A Current Quantity Intelligent Control Method and System Based on a Large Model

Abstract

This application relates to the technical field of electrical load control, and particularly to a current intelligent control method and system based on a large model. The current intelligent control method includes real-time acquisition of the current operating status of each charging pile and the current total load of the distribution cabinet; analysis based on the operating status of all charging piles and the overall load of the distribution cabinet; in the event of a risk, outputting the load demand trend of each charging pile within a preset time period based on a preset first LSTM model; and outputting the total load demand of the distribution cabinet within the preset time period based on a preset second LSTM model; if the difference between the rated maximum load and the total load demand of the distribution cabinet within the preset time period is greater than a preset capacity, assigning a priority to each charging pile; and generating a load allocation scheme based on the priority, real-time data, and predicted data. This application improves the timeliness and accuracy of load balance control for charging pile distribution cabinets.

Description

Technical Field

[0001] This application relates to the technical field of charging pile distribution cabinet control, and in particular to a current intelligent control method and system based on a large model. Background Technology

[0002] In today's society, with the increasing popularity of electric vehicles, the number of charging stations in residential communities is constantly increasing. As a key piece of equipment to ensure a stable power supply for charging stations, the load balancing control of the community charging station distribution cabinet is of paramount importance.

[0003] Currently, in terms of load balancing control of charging pile distribution cabinets in residential communities, most existing technologies rely on historical data and regular electricity consumption patterns for load prediction and allocation. This approach can, to some extent, cope with relatively regular daily electricity consumption.

[0004] However, in real-world applications, this control method based on historical data and conventional patterns has significant drawbacks. During peak electricity consumption periods, unexpected situations often arise, posing a significant challenge to load balancing control. For example, a large number of electric vehicles may suddenly connect to the charging station within a short period.

[0005] On the other hand, it is not uncommon for multiple high-power electric vehicles to simultaneously activate fast charging mode. Because high-power fast charging demands a huge amount of electricity, the simultaneous activation of fast charging by multiple such vehicles can cause drastic fluctuations in the load demand between charging stations.

[0006] Existing load balancing control methods, based on historical data and conventional electricity consumption patterns, are ill-suited to accurately address these unforeseen circumstances. This results in a significant lag in load balancing control, making it unable to respond promptly and effectively to sudden load changes. When the total load momentarily exceeds the distribution cabinet's capacity, it not only severely impacts the stability of the community's power supply, potentially causing power outages in some areas and disrupting residents' daily lives, but also threatens load safety, such as causing circuit overloads and short circuits, posing substantial safety hazards. Summary of the Invention

[0007] To improve the timeliness and accuracy of load balancing control in charging pile distribution cabinets, this application provides a current intelligent control method and system based on a large model.

[0008] Firstly, this application provides a current intelligent control method based on a large model, employing the following technical solution:

[0009] A current intelligent control method based on a large model includes the following steps:

[0010] Real-time acquisition of the current operating status of each charging pile and the current total load of the power distribution cabinet;

[0011] Based on the analysis of the working status of all charging piles and the overall load of the power distribution cabinet, when there are risk situations, the load demand change trend of each charging pile within a preset time period is output based on the preset first LSTM model; and the total load demand of the power distribution cabinet within a preset time period is output based on the preset second LSTM model.

[0012] If the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity, a priority will be set for each charging pile.

[0013] Based on priority, real-time data, and forecast data, a load allocation scheme is generated.

[0014] If the deviation between the actual total load and the total load demand exceeds the preset deviation value, or if the actual total load is greater than the third preset total load, the load demand will be re-predicted and the load allocation scheme will be adjusted.

[0015] In one embodiment, the risk situation is any one of the following three situations:

[0016] When the number of newly added electric vehicles charging within a first preset time exceeds the first preset proportion of the total number of charging piles, and the proportion of newly added vehicles charging in fast charging mode exceeds the second preset proportion, and when the current total load of the power distribution cabinet is greater than the first preset total load;

[0017] When the change in the current total load of the distribution cabinet exceeds the third preset ratio of the rated maximum load of the distribution cabinet within a second preset time period, and the current total load of the distribution cabinet is greater than the second preset total load;

[0018] When the current total load of the distribution cabinet is greater than the third preset total load;

[0019] Among them, the first preset total load is less than the second preset total load, and the second preset total load is less than the third preset total load.

[0020] In one embodiment: the current operating state includes charging status, current charging power, charging mode, battery type of the connected electric vehicle, current battery level of the electric vehicle, estimated time required to fully charge, and charging reservation information;

[0021] The priority generation method is to score each charging pile based on user information, equipment status, power grid efficiency and peak and off-peak status according to preset rules, and obtain the priority of each charging pile through weight fusion.

[0022] The user information includes user type, charging urgency, and historical credit score. Charging urgency = 1 - (estimated departure time - current time) / N; historical credit score is obtained based on the user's historical charging behavior.

[0023] The device status includes the charging pile's health, charging mode, the battery type of the connected electric vehicle, and the current battery level of the electric vehicle.

[0024] Power grid efficiency includes three-phase balance contribution and load stability.

[0025] In one embodiment: the step of generating a load allocation scheme based on priority, real-time data, and forecast data specifically includes:

[0026] Based on priority, a priority weight is generated for each charging pile, wherein the average priority weight of all charging piles is less than or equal to 1.

[0027] The predicted demand difference of charging piles, the actual remaining total load capacity, the total demand difference and the priority weight of the corresponding charging piles are used to calculate the expected power allocation of each charging pile, and a load allocation scheme is formed based on the expected power allocation of each charging pile.

[0028] The actual remaining total load capacity is calculated based on the difference between the rated maximum load of the distribution cabinet and the current total load, while the predicted total demand difference is calculated based on the difference between the maximum value of the total load demand and the rated maximum load of the distribution cabinet.

[0029] In one embodiment: if the sum of the expected allocated power of all charging piles is greater than the actual remaining total load capacity, the expected allocated power of each charging pile is adjusted based on the ratio of the actual remaining total load capacity to the sum of the expected allocated power of all charging piles.

[0030] In one embodiment, the control method further includes:

[0031] When analyzing the working status of a single charging pile and identifying charging piles with risks, the load demand change trend of each charging pile within a preset time period is output based on the preset first LSTM model, and the total load demand of the distribution cabinet within a preset time period is output based on the preset second LSTM model.

[0032] If the trend of load demand changes is greater than the safety threshold, and the difference between the rated maximum load and the total load demand of the distribution cabinet within the future preset time period is less than the preset capacity, the charging power of the charging pile is limited to within the safety threshold.

[0033] If the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity, a priority is set for each charging pile, and a load allocation scheme is generated based on the priority, real-time data and predicted data.

[0034] In one embodiment: a charging station is identified as a risky charging station when any one of the following conditions is met:

[0035] If the change in the current charging power of a single charging pile within a third preset time exceeds the fourth preset ratio of the rated power of that charging pile;

[0036] If the change in the current charging power of a single charging pile within a fourth preset time exceeds the fifth preset ratio of the rated power of the charging pile, and the current charging power reaches the sixth preset ratio of the rated power of the charging pile;

[0037] If the current charging power of a single charging pile reaches the seventh preset proportion of the rated power of that charging pile;

[0038] Among them, the fifth preset ratio is lower than the fourth preset ratio, the fourth preset ratio is lower than the sixth preset ratio, and the sixth preset ratio is lower than the seventh preset ratio.

[0039] In one embodiment, the control method includes:

[0040] Based on the load demand change trend of each charging pile, the load distribution of each phase circuit is calculated. If the load of a certain phase exceeds the fifth preset proportion of the rated load of that phase in the future, the charging pile connected to that phase will be adjusted, and the load will be transferred to the other two phases in proportion.

[0041] In one embodiment, the method for proportionally transferring the load is as follows:

[0042] The portion exceeding the fifth preset ratio is multiplied by the current load of that phase and divided by the sum of the loads of the two phases to obtain the allocated load quota for that phase;

[0043] When the sum of the current load condition of the second phase and the allocated load quota is greater than the fifth preset ratio, the remaining quota will be allocated to the third phase.

[0044] When the sum of the current load condition of the third phase and the allocated load quota is greater than the fifth preset ratio, the remaining quota will be allocated to the third phase.

[0045] The load quota transfer method is based on charging reservation information, which transfers excess charging reservations bound to this item to charging piles under the other two items.

[0046] Secondly, this application provides a current-based intelligent control system with a large model, employing the following technical solution:

[0047] A current-based intelligent control system, comprising:

[0048] The data acquisition module is used to acquire the current working status of each charging pile and the current total load of the power distribution cabinet in real time.

[0049] The load forecasting and analysis module analyzes the working status of all charging piles and the overall load of the distribution cabinet. When there are risks, it outputs the load demand trend of each charging pile within a preset time period based on the preset first LSTM model; and outputs the total load demand of the distribution cabinet within a preset time period based on the preset second LSTM model.

[0050] The scheme generation module sets a priority for each charging pile when the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity; and generates a load allocation scheme based on the priority, real-time data and predicted data.

[0051] If the load adjustment module finds that the deviation between the actual total load and the total load demand exceeds the preset deviation value, or that the actual total load is greater than the third preset total load, it will re-predict the load demand and adjust the load allocation scheme.

[0052] In summary, this application has the following beneficial effects: based on the analysis of real-time information, when a risk situation is identified, the load demand change trend of each charging pile and the total load demand of the distribution cabinet are predicted based on the preset LSTM model. Then, by analyzing the prediction results, a load allocation scheme is generated, thereby realizing the pre-load control of the distribution cabinet and charging piles. Attached Figure Description

[0053] Figure 1 A flowchart illustrating a current intelligent control method based on a large model, provided as one embodiment;

[0054] Figure 2 A flowchart illustrating a current intelligent control method based on a large model, provided as one embodiment;

[0055] Figure 3 A logic block diagram of a current-based intelligent control system based on a large model is provided for one embodiment.

[0056] In the diagram, 10 is the data acquisition module; 20 is the load forecasting and analysis module; 30 is the scheme generation module; and 40 is the scheme adjustment module. Detailed Implementation

[0057] The present application will be further described in detail below with reference to the accompanying drawings.

[0058] To better understand the purpose, technical solutions, and advantages of this application, it has been described and illustrated below with reference to the accompanying drawings and embodiments. However, those skilled in the art should understand that this application can be implemented without these details. In some cases, to avoid obscuring various aspects of this application due to unnecessary description, well-known methods, processes, systems, components, and / or circuits already described at a higher level will not be elaborated upon. It will be apparent to those skilled in the art that various modifications can be made to the embodiments disclosed in this application, and the general principles defined in this application can be applied to other embodiments and application scenarios without departing from the principles and scope of this application. Therefore, this application is not limited to the illustrated embodiments, but conforms to the broadest scope consistent with the scope of protection claimed in this application.

[0059] It should be noted that the descriptions of these embodiments are for the purpose of aiding understanding the present invention, but do not constitute a limitation thereof. Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0060] In the description of this application, "several" means one or more, "more than" means two or more, "greater than," "less than," and "exceeding" are understood to exclude the stated number, while "above," "below," and "within" are understood to include the stated number. The use of "first" and "second" in the description is merely for distinguishing technical features and should not be construed as indicating or implying relative importance, or implicitly indicating the number of indicated technical features, or implicitly indicating the order of the indicated technical features.

[0061] In the description of this application, the terms "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples.

[0062] like Figure 1 As shown, this application discloses a current intelligent control method based on a large model, including:

[0063] S100: Real-time acquisition of the current working status of each charging pile and the current total load of the power distribution cabinet.

[0064] In this step, the current working status includes charging status, current charging power, charging mode, battery type of the connected electric vehicle, current battery level of the electric vehicle, estimated time required to fully charge, and charging reservation information.

[0065] The charging status is divided into two types: charging and not charging. The charging mode is divided into two types: fast charging and slow charging. The charging reservation information includes the reserved charging time, reserved charging duration, and reserved charging mode.

[0066] S200 analyzes the working status of all charging piles and the load of the overall power distribution cabinet. When there is a risk, it outputs the load demand trend of each charging pile within a preset time period based on the preset first LSTM model; and outputs the total load demand of the power distribution cabinet within a preset time period based on the preset second LSTM model.

[0067] The risk situation mentioned in this step refers to any one of the following three situations:

[0068] 1. When the number of newly added electric vehicles charging within the first preset time exceeds the first preset proportion of the total number of charging piles, and the proportion of newly added vehicles charging in fast charging mode exceeds the second preset proportion, and when the current total load of the power distribution cabinet is greater than the first preset total load;

[0069] 2. When the change in the current total load of the distribution cabinet exceeds the third preset ratio of the rated maximum load of the distribution cabinet within the second preset time period, and the current total load of the distribution cabinet is greater than the second preset total load;

[0070] 3. When the current total load of the distribution cabinet is greater than the third preset total load.

[0071] Among them, the first preset total load is less than the second preset total load, and the second preset total load is less than the third preset total load.

[0072] The first scenario primarily addresses situations where a large number of electric vehicles are simultaneously connected to charging stations within a short period, especially when fast charging is prevalent. The second scenario mainly addresses situations where the load on the distribution cabinet is already relatively high but continues to grow rapidly. The third scenario involves directly comparing the load of the distribution cabinet to avoid the current total load being too high, preventing overload from occurring before any forecast is made.

[0073] In addition, the first LSTM model and the second LSTM model are trained based on historical data. The first LSTM model and the second LSTM model preferably adopt the LSTM model based on the attention mechanism. This type of model combines the convolutional neural network (CNN) and the long short-term memory network (LSTM). By introducing the attention mechanism, it can better capture the importance of different features in the time series, thereby optimizing the prediction effect and improving the accuracy of time series prediction.

[0074] The first LSTM model is trained by taking the historical operating status of the charging piles as input, while the second LSTM model is trained by taking the historical operating status of the charging piles and the total load data of the historical distribution cabinets as input.

[0075] S300: If the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity, a priority is set for each charging pile.

[0076] In this step, the preset capacity is used to make a judgment so that the current total load of the distribution cabinet can have enough room for adjustment.

[0077] The aforementioned priority generation method involves scoring users based on their information, equipment status, grid efficiency, and peak / off-peak conditions according to preset rules, and then using weighted fusion to obtain the priority of each charging pile.

[0078] Specifically, user information includes user type, charging urgency, and historical credit score.

[0079] User types are categorized into permanent residents, temporary visitors, and special groups, with pre-set scores assigned to each type. Charging urgency is calculated as 1 - (estimated departure time - current time) / N, where N can be set according to needs, but it's crucial to ensure that N's unit is consistent with other time units. Historical credit scores are based on users' past charging behavior; for example, points are deducted for customers who make reservations but don't charge, or who complete charging but don't leave on time. Scoring can be done directly using the credit score or by pre-setting scores based on score ranges.

[0080] The device status includes the charging pile's health, charging mode, the battery type of the connected electric vehicle, and the current battery level of the electric vehicle.

[0081] The health of the charging station is primarily calculated based on its failure rate over the past 30 days, with a score assigned accordingly. The charging mode is scored using preset scores for fast and slow charging. The battery type of the connected electric vehicle is scored based on a preset matching degree, where matching degree mainly refers to charging efficiency; higher charging efficiency results in a higher score. The current battery level of the electric vehicle also reflects the urgency of charging; therefore, the lower the remaining battery power, the higher the score.

[0082] Power grid efficiency includes three-phase balance contribution and load stability.

[0083] Among them, the three-phase balance contribution refers to the phase circuit conditions connected to the charging pile; the lower the load on the phase circuit, the higher the score. Load stability is judged based on the historical fluctuations in charging power; the smaller the fluctuations, the higher the score.

[0084] The scores for user information, equipment status, and grid efficiency are obtained by averaging the scores after normalizing each component. Taking user information as an example, the average of the normalized scores for user type, charging urgency, and historical credit score is the user information score.

[0085] The scoring for peak and off-peak periods can be achieved by simply pre-setting a normalized score for each of the different peak and off-peak periods. The fusion weights for weighted fusion can be set manually based on experience or obtained based on historical data analysis.

[0086] S400 generates a load allocation scheme based on priority, real-time data, and forecast data.

[0087] The specific steps in generating the load allocation scheme include:

[0088] Based on priority, a priority weight is generated for each charging pile, wherein the average priority weight of all charging piles is less than or equal to 1.

[0089] The predicted demand difference of charging piles, the actual remaining total load capacity, the total demand difference, and the priority weight of the corresponding charging piles are used to calculate the expected power allocation of each charging pile, and a load allocation scheme is formed based on the expected power allocation of each charging pile.

[0090] The above calculation formula is: the expected allocated power of the charging pile = the difference between the predicted demand of the charging pile × (actual remaining total load capacity ÷ difference between the predicted total demand) × the priority weight of the charging pile.

[0091] The actual remaining total load capacity is calculated based on the difference between the rated maximum load of the distribution cabinet and the current total load, while the predicted total demand difference is calculated based on the difference between the maximum value of the total load demand and the rated maximum load of the distribution cabinet.

[0092] In addition, the priority weights set in this embodiment are required to be greater than 1 in some cases. Therefore, the preferred method for generating priority weights is to increase the priority by a proportional scale.

[0093] Furthermore, if the sum of the projected allocated power of all charging stations exceeds the actual remaining total load capacity, the projected allocated power of each charging station is adjusted based on the ratio of the actual remaining total load capacity to the sum of the projected allocated power of all charging stations. This setting prevents the overall projected allocated power from becoming too high.

[0094] S500 If the deviation between the actual total load and the total load demand is found to exceed the preset deviation value, or if the actual total load is greater than the third preset total load, the load demand is re-predicted and the load allocation scheme is adjusted.

[0095] By setting this step, the actual load and the predicted load can be compared in real time, avoiding situations where the predicted results deviate from the actual situation, which could lead to an unreasonable load allocation scheme based on the prediction.

[0096] like Figure 2 As shown, in one embodiment, the control method further includes:

[0097] S600 performs operational status analysis on a single charging pile. When identifying charging piles with potential risks, it outputs the load demand change trend of each charging pile within a preset time period based on a preset first LSTM model, and outputs the total load demand of the distribution cabinet within a preset time period based on a preset second LSTM model.

[0098] S700 If the load demand change trend is greater than the safety threshold, and the difference between the rated maximum load and the total load demand of the distribution cabinet within the future preset time period is less than the preset capacity, the charging power of the charging pile is limited to within the safety threshold.

[0099] S800: If the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity, a priority is set for each charging pile, and a load allocation scheme is generated based on the priority, real-time data and predicted data.

[0100] Specifically, a charging station is considered a risky charging station if it meets any of the following conditions:

[0101] If the change in the current charging power of a single charging pile within a third preset time exceeds the fourth preset ratio of the rated power of that charging pile;

[0102] If the change in the current charging power of a single charging pile within a fourth preset time exceeds the fifth preset ratio of the rated power of the charging pile, and the current charging power reaches the sixth preset ratio of the rated power of the charging pile;

[0103] If the current charging power of a single charging pile reaches the seventh preset proportion of the rated power of that charging pile;

[0104] Among them, the fifth preset ratio is lower than the fourth preset ratio, the fourth preset ratio is lower than the sixth preset ratio, and the sixth preset ratio is lower than the seventh preset ratio.

[0105] In this embodiment, the working status of a single charging pile is analyzed to avoid the situation where a single charging pile is overloaded without prior prediction before a risk situation is identified in step S200. Furthermore, when a risky charging pile is identified, the total load demand of the power distribution cabinet is also predicted to further ensure that the power distribution cabinet does not experience excessive load.

[0106] In another embodiment, the control method further includes:

[0107] Based on the load demand change trend of each charging pile, the load distribution of each phase circuit is calculated. If the load of a certain phase exceeds the fifth preset proportion of the rated load of that phase in the future, the charging pile connected to that phase will be adjusted, and the load will be transferred to the other two phases in proportion.

[0108] In this embodiment, the method for transferring the load proportionally is as follows:

[0109] The portion exceeding the fifth preset ratio is multiplied by the current load of that phase and divided by the sum of the loads of the two phases to obtain the allocated load quota for that phase;

[0110] When the sum of the current load condition of the second phase and the allocated load quota is greater than the fifth preset ratio, the remaining quota will be allocated to the third phase.

[0111] When the sum of the current load condition of the third phase and the allocated load quota is greater than the fifth preset ratio, the remaining quota will be allocated to the third phase.

[0112] The second phase and the third phase are the other two phases in the three-phase circuit. Preferably, the load of the second phase is greater than that of the third phase, and the load is allocated from high to low during identification.

[0113] In addition, the load quota is transferred based on the charging reservation information, transferring the excess charging reservations bound to this item to the charging piles under the other two items.

[0114] This step is designed to balance the load of the three-phase circuit and further control the stability of the distribution cabinet.

[0115] like Figure 3 As shown, a current-based intelligent control system includes:

[0116] The data acquisition module 10 is used to acquire the current working status of each charging pile and the current total load of the power distribution cabinet in real time.

[0117] The load forecasting and analysis module 20 analyzes the working status of all charging piles and the overall load of the distribution cabinet. When there is a risk, it outputs the load demand change trend of each charging pile within a preset time period based on the preset first LSTM model; and outputs the total load demand of the distribution cabinet within a preset time period based on the preset second LSTM model.

[0118] The scheme generation module 30 sets a priority for each charging pile when the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity; and generates a load allocation scheme based on the priority, real-time data and predicted data.

[0119] If the scheme adjustment module 40 finds that the deviation between the actual total load and the total load demand exceeds the preset deviation value, or that the actual total load is greater than the third preset total load, it will re-predict the load demand and adjust the load allocation scheme.

[0120] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A large model-based current flow intelligent control method, characterized in that, Includes the following steps: Real-time acquisition of the current operating status of each charging pile and the current total load of the power distribution cabinet; Based on the analysis of the working status of all charging piles and the overall load of the power distribution cabinet, when there are risk situations, the load demand change trend of each charging pile within a preset time period is output based on the preset first LSTM model; and the total load demand of the power distribution cabinet within a preset time period is output based on the preset second LSTM model. If the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity, a priority will be set for each charging pile. Based on priority, real-time data, and forecast data, a load allocation scheme is generated. If the deviation between the actual total load and the total load demand exceeds the preset deviation value, or if the actual total load is greater than the third preset total load, the load demand will be re-predicted and the load allocation scheme will be adjusted. The step of generating a load allocation scheme based on priority, real-time data, and forecast data specifically includes: Based on priority, a priority weight is generated for each charging pile, wherein the average priority weight of all charging piles is less than or equal to 1. The predicted demand difference of charging piles, the actual remaining total load capacity, the total demand difference and the priority weight of the corresponding charging piles are used to calculate the expected power allocation of each charging pile, and a load allocation scheme is formed based on the expected power allocation of each charging pile. The actual remaining total load capacity is calculated based on the difference between the rated maximum load of the distribution cabinet and the current total load, while the predicted total demand difference is calculated based on the difference between the maximum value of the total load demand and the rated maximum load of the distribution cabinet.

2. The large model-based electric current intelligent control method according to claim 1, characterized in that, The risk situation is any one of the following three situations: When the number of newly added electric vehicles charging within a first preset time exceeds the first preset proportion of the total number of charging piles, and the proportion of newly added vehicles charging in fast charging mode exceeds the second preset proportion, and when the current total load of the power distribution cabinet is greater than the first preset total load; When the change in the current total load of the distribution cabinet exceeds the third preset ratio of the rated maximum load of the distribution cabinet within a second preset time period, and the current total load of the distribution cabinet is greater than the second preset total load; When the current total load of the distribution cabinet is greater than the third preset total load; Among them, the first preset total load is less than the second preset total load, and the second preset total load is less than the third preset total load.

3. The large model-based electric current intelligent control method according to claim 1, wherein, The current operating status includes charging status, current charging power, charging mode, battery type of the connected electric vehicle, current battery level of the electric vehicle, estimated time required to fully charge, and charging reservation information; The priority generation method is to score each charging pile based on user information, equipment status, power grid efficiency and peak and off-peak status according to preset rules, and obtain the priority of each charging pile through weight fusion. The user information includes user type, charging urgency, and historical credit score. Charging urgency = 1 - (estimated departure time - current time) / N; historical credit score is obtained based on the user's historical charging behavior; N is set according to needs, and the units of N and other times are kept consistent. The device status includes the charging pile's health, charging mode, the battery type of the connected electric vehicle, and the current battery level of the electric vehicle. Power grid efficiency includes three-phase balance contribution and load stability.

4. The large model-based electric current intelligent control method according to claim 1, characterized in that: If the sum of the expected allocated power of all charging piles is greater than the actual remaining total load capacity, the expected allocated power of each charging pile will be adjusted based on the ratio of the actual remaining total load capacity to the sum of the expected allocated power of all charging piles.

5. The large model-based electric current intelligent control method according to claim 1, wherein, The control method further includes: When analyzing the working status of a single charging pile and identifying charging piles with risks, the load demand change trend of each charging pile within a preset time period is output based on the preset first LSTM model, and the total load demand of the distribution cabinet within a preset time period is output based on the preset second LSTM model. If the trend of load demand changes is greater than the safety threshold, and the difference between the rated maximum load and the total load demand of the distribution cabinet within the future preset time period is less than the preset capacity, the charging power of the charging pile is limited to within the safety threshold. If the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity, a priority is set for each charging pile, and a load allocation scheme is generated based on the priority, real-time data and predicted data.

6. The large model-based electric quantity intelligent control method according to claim 5, characterized in that, A charging station is considered a risky charging station if it meets any of the following conditions: If the change in the current charging power of a single charging pile within a third preset time exceeds the fourth preset ratio of the rated power of that charging pile; If the change in the current charging power of a single charging pile within a fourth preset time exceeds the fifth preset ratio of the rated power of the charging pile, and the current charging power reaches the sixth preset ratio of the rated power of the charging pile; If the current charging power of a single charging pile reaches the seventh preset proportion of the rated power of that charging pile; Among them, the fifth preset ratio is lower than the fourth preset ratio, the fourth preset ratio is lower than the sixth preset ratio, and the sixth preset ratio is lower than the seventh preset ratio.

7. The large model-based electric quantity intelligent control method according to claim 5 or 6, characterized in that, The control method includes: Based on the load demand change trend of each charging pile, the load distribution of each phase circuit is calculated. If the load of a certain phase exceeds the fifth preset proportion of the rated load of that phase in the future, the charging pile connected to that phase will be adjusted, and the load will be transferred to the other two phases in proportion.

8. The large model-based electric quantity intelligent control method according to claim 7, characterized in that, The method for transferring the load proportionally is as follows: The portion exceeding the fifth preset ratio is multiplied by the current load of that phase and divided by the sum of the loads of the two phases to obtain the allocated load quota for that phase; When the sum of the current load condition of the second phase and the allocated load quota is greater than the fifth preset ratio, the remaining quota will be allocated to the third phase. When the sum of the current load condition of the third phase and the allocated load quota is greater than the fifth preset ratio, the remaining quota will be allocated to the third phase. The load quota transfer method is based on charging reservation information, which transfers excess charging reservations bound to this item to charging piles under the other two items.

9. A large model-based current flow intelligent control system applied to the large model-based current flow intelligent control method of any one of claims 1-8, characterized in that, include: The data acquisition module (10) is used to acquire the current working status of each charging pile and the current total load of the power distribution cabinet in real time; The load forecasting and analysis module (20) analyzes the working status of all charging piles and the load status of the overall distribution cabinet. When there is a risk, it outputs the load demand change trend of each charging pile within a preset time period based on the preset first LSTM model; and outputs the total load demand of the distribution cabinet within a preset time period based on the preset second LSTM model. The scheme generation module (30) sets a priority for each charging pile when the difference between the rated maximum load and the total load demand of the distribution cabinet within a preset time period is greater than the preset capacity; and generates a load allocation scheme based on the priority, real-time data and predicted data; the step of generating a load allocation scheme based on the priority, real-time data and predicted data specifically includes: generating a priority weight for each charging pile based on the priority, wherein the average priority weight of all charging piles is less than or equal to 1; calculating the expected allocation power of each charging pile based on the predicted demand difference, the actual remaining total load capacity, the total demand difference and the priority weight of the corresponding charging pile, and forming a load allocation scheme based on the expected allocation power of each charging pile; wherein the actual remaining total load capacity is calculated based on the difference between the rated maximum load of the distribution cabinet and the current total load, and the predicted total demand difference is calculated based on the difference between the maximum value in the total load demand and the rated maximum load of the distribution cabinet; If the scheme adjustment module (40) finds that the deviation between the actual total load and the total load demand exceeds the preset deviation value, or the actual total load is greater than the third preset total load, it will re-predict the load demand and adjust the load allocation scheme.

Images