Power Control Method, Device, System and Storage Medium for Microgrid System

Abstract

The present application discloses a power control method, device, system and storage medium of a microgrid system, and relates to the technical field of microgrid systems. The power control method of the microgrid system comprises: obtaining power load data of the microgrid system in a historical power generation period, and determining the power state of the microgrid system according to the power load data; when the power state is a power surplus state, generating a power load description text based on a large language model, power load data and meteorological data of the historical power generation period; predicting a target power load prediction result of the microgrid system based on a power load prediction model and the power load description text; determining a power dispatching plan according to the target power load prediction result and the target meteorological data, and dispatching the corresponding power generation nodes to operate based on the power dispatching plan to achieve a power balance state, so that power control can be performed from the root to achieve power balance, thereby eliminating the need to connect surplus power to the grid and avoiding power waste caused by line transmission loss.

Description

Technical Field

[0001] This application relates to the technical field of microgrid systems, and particularly to a power control method, device, system, and storage medium for microgrid systems. Background Art

[0002] A microgrid system, also known as a micro power grid or microgrid, is a small power generation and distribution system that combines distributed power sources, energy storage devices, energy conversion devices, and user equipment. Among them, the distributed power sources include multiple different types of power generation nodes, such as renewable energy power generation nodes like wind power generation nodes, photovoltaic power generation nodes, fuel cells, etc. These nodes provide the main electrical energy supply for the microgrid system. In related technologies, if there is a power surplus in the microgrid system, the surplus power is transmitted to the main grid at a low price. Due to line transmission losses during the process of grid connection of the surplus power, power waste is caused. Summary of the Invention

[0003] The main purpose of this application is to provide a power control method, device, system, and storage medium for a microgrid system, aiming to solve the problem of power waste caused by line transmission losses during the process of grid connection of surplus power.

[0004] To achieve the above purpose, this application proposes a power control method for a microgrid system, and the method includes:

[0005] Obtain the power load data of the microgrid system during the historical power generation period, and determine the power state of the microgrid system according to the power load data;

[0006] When the power state is a power surplus state, generate a power load description text based on a large language model, the power load data, and the meteorological data of the historical power generation period;

[0007] Based on a power load prediction model and the power load description text, predict the target power load prediction result of the microgrid system;

[0008] Determine a power dispatching plan according to the target power load prediction result and the target meteorological data, and dispatch the corresponding power generation nodes to operate based on the power dispatching plan to achieve a power balance state.

[0009] In one embodiment, the power load data includes a load value, and generating a power load description text based on a large language model, the power load data, and the meteorological data of the historical power generation period includes:

[0010] Input the load value and the meteorological data into the large language model, perform feature extraction on the load value and the meteorological data respectively, and obtain a load value feature and a meteorological data feature;

[0011] Determine a target prompt template associated with the load value feature and the meteorological data feature according to the similarity between the load value feature, the meteorological data feature, and each preset prompt template;

[0012] Embed the load value feature and the meteorological data feature into the target prompt template to obtain the power load description text.

[0013] In one embodiment, the power load prediction model includes a time series model and a neural network model. The steps of predicting the target power load prediction result of the microgrid system based on the power load prediction model and the power load description text include:

[0014] Extract features from the power load description text to obtain the power load impact features of the microgrid system;

[0015] Determine a first prediction result of the time series model based on the power load impact features, and determine a second prediction result of the neural network model based on the power load impact features;

[0016] Determine the target power load prediction result of the microgrid system according to the first prediction result and the second prediction result.

[0017] In one embodiment, the step of determining the target power load prediction result of the microgrid system according to the first prediction result and the second prediction result includes:

[0018] Obtain the weights of the time series model and the weights of the neural network model;

[0019] Determine the target power load prediction result of the microgrid system according to the product of the first prediction result and the weights of the time series model, and the sum of the product of the second prediction result and the weights of the neural network model.

[0020] In one embodiment, before the step of obtaining the weights of the time series model and the weights of the neural network model, it further includes:

[0021] Obtain historical power load data samples;

[0022] Based on the training data length corresponding to the time series model, divide the historical power load data samples into a first training data set and a first test data set, train an initial time series model based on the first training data set, and test the initial time series model based on the first test data set to obtain the weights of the time series model; and,

[0023] Divide the historical power load data samples into a second training data set and a second test data set based on the length of the training data corresponding to the neural network model, train an initial neural network model based on the second training data set, and test the initial neural network model based on the second test data set to obtain the weights of the neural network model.

[0024] In one embodiment, the target power load prediction result includes a target net load. Determining a power dispatch plan according to the target power load prediction result and target meteorological data, and dispatching the corresponding power generation nodes to operate based on the power dispatch plan includes:

[0025] Determine the power dispatch plan according to the target meteorological data, the target net load, and the corresponding relationship between the preset meteorological data, the preset load value, and the preset dispatch plan;

[0026] According to the power dispatch plan, determine the power generation amount of each power generation node, and determine the power generation power of each power generation node according to the power generation amount of each power generation node;

[0027] Control each of the power generation nodes to operate according to the corresponding power generation power.

[0028] In one embodiment, the microgrid system includes a plurality of power generation nodes and a plurality of power consumption nodes. The steps of obtaining the power load data of the microgrid system during the historical power generation period and determining the power state of the microgrid system according to the power load data include:

[0029] Determine the total power generation amount according to the power generation amounts of the respective power generation nodes of the microgrid system during the historical power generation period, and determine the total power consumption amount according to the power consumption amounts of the respective power consumption nodes of the microgrid system during the historical power generation period;

[0030] Determine the net load of the microgrid system during the historical power generation period according to the total power consumption amount and the total power generation amount, wherein the power load data further includes the net load;

[0031] When the net load is less than a preset value, determine that the power state is a power surplus state.

[0032] In addition, to achieve the above object, the present application also proposes a power control device for a microgrid system. The device includes: a memory, a processor, and a computer program stored on the memory and executable on the processor. The computer program is configured to implement the steps of the power control method for the microgrid system as described above.

[0033] In addition, to achieve the above object, the present application also proposes a microgrid system, which includes a plurality of power generation nodes, a plurality of power consumption nodes, and a power control device of the microgrid system. Each of the power generation nodes and each of the power consumption nodes are connected to the power control device of the microgrid system.

[0034] In addition, to achieve the above object, the present application also proposes a storage medium, which is a computer-readable storage medium. A computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the power control method of the microgrid system as described above are implemented.

[0035] Compared with the related art, the present application obtains the power load data of the microgrid system during the historical power generation period, determines the power state of the microgrid system according to the power load data; when the power state is a power surplus state, generates a power load description text based on a large language model, the power load data, and the meteorological data of the historical power generation period; predicts the target power load prediction result of the microgrid system based on a power load prediction model and the power load description text; determines a power dispatch plan according to the target power load prediction result and target meteorological data, and schedules the operation of the corresponding power generation nodes based on the power dispatch plan to achieve a power balance state. Since it can predict the future power load and formulate a power dispatch plan when detecting a power surplus, and perform power control from the root by executing the power dispatch plan to achieve power balance, there is no need to connect the surplus power to the grid, avoiding power waste caused by line transmission losses. And by combining the power load description text of the large language model with the power load prediction model, the respective advantages can be fully utilized to improve the accuracy and reliability of the target power load prediction result. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings herein are incorporated into the specification and constitute a part of the specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0037] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0038] Figure 1 It is a schematic flowchart provided for the first embodiment of the power control method of the microgrid system of the present application;

[0039] Figure 2 It is a schematic flowchart provided for the second embodiment of the power control method of the microgrid system of the present application;

[0040] Figure 3 A flowchart provided for the third embodiment of the power control method of the microgrid system of this application;

[0041] Figure 4 A flowchart provided for the fourth embodiment of the power control method of the microgrid system of this application;

[0042] Figure 5 A flowchart provided for the fifth embodiment of the power control method of the microgrid system of this application;

[0043] Figure 6 A structural diagram of the power control device of the microgrid system of this application.

[0044] The realization of the purpose, functional characteristics and advantages of this application will be further described with reference to the embodiments and the accompanying drawings. Specific embodiments

[0045] It should be understood that the specific embodiments described herein are only used to explain the technical solutions of this application and are not used to limit this application.

[0046] In order to better understand the technical solutions of this application, the following will be described in detail in conjunction with the drawings of the specification and specific embodiments.

[0047] In the related art, if there is a power surplus in the microgrid system, the surplus power is transmitted to the main power grid at a low price. Due to the line transmission loss during the grid connection process of the surplus power, power waste is caused.

[0048] In view of the above problems, obtain the power load data of the microgrid system during the historical power generation period, and determine the power state of the microgrid system according to the power load data; when the power state is a power surplus state, generate a power load description text based on the large language model, the power load data and the meteorological data of the historical power generation period; based on the power load prediction model and the power load description text, predict the target power load prediction result of the microgrid system; determine the power dispatching plan according to the target power load prediction result and the target meteorological data, and dispatch the corresponding power generation node to operate based on the power dispatching plan to achieve a power balance state.

[0049] Since it is able to predict future power loads and formulate power dispatching plans when detecting a power surplus, and conduct power control from the root by implementing the power dispatching plan to achieve power balance, there is no need to connect the surplus power to the grid, thus avoiding power waste caused by line transmission losses. Moreover, in order to improve the accuracy of the target power load prediction results, first generate a power load description text based on the large language model, the power load data, and the meteorological data of the historical power generation period, and then predict the target power load prediction results of the microgrid system based on the power load prediction model and the power load description text. By combining the power load description text of the large language model with the power load prediction model, the respective advantages can be fully utilized to improve the accuracy and reliability of the target power load prediction results.

[0050] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device capable of implementing the above functions, the power control device of the microgrid system, etc. Hereinafter, the power control device of the microgrid system will be taken as an example to illustrate this embodiment and the following embodiments.

[0051] Based on this, the embodiment of the present application provides a power control method for a microgrid system, referring to Figure 1 , Figure 1 is a schematic flowchart of an embodiment of the power control method for the microgrid system of the present application. In this embodiment, the power control method for the microgrid system includes steps S10 to S40:

[0052] Step S10, obtain the power load data of the microgrid system during the historical power generation period, and determine the power state of the microgrid system according to the power load data;

[0053] It should be noted that power load data refers to information related to the load of power nodes or power consumption nodes, which reflects the power demand of a specific user at a certain moment or within a certain period of time. Specifically, the power load data specifically includes but is not limited to: active power, reactive power, power factor, voltage, current, load curve, peak load, and load rate, etc.

[0054] The above-mentioned active power represents the electrical power demand for actual work and is one of the core indicators in electrical load data. It is usually measured in kilowatts (kW) or megawatts (MW) and reflects the active electrical energy consumed by electrical equipment. Reactive power is related to parameters such as the electromagnetic field and capacitance and inductance in the microgrid system. Although it does not directly do work, it has an important impact on the stability of the system and voltage regulation. It is usually measured in kilovars (kVar) or megavars (MVar). The power factor is the ratio of active power to apparent power and is an important parameter for evaluating the nature of the load (such as resistive, inductive, or capacitive). Improving the power factor helps reduce reactive power losses in the power grid and improve energy utilization efficiency. Voltage and current are the voltage and current levels at the load point and are very important for evaluating the operating state of the system and load characteristics. Their changes can reflect the power supply quality of the power grid and the operating conditions of electrical equipment. The load curve represents the curve of load variation over time during a historical power generation period (such as a day, a week, or a month). By analyzing the load curve, it is possible to understand the volatility and periodicity of the load, providing a basis for the scheduling and planning of the microgrid system. The peak load is the maximum load value that occurs within a specific time period and is an important reference index in the design and operation of the microgrid system. The magnitude of the peak load determines the maximum power supply capacity that the microgrid system needs to meet. The load factor represents the ratio of the average load to the peak load and is used to evaluate the smoothness or volatility of the load. The higher the load factor, the smoother the load; conversely, it indicates that the load fluctuates greatly.

[0055] In a feasible implementation, the method of obtaining the electrical load data of the microgrid system during the historical power generation period includes: usually collecting data through measuring devices such as smart meters and sensors installed at key nodes of the power grid and transmitting it to the central control system or data center for analysis and processing. By obtaining the above data, it provides a strong guarantee for the safe, stable, and efficient operation of the microgrid system.

[0056] It should be noted that the power state includes power surplus state, power balance state, power shortage state, etc.

[0057] In a feasible implementation, the total power generation is determined based on the power generation of each power generation node of the microgrid system during the historical power generation period, and the total power consumption is determined based on the power consumption of each power consumption node of the microgrid system during the historical power generation period; when the total power generation is greater than the total power consumption, the microgrid system is in a power surplus state, and the surplus power can be used for energy storage, output to the external power grid, or other forms of utilization. When the total power generation is equal to the total power consumption, the system is in a power balance state. When the total power generation is less than the total power consumption, the system is in a power shortage state. At this time, it is necessary to purchase power from the external power grid, activate the energy storage system, or take other measures to meet the power consumption demand.

[0058] However, if there is a power surplus in the microgrid system and the surplus power is transmitted to the main power grid at a low price, there will be power waste due to line transmission losses during the grid connection process of the surplus power. Therefore, step S20 is executed. When the power state is in a power surplus state, a power load description text is generated based on the large language model, the power load data, and the meteorological data of the historical power generation period.

[0059] When the power state is in a power surplus state, in order to avoid the problem of power waste caused by line transmission losses during the process of transmitting the surplus power to the main power grid at a low price. This application can predict the future power load and formulate a power dispatching plan when detecting a power surplus, and perform power control from the root by executing this power dispatching plan to achieve power balance, so as to avoid grid connection of the surplus power and prevent power waste caused by line transmission losses.

[0060] In order to improve the accuracy and reliability of the target power load prediction result and make the acquisition of the target power load prediction result more intelligent, this application combines the power load description text of the large language model with the power load prediction model, which can make full use of their respective advantages to improve the accuracy and reliability of the target power load prediction result.

[0061] It should be noted that when using the large language model to generate the power load description text, the power load description text can be used as an additional feature and input into the subsequent power load prediction model to facilitate the understanding of the power load prediction model. For example, the large language model can generate descriptions about weather conditions, power loads, etc., which may be related to the changes in power loads and meteorological data during the historical power generation period. Subsequently, by analyzing these data, the accuracy of the prediction result can be improved.

[0062] It should be noted that during the use of the large language model, some prompt information also needs to be added to improve the accuracy of the power load description text.

[0063] In a feasible implementation manner, the corresponding prompt information, power load data, meteorological data, etc. can be input into the large language model together, and the power load description text is generated through the large language model. For example, assuming the power load data is "Current time: 2023-04-01 14:00, Meteorological data: Sunny, Load value: 5000MW, Increased by 5% compared to the previous hour", the following input information can be designed: The description should include the meteorological data, load value, and the change in the load value compared to the previous hour during the historical preset period. The output power load description text can be: The current power load data shows that at 14:00 on April 1, 2023, the weather is sunny, the power load value is 5000MW, and it has increased by 5% compared to the previous hour, showing a certain upward trend.

[0064] Step S30: Based on the power load prediction model and the power load description text, predict the target power load prediction result of the microgrid system;

[0065] After obtaining the power load description text, the power load description text can be input into the power load prediction model to obtain the target power load prediction result of the microgrid system.

[0066] It should be noted that the power load prediction model is used to predict the future power load prediction result based on the power load data during the historical power generation period. Here, the power load prediction model can be a time series model, a deep learning model, a machine learning model, etc., and the corresponding power load prediction model can be selected according to actual needs.

[0067] In order to meet the power load prediction in different scenarios, multiple different power load prediction models can be set simultaneously for prediction to meet the power load prediction requirements in different scenarios. The more power load prediction models are selected, the more scenarios can be applicable. And by setting different power load prediction models, it can be verified whether the target power load prediction result output by a single prediction model is accurate, improving the accuracy of the power load prediction result.

[0068] In a feasible implementation manner, when the power load prediction models adopted are a time series model and a deep learning model, the power load description text can be input into the time series model to obtain the power load prediction result of the time series model, and at the same time, the power load description text can be input into the deep learning model to obtain the power load prediction result of the deep learning model. Finally, based on the power load prediction result of the time series model and the power load prediction result of the deep learning model, jointly determine the target power load prediction result of the microgrid system.

[0069] Step S40: Determine the power dispatch plan according to the target power load prediction result and the target meteorological data, and dispatch the corresponding power generation nodes to operate based on the power dispatch plan to achieve the power balance state.

[0070] After determining the power generation amounts of each power generation node, determine the power generation power required for each power generation node to reach that power generation amount, and control each power generation node to operate according to the corresponding power generation power, so that when each power generation node operates according to the corresponding power generation power, the microgrid system achieves the power balance state.

[0071] In this embodiment, since it is possible to predict the future power load and formulate a power dispatching plan when detecting a power surplus, and perform power control from the root by executing this power dispatching plan to achieve power balance, there is no need to connect the surplus power to the grid, thus avoiding power waste caused by line transmission losses. Moreover, in order to improve the accuracy of the target power load prediction result, first generate a power load description text based on the large language model, the power load data, and the meteorological data of the historical power generation period, and then predict the target power load prediction result of the microgrid system based on the power load prediction model and the power load description text. By combining the power load description text of the large language model with the power load prediction model, the respective advantages can be fully utilized to improve the accuracy and reliability of the target power load prediction result.

[0072] Based on the above embodiment, in the second embodiment of the present application, refer to Figure 2 , the power load data includes a load value, and here the load value can be a net load. The generation of the power load description text based on the large language model, the power load data, and the meteorological data of the historical power generation period in step S20 includes steps S21 to S23:

[0073] Step S21, input the load value and the meteorological data into the large language model, perform feature extraction on the load value and the meteorological data respectively, and obtain a load value feature and a meteorological data feature;

[0074] In a feasible implementation manner, since the dimensions and ranges of the load value and the meteorological data may be different, it is necessary to perform standardization or normalization processing on them so that all features are within the same numerical range, and then input the preprocessed load value and meteorological data into the large language model, which helps the large language model to converge faster and improve performance. Check for missing values in the data and take corresponding measures for filling, such as using the mean, median, interpolation method, or model-based prediction, etc. If the meteorological data contains categorical variables such as weather types, these categorical variables need to be converted into numerical features, and common methods include one-hot encoding or label encoding.

[0075] In a feasible implementation manner, after inputting the preprocessed load value and meteorological data into the large language model, the Transformer structure of the large language model can be used to perform feature extraction on the preprocessed load value and meteorological data to obtain a load value feature and a meteorological data feature, such as the change rate of the load, the moving average of the meteorological data, etc., and these features may help the model better understand the relationship between the data.

[0076] Step S22, determine the target prompt template associated with the load value feature and the meteorological data feature according to the similarity between the load value feature, the meteorological data feature, and each preset prompt template;

[0077] It should be noted that multiple prompt templates can be preset and stored in advance.

[0078] In a feasible embodiment, the preset slot value features of each slot corresponding to each preset prompt template can be obtained; the similarities between the load value feature, the meteorological data feature and the preset slot value features of each preset prompt template are determined, and the preset prompt template corresponding to the maximum similarity is determined as the target prompt template.

[0079] Step S23, embedding the load value feature and the meteorological data feature into the target prompt template to obtain the power load description text.

[0080] The slot values of the slots corresponding to the target prompt template are updated by using the load value feature and the meteorological data feature to obtain the power load description text.

[0081] Exemplarily, when given the input of load value, meteorological data and prompt information, the large language model first analyzes this input to understand its meaning and context. The model will extract the features in the input text, and these features may include vocabulary, part of speech, syntactic structure, semantic relationship, etc. Probability prediction: Based on the extracted features and the knowledge learned by the model, the large language model will assign a probability to each possible output word or sentence. This probability reflects the possibility of the word or sentence appearing in the current context. The model will select the most appropriate word or sentence according to the probability distribution to generate the description text. This process may be word-by-word generation such as an autoregressive text generation model, or it may be to generate the entire text at once such as an autoencoder text generation model, so as to generate the power load description text.

[0082] In this embodiment, the large language model processes the load value and the meteorological data to obtain the power load description text. The large language model can generate descriptions about weather conditions, power loads, etc., and these descriptions may be related to the changes in power loads and meteorological data during historical power generation periods. Subsequently, by analyzing these data, the accuracy of the prediction results can be improved.

[0083] Based on the above embodiments, in the third embodiment of the present application, referring to Figure 3 , the power load prediction model includes a time series model and a neural network model. Here, the time series model can be one of an ARMA model, an ARIMA model, a Holt-Winters model, etc., and the neural network model can be one of a feedforward neural network model, a convolutional neural network model, a recurrent neural network model, etc. Step S30 includes steps S31 to S33:

[0084] Step S31: Extract features from the power load description text to obtain the power load impact features of the microgrid system;

[0085] It should be noted that keyword extraction can be performed on the power load description text, and the power load impact features of the microgrid system can be constructed based on the extracted keywords. Among them, the power load impact features can be meteorological features, load features, etc. Here, the load features can be peak load, load rate, load change situation, etc.

[0086] Step S32: Determine the first prediction result of the time series model based on the power load impact features, and determine the second prediction result of the neural network model based on the power load impact features;

[0087] It should be noted that the first prediction result is obtained by the time series model prediction, and the second prediction result is obtained by the neural network model prediction.

[0088] In a feasible implementation manner, determining the first prediction result of the time series model based on the power load impact features includes: an appropriate time series model can be selected according to the characteristics of the data, the accuracy requirements of the prediction, and the limitations of the computing resources, etc. For example, the ARIMA model is suitable for short-term prediction and the case where the data is relatively stable. The LSTM model is good at dealing with long-term dependence problems and is suitable for power load prediction with complex time series characteristics. Input the power load impact features into the time series model to obtain the first prediction result; after obtaining the first prediction result, appropriate evaluation indicators can be used to evaluate the prediction result, such as mean square error, root mean square error, mean absolute percentage error, etc. At the same time, the prediction result can also be compared with the actual value to adjust the weights of the model, etc., to improve the prediction accuracy of the model.

[0089] Through the above steps, the first prediction result of the time series model based on the power load impact features can be determined, and at the same time, the second prediction result of the neural network model based on the power load impact features can be determined. By predicting with different models, different scenario requirements can be met.

[0090] Step S33: Determine the target power load prediction result of the microgrid system according to the first prediction result and the second prediction result.

[0091] In a feasible implementation manner, when the power load prediction results of the time series model and the deep learning model are the same, any one of the power load prediction results is determined as the target power load prediction result. Or, one of them can be selected as the target power load prediction result according to the actual situation.

[0092] In another feasible implementation, the weights of the time series model and the weights of the neural network model can be obtained; based on the product of the first prediction result and the weights of the time series model, and the sum of the product of the second prediction result and the weights of the neural network model, the target power load prediction result of the microgrid system is determined.

[0093] Among them, each power load prediction model has a corresponding weight, and the weight represents the importance of the prediction result of the power load prediction model to the target power load prediction result. The power load prediction model with more accurate prediction has a larger weight coefficient in the prediction stage. The weight coefficient corresponding to each power load prediction model can be determined in the training stage and can be determined according to the mean square error calculated from historical power load data. For example, the mean square error corresponding to each power load prediction model is calculated using the test data set respectively; the mean square error corresponding to each power load prediction model is normalized respectively to obtain the weights corresponding to each power load prediction model.

[0094] In this embodiment, multiple different power load prediction models are set simultaneously for prediction to meet the power load prediction requirements in different scenarios.

[0095] Further, before step S331, steps S110 to S130 are also included:

[0096] Step S110, obtaining historical power load data samples;

[0097] Step S120, dividing the historical power load data samples into a first training data set and a first test data set based on the training data length corresponding to the time series model, training an initial time series model based on the first training data set, and testing the initial time series model based on the first test data set to obtain the weights of the time series model; and

[0098] Step S130, dividing the historical power load data samples into a second training data set and a second test data set based on the training data length corresponding to the neural network model, training an initial neural network model based on the second training data set, and testing the initial neural network model based on the second test data set to obtain the weights of the neural network model.

[0099] It should be noted that the training data lengths of the time series model and the neural network model are different. Therefore, the training data sets and test data sets of the time series model and the neural network model are different. The time series model and the neural network model are trained separately.

[0100] It should be noted that during the training process, the number of training times, learning rate, etc. can also be set.

[0101] In a feasible embodiment, appropriate evaluation metrics can be used to evaluate the prediction results, such as mean squared error, root mean squared error, mean absolute percentage error, etc. The prediction accuracy of the model can be analyzed through these metrics to determine the weights of each prediction model.

[0102] In another feasible embodiment, the time series model and the neural network model can also be jointly trained to reduce the bias and variance of a single model by combining multiple models, thereby improving the overall prediction accuracy.

[0103] In this embodiment, the accuracy of each prediction model is improved by training different prediction models with different training data and test data.

[0104] Based on the third embodiment of the present application, in the fourth embodiment of the present application, referring to Figure 4 , the target power load prediction result includes the target net load, and step S40 includes S41 to S43:

[0105] Step S41, according to the target meteorological data, the target net load, and the corresponding relationship between the preset meteorological data, the preset load value, and the preset dispatching scheme, determine the power dispatching scheme;

[0106] The corresponding relationship between the meteorological data, the preset load value, and the preset dispatching scheme can be pre-constructed, and the corresponding power dispatching scheme can be determined based on this corresponding relationship.

[0107] In a feasible embodiment, based on the power load prediction model and the power load description text, predict the total power consumption and total power generation of the microgrid system in the future period, and determine the target power load prediction result according to the difference between the total power consumption and total power generation in the future period. That is, the target power load prediction result can be the target net load value.

[0108] Step S42, according to the power dispatching scheme, determine the power generation amount of each power generation node, and determine the power generation power of each power generation node according to the power generation amount of each power generation node;

[0109] Step S43, control each of the power generation nodes to operate according to the corresponding power generation power.

[0110] The power dispatching scheme includes but is not limited to the type of power generation nodes that need to be started and operated, the number of power generation nodes, the power generation power of each power generation node, the power generation duration of each power generation node, etc. The power dispatching schemes corresponding to different meteorological data and load values are different. The power generation amount of each power generation node can be obtained by parsing the power dispatching scheme.

[0111] After determining the power generation of each power generation node, determine the power generation power required for each power generation node to reach the power generation amount, and control each power generation node to operate according to the corresponding power generation power, so that when each power generation node operates according to the corresponding power generation power, the microgrid system achieves a power balance state.

[0112] Based on the above embodiments of the present application, in the fifth embodiment of the present application, refer to Figure 5 , the microgrid system includes multiple power generation nodes and multiple power consumption nodes. The power generation nodes are responsible for generating electricity, which may include coal-fired power plants, nuclear power plants, wind turbines, solar panels, etc. The power consumption nodes are the ultimate users of electricity, such as households, commercial facilities, and industrial users. These devices and facilities consume electricity to meet various needs. Since the power generation nodes and power consumption nodes each play important tasks and responsibilities in the microgrid system, they cooperate and work together to jointly ensure the stable operation and efficient energy utilization of the microgrid system, and can ensure the safety, reliability, and economy of the microgrid system.

[0113] Specifically, step S10 includes S11 to S13:

[0114] Step S11, determine the total power generation according to the power generation of each power generation node of the microgrid system during the historical power generation period, and determine the total power consumption according to the power consumption of each power consumption node of the microgrid system during the historical power generation period;

[0115] It should be noted that the historical power generation period is a period of historical power generation duration of the microgrid system, such as 3 hours, 4 hours, etc. This embodiment does not limit this.

[0116] Regarding the acquisition of power generation: For solar and wind power generation, there are usually special monitoring devices to record the power generation amount, and wind turbine generators and solar photovoltaic systems also have relevant real-time data.

[0117] Regarding the acquisition of power consumption: Distribution companies usually install monitoring devices to track the power consumption load of the distribution system in real time. These data can usually be obtained through a data acquisition and monitoring system. Distribution companies will regularly generate load reports to record the total power consumption in the area. For large-scale industrial users or commercial users, they usually have special energy management systems to record their actual power consumption. Households and small commercial users can obtain power consumption through smart meters, and these meters can provide detailed power consumption statistics.

[0118] In a feasible implementation, a power generation amount collection device and a power consumption amount collection device are set for the microgrid system. One end of the power generation amount collection device is connected to each power generation node, and the other end is connected to each power consumption node. The power generation amount collection device can count the power generation amounts of each power generation node to obtain the total power generation amount, and can also count the power consumption amounts of each power consumption node to obtain the total power consumption amount.

[0119] Step S12: Determine the net load of the microgrid system during the historical power generation period according to the total power consumption amount and the total power generation amount, where the power load data further includes the net load.

[0120] It should be noted that the power load data includes the net load, and the net load is the amount of electric power that the system needs to supply additionally or can output additionally during the historical power generation period.

[0121] In a feasible implementation, the net load of the microgrid system during the historical power generation period can be obtained according to the difference between the total power consumption amount and the total power generation amount.

[0122] Step S13: When the net load is less than the preset value, determine that the power state is a power surplus state.

[0123] It should be noted that the net load has positive and negative values. If the net load is positive, it means that the system needs additional power to meet the power consumption demand, and at this time, the power state of the microgrid system should be a shortage state; if the net load is negative, it means that the system has surplus power that can be output or stored, and at this time, the power state of the microgrid system should be a surplus state. Therefore, a preset value can be set. When the net load is less than the preset value, the power state is determined to be a power surplus state, and when the net load is greater than the preset value, the power state is determined to be a power shortage, where the preset value can be set to 0.

[0124] In this embodiment, the power state of the microgrid system can be analyzed according to the value of the net load, which is convenient for subsequent control related to the power state of the microgrid system.

[0125] It should be noted that the above examples are only for understanding the present application and do not constitute a limitation on the power control method of the microgrid system of the present application. Based on this technical concept, more forms of simple transformations are within the protection scope of the present application.

[0126] The present application provides a power control device for a microgrid system. The power control device 1 of the microgrid system includes: at least one processor 1001; and a memory 1002 communicatively connected to the at least one processor 1001, etc.; where the memory 1002 stores instructions executable by the at least one processor 1001, and the instructions are executed by the at least one processor 1001 so that the at least one processor 1001 can execute the power control method of the microgrid system in the following embodiments.

[0127] As Figure 6 shown, the power control device 1 of the microgrid system may include a processor 1001, which may perform various appropriate actions and processes according to a program stored in a memory 1002. Here, the program in the memory 1002 may be a program in a read-only memory (ROM: Read Only Memory) or a program loaded from a storage device into a random access memory (RAM: Random Access Memory). In the RAM, various programs and data required for the operation of the power control device 1 of the microgrid system are also stored. The processor 1001 and the memory 1002 (ROM and RAM) are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus. Generally, the following systems may be connected to the I / O interface: input devices including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; output devices including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; storage devices including, for example, a magnetic tape, a hard disk, etc.; and a communication device. The communication device may allow the power control device 1 of the microgrid system to communicate with other devices wirelessly or wiredly to exchange data. Although the power control device 1 of the microgrid system with various hardware is shown in the figure, it should be understood that it is not required to implement or have all the shown hardware, and more or less hardware may be alternatively implemented or had.

[0128] The power control device of the microgrid system provided in this application adopts the power control method of the microgrid system in the above embodiment, can perform power control from the root to achieve power balance, so that there is no need to grid the surplus power, and the power waste caused by line transmission loss can be avoided. Compared with the prior art, the beneficial effects of the power control device of the microgrid system provided in this application are the same as those of the power control method of the microgrid system provided in the above embodiment, and other technical features in the power control device of the microgrid system are the same as the features disclosed in the method of the previous embodiment, and will not be elaborated here.

[0129] It should be understood that each part disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0130] The above is only the specific implementation manner of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in this application, and all should be covered by the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

[0131] The present application provides a microgrid system, which includes a plurality of power generation nodes, a plurality of power consumption nodes, and a power control device of the microgrid system. Each of the power generation nodes and each of the power consumption nodes are connected to the power control device of the microgrid system.

[0132] The microgrid system provided by the present application includes the power control device of the microgrid system in the above embodiment, and can perform power control from the root to achieve power balance, so that there is no need to grid the surplus power, avoiding power waste caused by line transmission loss. Compared with the prior art, the beneficial effects of the microgrid system provided by the present application are the same as those of the power control method of the microgrid system provided by the above embodiment, and other technical features in this microgrid system are the same as those disclosed in the method of the previous embodiment, which will not be elaborated here.

[0133] The present application provides a computer-readable storage medium having computer-readable program instructions (i.e., computer programs) stored thereon, and the computer-readable program instructions are used to execute the power control method of the microgrid system in the above embodiment.

[0134] The computer-readable storage medium provided by the present application may be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: electrical connections with one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM) or flash memory, optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In this embodiment, the computer-readable storage medium may be any tangible medium that contains or stores a program, and this program can be used by or combined with an instruction execution system, device, or device. The program code contained on the computer-readable storage medium can be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination of the above.

[0135] The above computer-readable storage medium may be included in the power control device of the microgrid system; or it may exist separately and not be assembled into the power control device of the microgrid system.

[0136] The above computer-readable storage medium carries one or more programs. When the one or more programs are executed by a power control device of a microgrid system, the power control device of the microgrid system is enabled to: obtain power load data of the microgrid system during a historical power generation period, and determine the power state of the microgrid system according to the power load data; when the power state is a power surplus state, generate a power load description text based on a large language model, the power load data, and meteorological data of the historical power generation period; predict a target power load prediction result of the microgrid system based on a power load prediction model and the power load description text; determine a power dispatching scheme according to the target power load prediction result and target meteorological data, and dispatch the operation of corresponding power generation nodes based on the power dispatching scheme to achieve a power balance state.

[0137] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The above programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0138] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in the flowchart or block diagram may represent a module, a program segment, or a part of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and the combination of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0139] The modules involved in the embodiments of the present application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation to the unit itself in some cases.

[0140] The readable storage medium provided by the present application is a computer-readable storage medium, which stores computer-readable program instructions (i.e., computer programs) for executing the power control method of the above-mentioned microgrid system. It can perform power control from the root to achieve power balance, so that there is no need to grid the surplus power, avoiding power waste caused by line transmission loss. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as those of the power control method of the microgrid system provided in the above embodiments, and will not be elaborated here.

[0141] The above are only some embodiments of the present application, and do not limit the patent scope of the present application. Any equivalent structural transformation made by using the content of the specification and drawings of the present application under the technical concept of the present application, or direct / indirect application in other related technical fields, is included in the patent protection scope of the present application.

Claims

1. A power control method for a microgrid system, characterized in that: The method comprises: Acquire power load data of the microgrid system in a historical power generation period, and determine the power state of the microgrid system according to the power load data, wherein the power load data includes a load value; When the power state is a power surplus state, the load value and the meteorological data in the historical power generation period are input into a large language model, and feature extraction is performed on the load value and the meteorological data respectively to obtain load value features and meteorological data features; Determine a target prompt template associated with the load value feature and the meteorological data feature according to similarities between the load value feature, the meteorological data feature and each preset prompt template; Embedding the load value feature and the meteorological data feature into the target prompt template to obtain the power load description text; Extracting features from the power load description text to obtain power load impact features of the microgrid system, wherein the power load impact features include meteorological features and load features, and the load features include peak load, load rate, and load change; Determine a first prediction result of a time series model based on the power load impact feature, and determine a second prediction result of a neural network model based on the power load impact feature; Determine a target power load prediction result of the microgrid system according to the first prediction result and the second prediction result, wherein the target power load prediction result includes a target net load; A power dispatching plan is determined according to the target power load forecast result and the target meteorological data, and the corresponding power generation nodes are dispatched to operate based on the power dispatching plan to achieve a power balance state.

2. The method according to claim 1, characterized in that Determining the target power load prediction result of the microgrid system according to the first prediction result and the second prediction result includes: Obtaining the weight of the time series model and the weight of the neural network model; The target power load prediction result of the microgrid system is determined based on the product of the first prediction result and the weight of the time series model, and based on the sum of the products of the second prediction result and the weight of the neural network model.

3. The method according to claim 2, characterized in that Before the step of obtaining the weight of the time series model and the weight of the neural network model, the method further includes: Obtain historical power load data samples; Dividing the historical power load data samples into a first training data set and a first test data set based on the training data length corresponding to the time series model, training an initial time series model based on the first training data set, and testing the initial time series model based on the first test data set to obtain a weight of the time series model; and The historical power load data samples are divided into a second training data set and a second test data set based on the training data length corresponding to the neural network model, an initial neural network model is trained based on the second training data set, and the initial neural network model is tested based on the second test data set to obtain the weight of the neural network model.

4. The method according to claim 1, characterized in that Determining a power dispatching scheme according to the target power load forecast result and the target meteorological data, and dispatching the corresponding power generation nodes to operate based on the power dispatching scheme includes: Determine the power dispatching plan according to the target meteorological data, the target net load, and the corresponding relationship between the preset meteorological data, the preset load value and the preset dispatching plan; According to the power dispatching plan, the power generation of each power generation node is determined, and the power generation power of each power generation node is determined according to the power generation of each power generation node; Control each of the power generation nodes to operate according to the corresponding power generation power.

5. The method according to claim 1, characterized in that The microgrid system includes a plurality of power generation nodes and a plurality of power consumption nodes. The step of obtaining power load data of the microgrid system in a historical power generation period and determining the power state of the microgrid system according to the power load data includes: Determine the total power generation according to the power generation of each power generation node of the microgrid system during the historical power generation period, and determine the total power consumption according to the power consumption of each power consumption node of the microgrid system during the historical power generation period; Determine the net load of the microgrid system in the historical power generation period according to the total power consumption and the total power generation, wherein the power load data also includes the net load; When the net load is less than a preset value, the power state is determined to be a power surplus state.

6. A power control device for a microgrid system, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the power control method for the microgrid system according to any one of claims 1 to 5.

7. A microgrid system, characterized in that: The microgrid system comprises a plurality of power generation nodes, a plurality of power consumption nodes and a power control device of the microgrid system as claimed in claim 6, and each of the power generation nodes and each of the power consumption nodes are connected to the power control device of the microgrid system.

8. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, the steps of the power control method of the microgrid system according to any one of claims 1 to 5 are implemented.

Images