Methods, devices, electronic equipment and storage media for coordinated control of charging pile groups
By using a long short-term memory network deep learning model and a priority weight adjustment strategy, the resource allocation problem of charging pile groups under the conditions of grid load fluctuation and unstable user demand was solved, thereby improving grid stability and user experience.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGZHOU EVERBRIGHT POWER ENG CO LTD
- Filing Date
- 2026-01-15
- Publication Date
- 2026-07-31
AI Technical Summary
The existing charging pile network cannot effectively cope with sudden increases in user demand or fluctuations in grid load, resulting in disorderly charging that impacts the grid. Furthermore, when user demand and grid load are unstable, the allocation of charging resources cannot meet diverse needs.
A load prediction model based on a long short-term memory network deep learning structure is adopted, combined with charging situation awareness data, to dynamically determine the optimization target, and adjust the charging strategy by calculating priority weights to achieve reasonable allocation of charging resources.
Effectively address grid load fluctuations, improve grid stability and charging service experience, ensure priority is given to important or urgent charging needs, optimize resource allocation, and avoid the impact of disorderly charging on the grid.
Smart Images
Figure CN121799227B_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of charging management technology, and in particular to a method, device, electronic device and storage medium for coordinated control of charging pile groups. Background Technology
[0002] With the increasing popularity of electric vehicles, the demand for charging stations is growing rapidly. The surge in the number of charging stations can lead to uncontrolled charging and significant disruptions to local power grids. Meanwhile, users have diverse needs regarding charging speed and service experience. Currently, in large public charging stations or charging clusters, when multiple charging stations operate simultaneously, the allocation of charging resources is typically based on simple rules or algorithms. This method is effective when user demand and grid load are relatively stable, but it may not be effective in handling sudden increases in user demand or fluctuations in grid load. Summary of the Invention
[0003] To overcome the shortcomings of existing technologies, this invention provides a method, device, electronic equipment, and storage medium for coordinated control of charging pile groups. It comprehensively considers user needs and grid load, effectively responds to sudden increases in user demand and fluctuations in grid load, achieves reasonable allocation of charging resources, and improves the charging service experience and grid stability.
[0004] A first aspect of this application provides a method for coordinated control of a group of charging piles, the method comprising: Acquire charging status awareness data of the target charging pile group; The charging situation awareness data is input into a preset load forecasting model so that the load forecasting model outputs the future power grid area load forecast result; the load forecasting model is constructed based on a long short-term memory network deep learning structure. Based on the future power grid area load forecast results and the charging situation awareness data, the optimization target of the target charging pile group is dynamically determined; The priority weight of each charging request that satisfies the optimization objective is calculated based on the charging situation awareness data. The charging strategy of the target charging pile group is dynamically adjusted according to the priority weight.
[0005] In an optional implementation, calculating the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data includes: Key features are extracted from the charging situation awareness data and normalized; the key features include user charging demand features, electric vehicle status features, and power grid operation features. Construct a quantitative scoring function for each of the key features, and dynamically adjust the weights of the quantitative scoring functions according to the optimization objective; The initial score for each feature is calculated using the quantization scoring function based on the adjusted weights. The real-time feedback correction item is determined based on the real-time operating status of the power grid, and the priority weight is determined based on the initial score and the real-time feedback correction item.
[0006] In an optional implementation, the load prediction model includes two LSTM units: a first LSTM unit for extracting time-series features from the charging situation awareness data, and a second LSTM unit for capturing long-term dependencies in the charging situation awareness data.
[0007] In one optional implementation, the optimization objective includes at least one of minimizing peak grid load, minimizing user charging costs, and maximizing charging pile utilization.
[0008] In an optional implementation, the charging situation awareness data further includes grid load data, and the step of controlling the dynamic adjustment of the charging strategy for the target charging pile group according to the priority weight includes: Based on the power grid load data, determine whether the target charging pile group is currently in a peak power grid load period or a low power grid load period; When it is determined that the target charging pile group is in a peak period of grid load, the charging power of low-priority vehicles is reduced according to the priority weight; or the low-priority vehicles are added to the waiting queue, and the charging of vehicles in the waiting queue is restored in order of priority weight, so that high-priority vehicles can complete charging within the expected charging time; the low-priority vehicles are vehicles with priority weights lower than a preset threshold, and the high-priority vehicles are vehicles with priority weights higher than the preset threshold. When it is determined that the target charging pile group is in a period of low grid load, the charging power of all charging piles is increased in order of priority weight.
[0009] In an optional implementation, the charging situation awareness data further includes charging pile status data, and the method further includes: Determine whether the total power of the charging pile group exceeds the difference between the total capacity of the transformer in the area and the peak power consumption of other loads; the charging pile status data includes the total power of the charging pile group and vehicle SOC information; When it is determined that the difference exceeds the stated value, obtain the charging right purchase and sale prices offered by all users; The charging order sequence for each user among all users is determined based on the charging right purchase and sale price and the vehicle SOC information. The charging status of the target charging pile group is controlled according to the charging sequence.
[0010] In an optional implementation, the charging right purchase price includes a charging sales price and a charging purchase price, and determining the charging sequence for each user among all users based on the charging right purchase price and the vehicle SOC information includes: Based on the charging selling price, all users are sorted from low to high to obtain a selling user price sorting sequence; and based on the charging purchase price, all users are sorted from high to low to obtain a purchasing user price sorting sequence. The selling user price sorting sequence and the purchasing user price sorting sequence are used as the first sorting result; Based on the SOC gap, users whose charging selling price and charging purchase price are the same among all users are sorted from low to high to obtain a second sorting result; the SOC gap is determined based on the vehicle SOC information. Based on the time of access to the charging pile, users with the same SOC difference among all users are sorted from earliest to latest to obtain the third sorting result; The first sorting result, the second sorting result, and the third sorting result are merged to generate the charging sequence.
[0011] A second aspect of this application provides a charging pile group collaborative control device, the device comprising: The acquisition module is used to acquire charging status awareness data of the target charging pile group; The prediction module is used to input the charging situation awareness data into a preset load prediction model so that the load prediction model outputs the future power grid area load prediction results; the load prediction model is constructed based on a long short-term memory network deep learning structure. The determination module is used to dynamically determine the optimization target of the target charging pile group based on the future power grid area load forecast results and the charging situation perception data; The calculation module is used to calculate the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data. The adjustment module is used to control the target charging pile group to dynamically adjust the charging strategy according to the priority weight.
[0012] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the charging pile group collaborative control method.
[0013] A fourth aspect of this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-described charging pile group collaborative control method.
[0014] In summary, the charging pile group collaborative control method, device, electronic equipment, and storage medium provided in this application have at least one of the following beneficial effects: 1. The acquired charging situation awareness data is input into a pre-built load forecasting model constructed based on a long short-term memory (LSTM) deep learning structure. The model outputs future regional power grid load forecasts. The LTM deep learning structure can handle data with time-series characteristics and has good predictive capabilities for complex data such as power grid load that changes over time. By predicting future regional power grid load, the fluctuation of power grid load can be known in advance, thus providing a basis for addressing power grid load fluctuation issues.
[0015] 2. Based on future regional power grid load forecasts and user-set expected charging data, the optimization target for the target charging pile group is dynamically determined. By comprehensively considering both power grid load and user demand, the optimization target can adapt to different situations, taking into account both the power grid's capacity and user needs for charging speed and service experience. This solves the problem that existing methods cannot effectively address unstable user demand and power grid load conditions.
[0016] 3. By calculating priority weights, different charging requests are distinguished and sorted, providing a quantitative basis for the subsequent reasonable allocation of charging resources. This helps to prioritize important or urgent charging needs when resources are limited, and further optimize the allocation of charging resources.
[0017] 4. The charging strategy is dynamically adjusted based on priority weights, which can flexibly allocate charging resources according to real-time conditions. When faced with sudden increases in user demand or fluctuations in grid load, the charging schedule can be adjusted in a timely manner to avoid disorderly charging from impacting the grid. At the same time, it better meets the diverse needs of users and solves the problem that existing methods cannot effectively cope with complex situations. Attached Figure Description
[0018] Figure 1 This is a flowchart illustrating a collaborative control method for a group of charging piles, as shown in an embodiment of this application. Figure 2 This is a functional block diagram of a charging pile group collaborative control device shown in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device shown in an embodiment of this application. Detailed Implementation
[0019] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0020] The following will clearly and completely describe the concept, specific structure, and technical effects of the present invention in conjunction with embodiments and accompanying drawings, so as to fully understand the purpose, features, and effects of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are all within the scope of protection of the present invention. Furthermore, all connections / linkages involved in the patent do not simply refer to direct contact between components, but rather to the ability to form a better connection structure by adding or reducing connecting accessories according to specific implementation conditions. The various technical features in this invention can be combined interactively without contradicting each other.
[0021] Reference Figure 1 The diagram shown is a flowchart illustrating a collaborative control method for a charging pile group according to an embodiment of this application. The collaborative control method for a charging pile group includes the following steps.
[0022] S11, acquire charging status awareness data of the target charging pile group.
[0023] The target charging pile group refers to the total number of charging piles within a certain area (the area divided by the power distribution network).
[0024] In some embodiments, smart sensors installed on charging piles, data interfaces connected to the power grid system, and vehicle-mounted terminals communicating with the vehicle can be used to acquire real-time charging situational awareness data, including but not limited to: power grid load data, charging pile status data, vehicle battery information (SOC), and user-set expected charging data, whereby the expected charging data includes expected charging time and cost preference data. Specifically, through data interaction with the power grid dispatch center, real-time power, voltage, frequency, and other parameters of the current power grid, as well as historical load data for different time periods, are obtained as power grid load data. For example, the total load value of the current power grid is acquired every 5 minutes, and load records for each time point in the past 24 hours are also acquired. Simultaneously, the working status of the charging piles is monitored in real time through smart sensors to obtain charging pile status data, such as whether charging is in progress, charging power, charging duration, and charging pile fault information. For example, when the charging pile starts charging the vehicle, the charging start time is recorded; during the charging process, the current charging power is recorded every minute. Vehicle battery information (SOC) is also acquired through communication between the vehicle-mounted terminal and the vehicle battery management system (BMS). For example, the vehicle's SOC value is acquired every 10 seconds. In addition, when users make a charging reservation through the charging station's operating interface or mobile app, they can set their expected full charge time and cost preferences (such as choosing to charge during low-price periods, accepting a certain degree of extended charging time to save money, etc.).
[0025] By comprehensively collecting data from the power grid, charging piles, vehicles, and users, this method fully considers various factors that influence charging decisions, providing rich basic information for subsequent precise control and optimization. Compared with traditional charging control methods that only consider partial data, it has higher accuracy and adaptability.
[0026] In some embodiments, after acquiring charging situation awareness data, the data can be preprocessed. For example, data cleaning can be performed to remove outliers (such as negative charging power or exceeding the charging pile's rated value), and missing data can be filled using linear interpolation.
[0027] S12, the charging situation awareness data is input into a preset load forecasting model so that the load forecasting model outputs the future power grid area load forecasting results.
[0028] In some embodiments, the load prediction model is constructed based on a Long Short-Term Memory (LSTM) deep learning architecture, specifically a stacked LSTM network with two layers. The first LSTM layer extracts time-series features from the charging situation awareness data to capture short-term dynamic changes. The second LSTM layer further captures long-term dependencies in the charging situation awareness data, enhancing the ability to model periodic patterns. To improve training efficiency and stability, a batch normalization layer is introduced after each LSTM layer to accelerate model convergence and improve training stability. A fully connected layer is connected after the second LSTM layer to map high-dimensional features to low-dimensional load prediction results. Furthermore, to enhance the model's sensitivity to key time steps, an attention mechanism is introduced to weight and fuse the hidden states of the LSTM output, giving higher weight to key information such as load spikes or peak periods in the prediction.
[0029] The load forecasting model employs a composite loss function, combining mean squared error (MSE) and symmetric average absolute percentage error (SMAPE) to simultaneously optimize the absolute and relative errors of the forecast values. Its expression is as follows: ; Where y is the actual load value, To predict load values, and For the weighting coefficients, satisfying .
[0030] MSE is used to penalize the absolute deviation between the predicted and the true values, and is defined as: ; MAPE is used to balance relative errors under different dimensions, and is defined as follows: .
[0031] To improve the model's generalization ability, a dynamic adjustment strategy is employed during the training optimization phase. First, the Adam optimizer is used to dynamically adjust the learning rate, initially set to 0.001. During training, a learning rate decay strategy is used, decreasing the rate to 0.9 times its original value every 10 epochs. Simultaneously, an early stopping mechanism is introduced: training is terminated to prevent overfitting if the validation set loss does not decrease for five consecutive epochs. When the validation set loss does not decrease for five consecutive epochs, training automatically terminates and the optimal model parameters are preserved.
[0032] By stacking LSTM layers to enhance feature extraction capabilities and introducing an attention mechanism to dynamically focus on key time steps, the problem of information loss in long sequence prediction using traditional LSTM is solved. Furthermore, by combining MSE and SMAPE, the optimization of both absolute and relative errors is achieved, improving the model's prediction accuracy under different load dimensions. In addition, the Adam optimizer and learning rate decay strategy are used to accelerate convergence, while an early stopping mechanism is employed to enhance the model's generalization ability.
[0033] By analyzing the grid load, charging pile usage, and vehicle charging demand at different times of the day over the past week, the trend of total grid load change at each point in time over the next 24 hours is predicted with a resolution of 15 minutes per point.
[0034] Once the load forecasting model is trained, the preprocessed charging situation awareness data is input into the load forecasting model, which can then output the power grid area load forecast results for each time point in the next 24 hours.
[0035] S13. Based on the future power grid area load forecast results and the charging situation awareness data, dynamically determine the optimization target of the target charging pile group.
[0036] The optimization objectives include at least one of minimizing peak grid load, minimizing user charging costs, and maximizing charging pile utilization. Minimizing peak grid load is used to suppress charging demand during peak hours and guide users to charge during off-peak periods; minimizing user charging costs is used to prioritize charging during periods of lower electricity prices and reduce peak electricity costs; maximizing charging pile utilization is used to evenly distribute charging power and avoid some charging piles being idle or overloaded.
[0037] In some embodiments, load thresholds are set to divide different load periods based on future grid area load forecasts. For example, periods with loads above a certain higher threshold are classified as peak load periods, periods with loads below a certain lower threshold are classified as off-peak load periods, and the remaining periods are classified as normal load periods. When it is predicted that the grid will enter a peak load period, the primary optimization objective is set to minimize the peak grid load.
[0038] When the grid is predicted to be in a period of low load, the optimization objective focuses on maximizing the utilization rate of charging stations. By improving the efficiency of charging station utilization, operators' revenue is increased, while also helping to balance the grid load.
[0039] Simultaneously, the system analyzes user-defined expected charging times and cost preferences from the charging situation awareness data. If most users want to complete charging within a specific timeframe and are cost-sensitive, the optimization objective is further adjusted to minimize user charging costs while ensuring grid safety. Specifically, each user is assigned a weight based on their expected charging time and cost preferences. Users who are cost-sensitive and want to complete charging within a specific timeframe are given a higher weight. During the optimization process, priority is given to meeting the charging needs of these users. By rationally arranging the charging sequence and power, charging during lower-price periods is prioritized while meeting users' expected charging times, thereby improving user satisfaction.
[0040] If the current problem requires optimization with multiple objectives, the electronic device can be transformed into a single-objective optimization problem by weighted summation.
[0041] It should be noted that the optimization target is not fixed, but dynamically adjusted according to the real-time operation of the power grid and user needs, thereby achieving a balance between the interests of the power grid and users and improving the flexibility and adaptability of the collaborative control of charging pile groups.
[0042] S14, calculate the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data.
[0043] Once the charging situation awareness data and optimization goals are determined, electronic devices can adopt multi-dimensional dynamic priority scoring. At the same time, a dynamic weight adjustment mechanism guided by optimization goals and a real-time feedback correction closed loop are introduced, enabling the charging resource allocation strategy to adapt to changes in the grid operation status, ultimately achieving a leap from "passive response" to "active optimization".
[0044] In an optional implementation, calculating the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data includes: Key features are extracted from the charging situation awareness data and normalized; the key features include user charging demand features, electric vehicle status features, and power grid operation features. Construct a quantitative scoring function for each of the key features, and dynamically adjust the weights of the quantitative scoring functions according to the optimization objective; The initial score for each feature is calculated using the quantization scoring function based on the adjusted weights. The real-time feedback correction item is determined based on the real-time operating status of the power grid, and the priority weight is determined based on the initial score and the real-time feedback correction item.
[0045] In some embodiments, the electronic device first extracts the following key features from the charging situation awareness data: User charging demand characteristics: expected charging time window, expected charging amount, preferred charging power, and historical charging periods (such as whether peak hours are frequently selected); Electric vehicle status characteristics: current state of charge and battery health (SOH), etc. Power grid operation characteristics: current electricity price and real-time load, etc.
[0046] All key features are normalized to map the values to the [0, 1] interval, thus eliminating dimensional differences.
[0047] Next, a quantitative scoring function is constructed for each key feature to quantify the contribution of each feature to the priority of charging requests. Specifically, for user charging demand features, weight allocation is based on business priority, where time urgency > charging amount > power preference > historical peak suppression, and the percentage of historical peaks needs to be calculated in reverse (1- To achieve a suppression effect, the quantification function of user charging demand characteristics is: ; in, The time urgency is scored based on how urgency the expected charging time window is; for example, the shorter the time window, the higher the score. The expected charging capacity is scored based on the ratio of expected charging capacity to battery capacity; the higher the expected charging capacity, the higher the score. A power preference score is assigned based on the user's preference for charging power, with a higher score for fast charging preference than for slow charging. The system scores users based on the percentage of peak charging times in their historical charging history. The higher the percentage, the lower the score (to achieve a suppression effect).
[0048] For the state characteristics of electric vehicles, setting a higher SOC weight (e.g., 0.7) directly impacts users' range anxiety, while a SOH weight of 0.3 is used to limit the overload risk of older batteries. Therefore, the quantitative scoring function for the state characteristics of electric vehicles is set as follows: ; in, The current state of charge (SOC) is scored based on the percentage of the current SOC; the higher the SOC, the higher the score. The battery health is rated based on a percentage of battery health; the higher the SOH (State of Health), the higher the score.
[0049] Based on the characteristics of power grid operation, a higher weight is set for electricity price sensitivity (e.g., 0.6) to guide users to stagger peak hours, and a load pressure weight of 0.4 to suppress the risk of system overload. The quantitative scoring function for the characteristics of power grid operation is then set as follows: ; Electricity price sensitivity is scored based on the ratio of the current electricity price to the average electricity price; the lower the electricity price, the higher the score. The real-time load pressure score is based on the ratio of the real-time load to the maximum load capacity of the power grid; the lower the load, the higher the score.
[0050] Furthermore, the electronic device can dynamically adjust the weights of each scoring function according to the optimization objective. The weights are determined as follows: If the optimization objective is to minimize the peak load of the power grid, the weight of the power grid operation characteristic score can be set to be relatively high, such as power grid operation characteristic score weight = 0.5, user charging demand characteristic weight = 0.3, and electric vehicle status characteristic weight = 0.2; if the optimization objective is to minimize the user charging cost, the weight of the user charging demand characteristic can be set to be relatively high, such as user charging demand characteristic weight = 0.5, power grid operation characteristic score weight = 0.3, and electric vehicle status characteristic weight = 0.2; if the optimization objective is to maximize the utilization rate of charging piles, the weight of the electric vehicle status characteristic can be set to be relatively high, such as electric vehicle status characteristic weight = 0.5, power grid operation characteristic score weight = 0.3, and user charging demand characteristic weight = 0.2.
[0051] For each charging request, an initial priority score is calculated based on the scores of its key features and their corresponding weights. For each feature, calculate its weighted score. : ; in, To dynamically adjust the feature weights according to the optimization objective, Feature scoring function Sum the weighted scores of all features to obtain the initial total score. : .
[0052] Finally, the electronic equipment incorporates a real-time feedback correction term based on the real-time state of the power grid. This real-time feedback correction term can be adjusted according to factors such as changes in grid load and electricity price fluctuations. The final priority weight is obtained by multiplying the real-time feedback correction term by the initial total score.
[0053] Through the aforementioned optional implementation methods, multi-dimensional dynamic scoring and optimization-oriented weight adjustments comprehensively consider user needs, vehicle status, and grid operation characteristics to accurately calculate the priority weight of charging requests. Furthermore, the real-time feedback correction mechanism enables the strategy to adapt to grid changes, avoiding the limitations of fixed weights. This achieves a shift in charging resource allocation from "passive response" to "proactive optimization," effectively improving grid stability, user satisfaction, and charging pile utilization.
[0054] S15, dynamically adjust the charging strategy of the target charging pile group according to the priority weight.
[0055] Based on the calculated priority weight of each charging request, electronic devices can coordinately control the charging stations through the charging station cluster controller. The specific charging strategy is as follows: During peak grid load periods: Automatically reduce the charging power of non-emergency charging tasks (those with lower priority weights, e.g., weights below a preset threshold of 0.5) or queue them for waiting. For example, reduce the charging power of low-priority charging requests to half of their original power, or suspend their charging and add them to the waiting queue, resuming charging sequentially according to priority. Prioritize the charging needs of high-priority vehicles (those with higher priority weights, e.g., weights above a preset threshold of 0.5) to ensure they can complete charging within the user's expected timeframe.
[0056] During periods of low grid load: Appropriately increase the charging power of all charging piles to speed up charging. At the same time, based on users' cost preferences, prioritize the allocation of charging resources to users who choose to charge during low-price periods to improve operator revenue and user satisfaction.
[0057] In some implementations, during the charging process, the grid load, charging station status, and vehicle battery information are continuously monitored, and the charging strategy is adjusted in real time according to the actual situation. For example, if the grid load suddenly decreases, the charging power of the vehicle being charged can be appropriately increased; if a charging station malfunctions, the charging request on that charging station is promptly transferred to other available charging stations.
[0058] By dynamically adjusting the charging power and charging sequence, coordinated control of the charging pile group is achieved, effectively shaving peaks and filling valleys, improving the stability of the power grid and the operating efficiency of the charging pile group, while fully considering user needs and optimizing the user experience.
[0059] In an optional implementation, the method further includes: Determine whether the total power of the charging pile group exceeds the difference between the total capacity of the transformer in the area and the peak power consumption of other loads; the charging pile status data includes the total power of the charging pile group and vehicle SOC information; When it is determined that the difference exceeds the stated value, obtain the charging right purchase and sale prices offered by all users; The charging order sequence for each user among all users is determined based on the charging right purchase and sale price and the vehicle SOC information. The charging status of the target charging pile group is controlled according to the charging sequence.
[0060] If the total power of the target charging pile group in a certain area exceeds the difference between the total capacity of the transformer in the area and the peak power consumption of other loads, then the transformer in the community may be overloaded due to all electric vehicles charging at the same time during a certain period of time, which may easily lead to events such as tripping.
[0061] In some embodiments, when it is determined that the total power of the charging pile group exceeds the difference between the total capacity of the transformer in the area and the peak power consumption of other loads, the electronic device can collect the charging right buying and selling prices offered by all users participating in the micromarket transaction, including the charging right selling price. and charging purchase price Where i is the user ID. Next, the price is normalized, mapping it to the [0,1] interval to eliminate dimensional differences: , ; in, The normalized selling price of charging rights. and The minimum and maximum selling prices for all users. The normalized purchase price for charging. and These are the minimum and maximum purchase prices.
[0062] At the same time, obtain the current state of charge of all electric vehicles connected to the charging pile. This information is used to determine the vehicle's State of Charge (SOC) and to calculate its difference from a target SOC (such as 100% when fully charged). : ; Next, regarding Normalization is performed: ; in, For normalized N is the total number of users participating in the sorting.
[0063] Furthermore, electronic devices can extract the selling price declared by the user selling charging rights. (The lower the price, the stronger the willingness to relinquish charging rights), the purchase price declared by users purchasing charging rights. (The higher the price, the more urgent the need for charging), and the difference after normalization. Then, sorting is performed according to hierarchical sorting rules, which specifically include: First-level sorting: Based on the price of buying and selling charging rights, the first sorting result is obtained.
[0064] Selling users: By Sort the users by price from low to high (those with the lowest price are given priority in charging, indirectly supporting other users' charging), and obtain the selling user price sorting sequence. Purchased users: By Sort the users from highest to lowest price (those with higher prices have priority in getting charging rights), and obtain the price ranking sequence of the purchasing users; The selling user price sorting sequence and the purchasing user price sorting sequence are used together as the first sorting result.
[0065] Second-level ranking: When the buy and sell prices are the same, a second ranking result is obtained based on the urgency of SOC.
[0066] For all users with the same price (whether selling or buying), by Sort by low to high (prioritize users with small SOC differences to avoid excessive delays for low-battery vehicles).
[0067] Third-level sorting: When the SOC difference is still the same, sort by access time to obtain the third sorting result.
[0068] right For the same user, they are sorted from earliest to latest according to the time they connect to the charging station (first come, first served).
[0069] Then, the hierarchical sorting results are merged, and the final charging sequence is output. ,in Users with the highest priority (can be charged first). Users with the lowest priority (may be delayed or unable to charge).
[0070] For example, suppose user A: selling price = 0.8 yuan / kWh, ΔSOC = 0.3; user B: selling price = 0.8 yuan / kWh, ΔSOC = 0.5; user C: purchasing price = 1.2 yuan / kWh, ΔSOC = 0.2. Then the ranking result is user C (highest purchasing price), user A (same selling price, but smaller SOC difference), user B (same selling price, but larger SOC difference).
[0071] Through the above optional implementation methods, resource allocation is achieved through charging right trading. Users with lower selling prices or higher purchasing prices have higher priority. When prices are the same, it is necessary to avoid excessive delays for vehicles with low battery levels. Therefore, they are sorted from low to high according to the SOC difference, which quantifies the willingness to trade and the urgency of charging, thus achieving fair, efficient and dynamic allocation of charging resources.
[0072] Reference Figure 2 The diagram shown is a functional block diagram of the charging pile group collaborative control device according to an embodiment of this application.
[0073] In some embodiments, the charging pile group collaborative control device 20 may include multiple functional modules composed of computer program segments. The computer programs of each program segment of the charging pile group collaborative control device 20 may be stored in the memory of an electronic device and executed by at least one processor to perform (see details). Figure 1 (Description) The function of coordinated control of charging pile groups. Based on the functions it performs, it can be divided into multiple functional modules. These functional modules may include: an acquisition module 201, a prediction module 202, a determination module 203, a calculation module 204, and an adjustment module 205. The module referred to in this application is a series of computer program segments that can be executed by at least one processor and perform a fixed function, stored in memory. In this embodiment, the functions of each module will be detailed in subsequent embodiments.
[0074] The acquisition module 201 is used to acquire charging situation awareness data of the target charging pile group; the charging situation awareness data includes the expected charging data set by the user.
[0075] The prediction module 202 is used to input the charging situation perception data into a preset load prediction model so that the load prediction model outputs the future power grid area load prediction results; the load prediction model is constructed based on a long short-term memory network deep learning structure.
[0076] The determining module 203 is used to dynamically determine the optimization target of the target charging pile group based on the future power grid area load forecast results and the expected charging data.
[0077] The calculation module 204 is used to calculate the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data.
[0078] The adjustment module 205 is used to control the target charging pile group to dynamically adjust the charging strategy according to the priority weight.
[0079] It should be understood that the various variations and specific embodiments of the charging pile group collaborative control method provided in the above embodiments are also applicable to the charging pile group collaborative control device of this embodiment. Through the foregoing detailed description of the charging pile group collaborative control method, those skilled in the art can clearly understand the implementation method of the charging pile group collaborative control device in this embodiment. For the sake of brevity, it will not be described in detail here.
[0080] See Figure 3The diagram shown is a schematic representation of the structure of an electronic device according to an embodiment of this application. In a preferred embodiment of this application, the electronic device 3 includes a memory 31, at least one processor 32, and at least one communication bus 33.
[0081] Those skilled in the art should understand that Figure 3 The structure of the electronic device shown does not constitute a limitation of the embodiments of this application. It can be a bus structure or a star structure. The electronic device 3 may also include more or fewer other hardware or software than shown, or different component arrangements.
[0082] In some embodiments, the electronic device 3 is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits (ASICs), programmable gate arrays (FPGAs), digital processors, and embedded devices. The electronic device 3 may also include user equipment, which includes, but is not limited to, any electronic product capable of human-computer interaction with a user via a keyboard, mouse, remote control, touchpad, or voice control device, such as a personal computer, tablet computer, smartphone, or digital camera.
[0083] In the embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, computer-readable storage media, and electronic devices can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple components or modules may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices, components, or modules may be electrical, mechanical, or other forms.
[0084] The components described as separate parts may or may not be physically separate. The components shown as components may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the components can be selected to achieve the purpose of this embodiment according to actual needs.
[0085] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each component can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.
[0086] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: USB flash drive, portable hard drive, read-only memory (ROM). Various media that can store program code, such as only memory, random access memory (RAM), magnetic disks or optical disks.
[0087] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.
[0088] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.
[0089] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the embodiments described. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of this application.
Claims
1. A method for coordinated control of charging pile groups, characterized in that, The method includes: Acquire charging status awareness data of the target charging pile group; The charging situation awareness data is input into a preset load forecasting model so that the load forecasting model outputs the future power grid area load forecast result; the load forecasting model is constructed based on a long short-term memory network deep learning structure. Based on the future power grid area load forecast results and the charging situation awareness data, the optimization target of the target charging pile group is dynamically determined; The priority weight of each charging request that satisfies the optimization objective is calculated based on the charging situation awareness data. The charging strategy of the target charging pile group is dynamically adjusted according to the priority weight; The step of calculating the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data includes: Key features are extracted from the charging situation awareness data and normalized; the key features include user charging demand features, electric vehicle status features, and power grid operation features. Construct a quantitative scoring function for each of the key features, and dynamically adjust the weights of the quantitative scoring functions according to the optimization objective; The initial score for each feature is calculated using the quantization scoring function based on the adjusted weights. The real-time feedback correction item is determined based on the real-time operating status of the power grid, and the priority weight is determined based on the initial score and the real-time feedback correction item.
2. The charging pile group collaborative control method according to claim 1, characterized in that, The load prediction model includes two LSTM units. The first LSTM unit is used to extract the time series features of the charging situation awareness data, and the second LSTM unit is used to capture the long-term dependencies in the charging situation awareness data.
3. The charging pile group collaborative control method according to claim 1, characterized in that, The optimization objectives include at least one of minimizing peak grid load, minimizing user charging costs, and maximizing charging pile utilization.
4. The charging pile group collaborative control method according to claim 1, characterized in that, The charging situation awareness data also includes grid load data, and the step of controlling the dynamic adjustment of the charging strategy for the target charging pile group according to the priority weight includes: Based on the power grid load data, determine whether the target charging pile group is currently in a peak power grid load period or a low power grid load period; When it is determined that the target charging pile group is in a peak period of grid load, the charging power of low-priority vehicles is reduced according to the priority weight; or the low-priority vehicles are added to the waiting queue, and the charging of vehicles in the waiting queue is restored in order of priority weight, so that high-priority vehicles can complete charging within the expected charging time; the low-priority vehicles are vehicles with priority weights lower than a preset threshold, and the high-priority vehicles are vehicles with priority weights higher than the preset threshold. When it is determined that the target charging pile group is in a period of low grid load, the charging power of all charging piles is increased in order of priority weight.
5. The charging pile group collaborative control method according to claim 1, characterized in that, The charging situation awareness data also includes charging pile status data, and the method further includes: Determine whether the total power of the charging pile group exceeds the difference between the total capacity of the transformer in the area and the peak power consumption of other loads; the charging pile status data includes the total power of the charging pile group and vehicle SOC information; When it is determined that the difference exceeds the stated value, obtain the charging right purchase and sale prices offered by all users; The charging order sequence for each user among all users is determined based on the charging right purchase and sale price and the vehicle SOC information. The charging status of the target charging pile group is controlled according to the charging sequence.
6. The charging pile group collaborative control method according to claim 5, characterized in that, The charging right trading price includes the charging sales price and the charging purchase price. Determining the charging sequence for each user based on the charging right trading price and the vehicle SOC information includes: Based on the charging selling price, all users are sorted from low to high to obtain a selling user price sorting sequence; and based on the charging purchase price, all users are sorted from high to low to obtain a purchasing user price sorting sequence. The selling user price sorting sequence and the purchasing user price sorting sequence are used as the first sorting result; Based on the SOC gap, users whose charging selling price and charging purchase price are the same among all users are sorted from low to high to obtain a second sorting result; the SOC gap is determined based on the vehicle SOC information. Based on the time of access to the charging pile, users with the same SOC difference among all users are sorted from earliest to latest to obtain the third sorting result; The first sorting result, the second sorting result, and the third sorting result are merged to generate the charging sequence.
7. A collaborative control device for a group of charging piles, characterized in that, The device includes: The acquisition module is used to acquire charging status awareness data of the target charging pile group; The prediction module is used to input the charging situation awareness data into a preset load prediction model so that the load prediction model outputs the future power grid area load prediction result; the load prediction model is constructed based on a long short-term memory network deep learning structure. The determination module is used to dynamically determine the optimization target of the target charging pile group based on the future power grid area load forecast results and the charging situation perception data; The calculation module is used to calculate the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data. An adjustment module is used to control the target charging pile group to dynamically adjust the charging strategy according to the priority weight; The step of calculating the priority weight of each charging request that satisfies the optimization objective based on the charging situation awareness data includes: Key features are extracted from the charging situation awareness data and normalized; the key features include user charging demand features, electric vehicle status features, and power grid operation features. Construct a quantitative scoring function for each of the key features, and dynamically adjust the weights of the quantitative scoring functions according to the optimization objective; The initial score for each feature is calculated using the quantization scoring function based on the adjusted weights. The real-time feedback correction item is determined based on the real-time operating status of the power grid, and the priority weight is determined based on the initial score and the real-time feedback correction item.
8. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the charging pile group collaborative control method according to any one of claims 1 to 6.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the charging pile group collaborative control method according to any one of claims 1 to 6.