector based on the MLP fusion model, and trains a CNN-LSTM load prediction model based on the fusion feature vector; A strategy adjustment module that obtains the load prediction result based on the CNN-LSTM prediction model in combination with real-time charging information, selects a charging mode based on the predicted load result, user demand, and real-time electricity price, and dynamically adjusts the charging strategy based on the charging mode. The charging modes include the immediate charging mode, the timing mode, and the intelligent charging mode.

[0014] In a third aspect, a terminal is provided, including: A processor and a memory, wherein, the memory is used to store a computer program, the processor is used to call and run the computer program from the memory, so that the terminal executes the method of the terminal described above.

[0015] In a fourth aspect, a computer-readable storage medium is provided. Instructions are stored in the computer-readable storage medium. When it runs on a computer, it causes the computer to execute the methods described in the above aspects.

[0016] The beneficial effects of the present invention are as follows. The intelligent control method, system, terminal and storage medium for orderly charging in the power distribution area provided by the present invention establish a two-way communication connection based on power line carrier communication technology and protocol conversion technology, obtain multi-faceted data, process it through a fusion model and a load prediction model, select a charging mode according to user needs and electricity prices based on the prediction results, and dynamically adjust the strategy, achieving beneficial effects in multiple aspects such as improving system compatibility, enhancing the accuracy of load prediction, improving the level of intelligent management to optimize the user experience, reducing communication costs, and optimizing charging load management, providing an efficient, intelligent and economically stable solution for electric vehicle charging management.

[0017] In addition, the design principle of the present invention is reliable and the structure is simple, having a very broad application prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the present invention, the drawings required for description will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0019] Figure 1 is a schematic flowchart of the intelligent control method for orderly charging in the power distribution area according to an embodiment of the present invention.

[0020] Figure 2 is the communication architecture of protocol conversion according to an embodiment of the present invention.

[0021] Figure 3 is a full-link communication scheme diagram for constructing orderly charging using power line carrier communication (PLC) technology according to an embodiment of the present invention.

[0022] Figure 4 is a detailed flowchart of power line carrier communication technology according to an embodiment of the present invention.

[0023] Figure 5 is a schematic block diagram of the intelligent control system for orderly charging in the power distribution area according to an embodiment of the present invention.

[0024] Figure 6 A schematic structural diagram of a terminal provided by an embodiment of the present invention. Detailed implementation manners

[0025] In order to enable those skilled in the art of the present technology to better understand the technical solutions in the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0026] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments, and are not intended to limit the present invention.

[0027] The method for intelligent control of orderly charging in a power distribution area provided by an embodiment of the present invention is executed by a computer device. Correspondingly, the intelligent control system for orderly charging in a power distribution area runs in the computer device.

[0028] Figure 1 It is a schematic flowchart of the method for intelligent control of orderly charging in a power distribution area according to an embodiment of the present invention. Among them, Figure 1 The execution subject can be an intelligent control system for orderly charging in a power distribution area. According to different requirements, the order of the steps in this flowchart can be changed, and some can be omitted.

[0029] As Figure 1 shown, the method includes: S1. Establish a two-way communication connection between the charging pile and the power distribution area usage and acquisition terminal based on the power line carrier communication technology and protocol conversion technology; Compared with traditional communication methods such as 4G and Wi-Fi, power line carrier communication is not affected by the wireless signal coverage range and network stability, improving the communication stability and reliability. The protocol conversion technology enables charging piles of different brands and models to be compatible with the power distribution area usage and acquisition terminal, solves the problem of incompatible communication between charging piles, and enhances the adaptability and flexibility of the entire charging system.

[0030] S2. After the connection is established, the power distribution area usage and acquisition terminal obtains charging information, user requirements, and real-time electricity prices. The charging information includes the charging vehicle number, charging frequency, SOC, charging habits, and historical power grid load data; These data can comprehensively reflect the user's charging behavior patterns, the vehicle battery status, and the operating conditions of the power grid, which helps to deeply analyze the changing rules of charging demand, thus providing strong support for an accurate load prediction model and further realizing the reasonable allocation of charging time and power.

[0031] S3. Construct an MLP fusion model and fuse the multi-dimensional data in the charging information into a fused feature vector based on the MLP fusion model, and train a CNN-LSTM load prediction model based on the fused feature vector; The MLP fusion model can effectively integrate the multi-dimensional data in the charging information, mine the internal correlations between different data, and transform them into more representative fused feature vectors, thereby providing more comprehensive and accurate input features for load prediction. The CNN-LSTM load prediction model trained based on these fused feature vectors fully utilizes the ability of CNN to extract local features of data and the advantage of LSTM in processing time series data, and can accurately capture the changing trends and complex rules of charging load, significantly improving the accuracy and reliability of load prediction.

[0032] S4. Obtain the load prediction result based on the CNN-LSTM prediction model combined with real-time charging information, select a charging mode based on the predicted load result, user demand, and real-time electricity price, and dynamically adjust the charging strategy based on the charging mode. The charging modes include immediate charging mode, timed charging mode, and intelligent charging mode.

[0033] The immediate charging mode meets the user's immediate charging needs, the timed charging mode is convenient for users to charge at specific times, and the intelligent charging mode dynamically adjusts during the peak period of the power grid. This greatly improves the user experience and meets the personalized needs of different user groups. Dynamically adjusting the charging strategy can optimize the charging time and power according to the real-time conditions of the power grid, avoid power grid overload, improve the operation efficiency of the power grid, and realize the intelligentization of charging management. By reasonably using the low-price period for charging, the user's charging cost is reduced, while the charging load management is optimized, and the charging during the peak period is centrally scheduled, effectively reducing the system load pressure.

[0034] Optionally, as an embodiment of the present invention, referring to Figure 2 , the protocol conversion technology in step S1 specifically includes developing a dedicated protocol conversion module, which has a CAN interface and an Ethernet interface for connecting the charging pile and the transformer area power consumption acquisition terminal.

[0035] The figure shows multiple charging pile guns (such as gun No. 1, gun No. x, etc.), representing different charging interfaces or charging piles at different locations, which reflects that the system can be compatible with multiple charging pile devices and adapt to the diversity of the distribution and use of charging piles in practical applications.

[0036] Each charging pile gun is connected to the corresponding protocol conversion module through specific cables (e.g., gun No. 1 is connected to module No. 1, gun No. x is connected to module No. x, etc.). This one-to-one connection method ensures that the data of each charging pile can be independently and accurately protocol-converted, avoiding data confusion and transmission errors.

[0037] The protocol conversion module converts the charging pile data from the CAN protocol to the Ethernet communication protocol or converts the data of the substation access terminal from the Ethernet communication protocol to the CAN protocol. These modules have CAN interfaces and Ethernet interfaces, realizing the physical connection and data conversion between two different protocols. As can be seen from the figure, protocol conversion modules with different numbers respectively process the data from the corresponding charging pile guns, ensuring the orderliness and accuracy of data conversion.

[0038] The modules are connected to the main control board through specific control lines, which enables the main control board to control and manage each protocol conversion module, such as coordinating data transmission, monitoring the working status of the modules, etc., to ensure the stable operation of the entire communication system.

[0039] The main control board is connected to the external network (such as the network where the substation access terminal is located) through the Ethernet interface to realize data interaction with a wider system. The use of the Ethernet interface provides a high-speed and stable data transmission channel, which can meet the real-time transmission requirements of a large amount of data between the charging pile and the substation access terminal.

[0040] The connection of the control line between the main control board and the protocol conversion module enables it to receive the Ethernet data after protocol conversion from the charging pile and further process or forward these data to the substation access terminal. At the same time, the main control board can also receive control instructions from the substation access terminal through the Ethernet, and then send the instructions to the corresponding protocol conversion module through the control line. The protocol conversion module then converts the instructions into the CAN protocol format and sends them to the charging pile, realizing the two-way transmission of information.

[0041] Optionally, as an embodiment of the present invention, when the charging pile sends data, the protocol conversion module receives the CAN protocol data through the CAN interface. It parses the data frame according to the CAN protocol specification, including identifying the start bit, arbitration field (determining data priority and sending source), control field, data field (containing the actual information to be transmitted, such as charging pile status, charging parameters, etc.), CRC check field (for data integrity check) and end bit.

[0042] Extract key information from the parsed CAN data field, such as the ID of the charging pile (used to distinguish different charging piles), charging current, voltage, SOC (state of charge of the battery), charging status (charging, charging completed, fault, etc.), and temporarily store this information in the memory buffer of the conversion module, preparing for Ethernet protocol encapsulation.

[0043] Write a conversion program to repackage the information parsed from the CAN protocol in the format of the Ethernet protocol. The Ethernet protocol frame includes the destination MAC address (determining the receiving device, here it is the MAC address of the substation collection terminal), the source MAC address (the MAC address of the conversion module itself or set according to the system configuration), the Ethernet type (identifying the upper-layer protocol type, such as a specific type used to represent charging pile data here), the data (i.e., the charging pile information converted from the CAN protocol), and the frame check sequence (used to ensure the integrity of the Ethernet frame).

[0044] Send the encapsulated Ethernet frame through the Ethernet interface, and the data will be transmitted to the substation collection terminal according to the network configuration (such as the settings of devices like switches and routers, if any). During the sending process, follow the rules of Ethernet communication, such as the collision detection and retransmission mechanism (if it is in half-duplex mode), etc., to ensure that the data can be accurately transmitted to the target device.

[0045] Optionally, as an embodiment of the present invention, when the substation collection terminal sends data, the protocol conversion module receives the Ethernet protocol data through the Ethernet interface. It parses the data frame according to the Ethernet protocol specification and extracts information such as the destination address (determining the target charging pile), the source address (substation collection terminal address), data type, and data content.

[0046] According to the information such as the control instructions in the data content, parse relevant parameters, such as the power value in the charging power adjustment instruction, the charging time setting, operation instructions such as starting or stopping charging, and the identification of the target charging pile, etc., and temporarily store this information in the memory buffer of the conversion module, preparing for CAN protocol encapsulation.

[0047] Reorganize the parsed control instructions and parameters according to the data frame structure of the CAN protocol. Determine the arbitration field of the CAN data frame (set according to factors such as the priority of the target charging pile), the control field (such as indicating information like data length), the data field (fill in the parameters related to the control instructions, such as the adjusted charging power value, charging time, etc.), and calculate the CRC check field to ensure the correctness of the data. At the same time, set the appropriate start bit and end bit of the CAN data frame.

[0048] The encapsulated CAN protocol data is sent to the CAN bus through the CAN interface. According to the ID or address information of the target charging pile, the data will be broadcast on the CAN bus or transmitted directionally to the corresponding charging pile. During the sending process, the communication rules of the CAN bus, such as the identifier arbitration mechanism, error detection and handling mechanism, etc., should be followed to ensure that the data can be accurately transmitted to the target charging pile.

[0049] Optionally, as an embodiment of the present invention, referring to Figure 3 , in step S1, the power line carrier communication (PLC) technology is adopted to construct a full-link communication scheme for orderly charging. The electric vehicle exchanges data through the CAN protocol of the charging gun, and after protocol conversion and PLC modulation, it is connected to the user acquisition system and the main control system through the power line. The user acquisition system and the main control system obtain the charging pile information after PLC demodulation. At the same time, the main control system can also send the charging control instruction to the user acquisition system and the charging pile through the power carrier to realize the two-way transmission of information. In addition, this design scheme also fully considers the AC and DC modes of the charging pile, as well as the randomness and dynamics of the connection power line phase, and proposes a new full-phase recognition aggregation design scheme. A PLC modulation device is designed for each path, which is connected to the main control board through an exchanger and displayed on the screen, ensuring the full reception of the user's charging information.

[0050] Optionally, as an embodiment of the present invention, the operation process of the power line carrier communication (PLC) technology is specifically as Figure 4 shown, and can be summarized as: Based on the modulation technology, the data sent by the charging pile or the user acquisition terminal of the substation area is converted into a high-frequency signal suitable for transmission through the power line; Based on the gain mechanism, the high-frequency signal is amplified and the amplified high-frequency signal is transmitted through the power line; Receive the amplified high-frequency signal transmitted through the power line, and based on the demodulation algorithm, demodulate the high-frequency signal into the original low-frequency digital signal; The user acquisition terminal of the substation area or the charging pile receives the low-frequency digital signal, and completes the two-way communication connection between the charging pile and the user acquisition terminal of the substation area.

[0051] Optionally, as an embodiment of the present invention, the charging information in step 2 specifically includes: Charging trips: The number of times a user charges, indicating the charging demand frequency of each user; Charging frequency: The number of times a user charges per day or per week, reflecting the regularity of charging behavior; SOC: The current charging state of the battery, which determines the demand for each charge; Charging habit: Including behavioral characteristics such as charging time period and charging duration, reflecting the user's personalized charging preference; Historical power grid load data: The load data of the power grid over a period of time, which reflects the power consumption pattern and load fluctuations of the power grid; These data need to be preprocessed before model training, including: Data cleaning: Remove missing values or outliers to ensure the integrity and accuracy of the data.

[0052] Time synchronization: Align data from different sources according to timestamps for subsequent analysis.

[0053] Normalization processing: Standardize or normalize each feature to eliminate the influence of different feature scales.

[0054] Extract meaningful feature data from the data: Statistical features of user charging frequency: Such as average daily charging times, charging period distribution, charging duration, etc.; SOC change pattern: Such as the increase and decrease rate of SOC, the SOC distribution after charging, etc.; Power grid load fluctuation characteristics: Trends, periodicity, sudden fluctuations, etc. of historical power grid load data.

[0055] Optionally, as an embodiment of the present invention, in step S3, constructing the MLP fusion model specifically includes: Construct a data set based on historical charging information and the corresponding historical fusion multi-dimensional vector, and divide the data set into a training set, a validation set and a test set; Determine the number of input layer neurons based on the multi-dimensional data feature quantity in the basic information, set the hidden layer, and determine the output layer based on the output fusion feature vector to construct an initial MLP fusion model, where the input vector is historical charging information and the output vector is the historical fusion multi-dimensional vector; For each sample in the training set, input the input feature vector into the initial MLP fusion model and calculate the predicted output vector of the model through forward propagation; Calculate the loss value between the predicted output vector and the actual output feature vector according to the loss function, and use the backpropagation algorithm to optimize the parameters of the initial MLP fusion model; Calculate the evaluation index based on the validation set combined with the root mean square error function every preset number of training times, and adjust the model structure based on the change trend of the evaluation index; Conduct a final evaluation on the trained initial MLP fusion model based on the test set to obtain the final MLP fusion model.

[0056] Optionally, as an embodiment of the present invention, in step S3, training the CNN-LSTM load prediction model specifically includes: Construct a data set based on the historical fusion feature vector and the historical load result, and divide the data set into a training set, a validation set and a test set; Determine the parameters of the convolutional layer, including the number of convolutional layers, the size of the convolutional kernel, and the stride parameter; Configure the pooling layer based on the max pooling method to reduce the dimension of the features output by the convolutional layer by setting the size of the pooling window; Determine the number of LSTM layers and the number of neurons in each LSTM layer; Configure the output layer based on the dimension of the load prediction result to obtain the initial CNN-LSTM load prediction model; the historical fusion feature vector is the input vector of the model, and the historical load result is the output vector of the model; For each sample in the training set, input the input feature vector into the CNN-LSTM load prediction model for forward propagation calculation. In the convolutional layer, perform a convolution operation by sliding the convolutional kernel on the historical fusion feature vector to extract feature data. After passing through the pooling layer, input the feature data into the LSTM layer to learn the temporal pattern in the data, and finally output the load prediction result through the output layer; Calculate the loss value between the predicted output vector and the actual output feature vector according to the loss function, and use the backpropagation algorithm to optimize the parameters of the initial CNN-LSTM load prediction model; Calculate the evaluation index based on the validation set combined with the root mean square error function every preset number of training times, and adjust the model structure based on the change trend of the evaluation index; Perform a final evaluation on the trained initial CNN-LSTM load prediction model based on the test set to obtain the final CNN-LSTM load prediction model.

[0057] Optionally, as an embodiment of the present invention, step S4 specifically includes: When it is determined that the load prediction result is abnormal, do not charge; When it is determined that the load prediction result is lower than the load threshold, there is no special time requirement in the user demand, and the stepped electricity price is lower than the electricity price threshold, select the immediate charging mode, and the charging pile immediately starts to charge the vehicle at a preset high power; for example, during low-load periods such as morning or night, when the user uses the immediate charging mode, the charging pile can continuously charge the vehicle at a higher power until it is full or the user manually stops charging.

[0058] When it is determined that the fluctuation of the load prediction result is less than the fluctuation threshold and there is a time requirement in the user demand, the timing mode is selected. The user sets the charging time, configures the timed charging power based on the charging time, and the charging pile performs charging operations based on the charging time and the timed charging power. If the user does not specify an end time, the system will calculate the approximate charging end time according to the remaining battery power and the preset charging power. At the same time, if the grid load changes during the charging period set by the user and approaches or exceeds the predetermined load limit, the system may appropriately reduce the charging power to ensure the stable operation of the grid, but still try to meet the user's demand for completing charging within the specified time. For example, if the user sets to charge during the low valley electricity price period at night, the system will start charging during this period and fine-tune the charging power according to the real-time grid load situation.

[0059] In other cases, the intelligent mode is selected. Based on the load prediction result, the low load period and the high load period are obtained and combined with the real-time electricity price to automatically adjust the charging period or the charging power. For example, during the peak period with a higher electricity price during the day, the charging power is reduced or charging is paused, and then the charging power is increased for charging when the electricity price is lower at night, ensuring that while meeting the user's charging demand, the user's charging cost is reduced and the pressure on the grid during the peak period is alleviated.

[0060] In the intelligent mode, it is also possible to: dynamically adjust according to the current SOC of the user's battery, combined with the charging rule (such as the optimal power range for charging per hour). For users with a low SOC, the charging pile will accelerate charging; for users with a higher SOC, the charging pile will slow down the charging power to avoid wasting electricity. For example, when the battery SOC is low, the system determines that the vehicle urgently needs to replenish power, and will appropriately increase the charging power, but at the same time will ensure that it will not cause too much impact on the grid in combination with the grid load situation; when the SOC is close to the full state, the charging power is gradually reduced to protect the battery and avoid overcharging.

[0061] In the intelligent mode, it is also possible to: if there are conflicts in the charging requests of multiple users (for example, multiple users choose to charge at the same time), the system will give priority to the users who choose an earlier charging time according to the first-come, first-served principle, and at the same time, in combination with the grid load situation, allocate charging resources at a lower grid load through the time priority strategy. For example, during the peak period, if multiple users request charging, the system will queue up according to the order in which the charging requests are submitted by the users, and then, according to the grid load prediction, give priority to allocating a higher charging power to the users who are at the front of the queue and can charge during the relatively low grid load period, ensuring that each user can complete charging within a reasonable time and maintaining the stable operation of the grid.

[0062] During the charging process, the load prediction results are obtained in real time to determine whether there is a sudden increase or abnormality in the load. If it is determined that there is, the charging power of the device being charged is adjusted or the charging is stopped. For example, during the charging process, if the grid load suddenly rises above the safety threshold, the system will reduce the charging power of the vehicle being charged. If the load continues to rise, the charging of some vehicles may be suspended. After the grid load returns to normal, the charging power will be reallocated and the charging will be resumed according to factors such as the queuing situation and the user's battery SOC.

[0063] In some embodiments, the intelligent control system for orderly charging in the substation area may include multiple functional modules composed of computer program segments. The computer programs of each program segment in the intelligent control system for orderly charging in the substation area can be stored in the memory of the computer device and executed by at least one processor to execute (see Figure 1 description) the functions of intelligent control for orderly charging in the substation area.

[0064] In this embodiment, according to the functions it performs, the intelligent control system for orderly charging in the substation area can be divided into multiple functional modules, as Figure 5 shown. The functional modules of the system may include: a connection establishment module, an information acquisition module, a model construction module, and a strategy adjustment module. The module referred to in the present invention means a series of computer program segments that can be executed by at least one processor and can complete fixed functions, and are stored in the memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0065] The connection establishment module establishes a two-way communication connection between the charging pile and the substation area user collection terminal based on the power line carrier communication technology and protocol conversion technology; The information acquisition module, after establishing the connection, the substation area user collection terminal acquires charging information, user requirements, and real-time electricity prices. The charging information includes the number of charging times, charging frequency, SOC, charging habits, and historical grid load data; The model construction module constructs an MLP fusion model and fuses multi-dimensional data in the charging information into a fusion feature vector based on the MLP fusion model, and trains a CNN-LSTM load prediction model based on the fusion feature vector; The strategy adjustment module obtains the load prediction result based on the CNN-LSTM prediction model combined with real-time charging information, selects a charging mode based on the predicted load result, user requirements, and real-time electricity price, and dynamically adjusts the charging strategy based on the charging mode. The charging modes include the immediate charging mode, the timed charging mode, and the intelligent charging mode.

[0066] The connection establishment module utilizes power line carrier communication and protocol conversion technologies to reduce communication costs and improve compatibility; the information acquisition module provides rich data for the system to facilitate accurate decision-making; the model construction module enhances the accuracy of load forecasting; the strategy adjustment module selects the charging mode according to various factors and dynamically adjusts the strategy, improving the user experience, optimizing the grid operation efficiency, reducing costs, enhancing system stability, and promoting the intelligent and efficient development of electric vehicle charging management.

[0067] Figure 6 FIG. 4 is a schematic structural diagram of a terminal provided by an embodiment of the present invention, and the terminal can be used to execute the method for intelligent control of orderly charging in a substation area provided by the embodiment of the present invention.

[0068] Among them, the terminal may include: a processor, a memory, and a communication unit. These components communicate through one or more buses. Those skilled in the art can understand that the structure of the server shown in the figure does not constitute a limitation to the present invention. It can be a bus structure, a star structure, and may also include more or fewer components than shown in the figure, or combine some components, or have different component arrangements.

[0069] Among them, the memory can be used to store the execution instructions of the processor. The memory can be implemented by any type of volatile or non-volatile storage terminal or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. When the execution instructions in the memory are executed by the processor, the terminal can execute some or all of the steps in the above method embodiments.

[0070] The processor is the control center of the storage terminal. It connects various parts of the entire electronic terminal through various interfaces and lines, and executes various functions and / or processes data of the electronic terminal by running or executing software programs and / or modules stored in the memory, and calling the data stored in the memory. The processor can be composed of an integrated circuit (IC). For example, it can be composed of a single packaged IC, or can be composed of multiple packaged ICs with the same or different functions connected together. For example, the processor may only include a central processing unit (CPU). In the embodiment of the present invention, the CPU can be a single operation core or can include multiple operation cores.

[0071] A communication unit for establishing a communication channel so that the storage terminal can communicate with other terminals. It receives user data sent by other terminals or sends user data to other terminals.

[0072] The present invention also provides a computer-readable storage medium. The computer storage medium can store a program, and when the program is executed, it may include some or all of the steps in the embodiments provided by the present invention. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), a random access memory (RAM), or the like.

[0073] Therefore, the technical effects achievable in this embodiment can be referred to the descriptions above and will not be elaborated here.

[0074] Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be realized by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solutions in the embodiments of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product is stored in a storage medium such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, etc., which can store program codes, and includes several instructions to enable a computer terminal (which can be a personal computer, a server, or a second terminal, a network terminal, etc.) to execute all or part of the steps of the methods described in the embodiments of the present invention.

[0075] For the same or similar parts among the various embodiments in this specification, reference can be made to each other. In particular, for the terminal embodiments, since they are basically similar to the method embodiments, the descriptions are relatively simple, and the relevant parts can refer to the descriptions in the method embodiments.

[0076] In several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the systems or modules can be in electrical, mechanical, or other forms.

[0077] The module described as a separation component may or may not be physically separated. The component shown as a module may or may not be a physical module, that is, it may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0078] In addition, in each embodiment of the present invention, each functional module can be integrated into a processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0079] Although the present invention has been described in detail by referring to the drawings and in combination with the preferred embodiments, the present invention is not limited thereto. Without departing from the spirit and essence of the present invention, those of ordinary skill in the art can make various equivalent modifications or substitutions to the embodiments of the present invention, and these modifications or substitutions should all be within the scope of the present invention. / Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, and all should be covered within the protection scope of the present invention.

Claims

1. A method for intelligent control of orderly charging in a substation, characterized in that: The following steps are involved: S1, establish a two-way communication connection between the charging pile and the user terminal in the substation based on power line carrier communication technology and protocol conversion technology; S2, after the connection is established, the substation uses the terminal to obtain charging information, user demand and real-time electricity prices. The charging information includes the number of charging vehicles, charging frequency, SOC, charging habits and historical grid load data; S3, constructing an MLP fusion model and fusing the multi-dimensional data in the charging information into a fusion feature vector based on the MLP fusion model, and training a CNN-LSTM load forecasting model based on the fusion feature vector; S4, based on the CNN-LSTM prediction model combined with real-time charging information, obtains load forecast results, selects charging mode based on predicted load results, user demand and real-time electricity price, and dynamically adjusts charging strategy based on charging mode. Charging modes include instant charging mode, timing mode and smart charging mode.

2. The method for intelligent control of orderly charging in a substation according to claim 1, characterized in that: Step S1 specifically includes: Convert the data sent by the charging pile from the CAN protocol to the Ethernet protocol, or convert the data sent by the user terminal in the area from the Ethernet protocol to the CAN protocol; Based on modulation technology, the data sent by the charging pile or the area user terminal is converted into a high-frequency signal suitable for transmission through the power line; Amplifying the high frequency signal based on a gain mechanism and transmitting the amplified high frequency signal through a power line; receiving the amplified high-frequency signal transmitted through the power line, and demodulating the high-frequency signal into the original low-frequency digital signal based on a demodulation algorithm; The area terminal or charging pile receives the low-frequency digital signal to complete the two-way communication connection between the charging pile and the area terminal.

3. The method for intelligent control of orderly charging in a substation according to claim 2, characterized in that: Converting the data sent by the charging pile from the CAN protocol to the Ethernet protocol specifically includes: When the charging pile sends data, the data is parsed into a CAN data frame based on the CAN protocol specification. The data frame includes a start bit, an arbitration field, a control field, a data field, a CRC check field, and an end bit. Extract key information from the parsed CAN data field and temporarily store the information in the memory buffer; Encapsulate the key information into an Ethernet protocol frame in the format of the Ethernet protocol. The Ethernet protocol frame includes the destination address of the acquisition terminal in the station area, the conversion source address, the Ethernet type, the data information and the frame check sequence; Send the encapsulated Ethernet frame through the Ethernet interface.

4. The method for intelligent control of orderly charging in a substation according to claim 2, characterized in that: Converting the data sent by the terminal from Ethernet protocol to CAN protocol specifically includes: When the area user terminal sends data, the Ethernet data frame is parsed based on the Ethernet protocol specification. The Ethernet data frame includes the target charging pile address, the area user terminal source address, the data type and the data content; Parse relevant parameters according to the control instructions in the data content, and temporarily store the relevant parameters in the memory buffer; The control instructions and related parameters are encapsulated into a CAN protocol frame according to the format of the CAN protocol. The CAN protocol frame includes a start bit, an arbitration field, a control field, a data field, a CRC check field and an end bit; Send the encapsulated CAN protocol data to the CAN bus.

5. The method for intelligent control of orderly charging in a substation according to claim 1, characterized in that: In step S3, constructing the MLP fusion model specifically includes: A data set is constructed based on historical charging information and the corresponding historical fusion multidimensional vector, and the data set is divided into a training set, a validation set, and a test set; Determine the number of input layer neurons based on the number of multidimensional data features in the basic information, set the hidden layer, determine the output layer based on the output fusion feature vector, and build the initial MLP fusion model. The input vector is the historical charging information, and the output vector is the historical fusion multidimensional vector. For each sample in the training set, the input feature vector is fed into the initial MLP fusion model, and the predicted output vector of the model is calculated through forward propagation; The loss value between the predicted output vector and the actual output feature vector is calculated according to the loss function, and the parameters of the initial MLP fusion model are optimized using the back propagation algorithm; The preset number of training times per interval is used to calculate the evaluation index based on the validation set combined with the root mean square error function, and the model structure is adjusted based on the changing trend of the evaluation index; The trained initial MLP fusion model is finally evaluated based on the test set to obtain the final MLP fusion model.

6. The method for intelligent control of orderly charging in a substation according to claim 1, characterized in that: In step S3, training the CNN-LSTM load forecasting model specifically includes: Construct a data set based on historical fusion feature vectors and historical load results, and divide the data set into a training set, a validation set, and a test set; Determine the parameters of the convolutional layer, including the number of convolutional layers, convolution kernel size, and step size parameters; Configure the pooling layer based on the maximum pooling method, and reduce the dimension of the features output by the convolutional layer by setting the pooling window size; Determine the number of LSTM layers and the number of neurons in each LSTM layer; The output layer is configured based on the dimension of the load forecasting result to obtain the initial CNN-LSTM load forecasting model; the historical fusion feature vector is the input vector of the model, and the historical load result is the output vector of the model; For each sample in the training set, the input feature vector is input into the CNN-LSTM load forecasting model for forward propagation calculation. In the convolution layer, the convolution kernel slides on the historical fusion feature vector to perform convolution operation and extract feature data. After the pooling layer, the feature data is input into the LSTM layer to learn the time series pattern in the data. Finally, the load forecast result is output through the output layer. The loss value between the predicted output vector and the actual output feature vector is calculated according to the loss function, and the parameters of the initial CNN-LSTM load forecasting model are optimized using the back propagation algorithm; The preset number of training times per interval is used to calculate the evaluation index based on the validation set combined with the root mean square error function, and the model structure is adjusted based on the changing trend of the evaluation index; The trained initial CNN-LSTM load forecasting model is finally evaluated based on the test set to obtain the final CNN-LSTM load forecasting model.

7. The method for intelligent control of orderly charging in a substation according to claim 1, characterized in that: Step S4 specifically includes: When it is determined that the load forecast result is abnormal, charging is not performed; When it is determined that the load forecast result is lower than the load threshold, there is no special time requirement in the user demand, and the tiered electricity price is lower than the electricity price threshold, the instant charging mode is selected, and the charging pile immediately starts charging the vehicle at the preset high power; When it is determined that the load forecast result fluctuation is less than the fluctuation threshold and the user demand has a time requirement, the timing mode is selected, the user sets the charging time, and the timing charging power is configured based on the charging time. The charging pile performs charging operations based on the charging time and the timing charging power; In other cases, select the smart mode to obtain low-load and high-load periods based on load forecast results and combine with real-time electricity prices to automatically adjust the charging period or charging power; During the charging process, load forecast results are obtained in real time to determine whether there is a sudden increase or abnormality in the load. If so, the charging power of the device being charged is adjusted or charging is stopped.

8. A method for intelligent control of orderly charging in a substation, characterized in that: When the system is implemented, the method for intelligent control of orderly charging in a substation area according to any one of claims 1 to 7 is executed, and the system includes: The connection establishment module establishes a two-way communication connection between the charging pile and the user terminal in the substation based on the power line carrier communication technology and protocol conversion technology; Information acquisition module: After the connection is established, the substation uses the terminal to obtain charging information, user demand and real-time electricity prices. The charging information includes the number of charging vehicles, charging frequency, SOC, charging habits and historical grid load data; Model building module, constructing an MLP fusion model and fusing the multi-dimensional data in the charging information into a fusion feature vector based on the MLP fusion model, and training the CNN-LSTM load forecasting model based on the fusion feature vector; The strategy adjustment module obtains load forecast results based on the CNN-LSTM prediction model combined with real-time charging information, selects the charging mode based on the predicted load results, user demand and real-time electricity price, and dynamically adjusts the charging strategy based on the charging mode. The charging modes include instant charging mode, timing mode and smart charging mode.

9. A terminal, characterized in that: include: A memory for storing an intelligent control program for orderly charging in a substation; A processor is used to implement the steps of the method for intelligent control of orderly charging in an area as described in any one of claims 1 to 7 when executing the intelligent control program for orderly charging in an area.

10. A computer-readable storage medium, characterized in that: The readable storage medium stores a smart control program for orderly charging in a substation, and when the smart control program for orderly charging in a substation is executed by a processor, the steps of the smart control method for orderly charging in a substation are implemented as described in any one of claims 1 to 7.

Images