Power load control method and system for virtual power plant

Abstract

The present invention relates to a power load control method and system for a virtual power plant. The method comprises: acquiring power load data and corresponding power supply capability data of a virtual power plant, and on the basis of the power load data and the corresponding power supply capability data, generating a control strategy training sample set; on the basis of the stability of a power system, analyzing each piece of power supply capability data in the training sample set, so as to obtain a plurality of pieces of supply capability sub-data of each power supply capability; constructing a power load prediction and control model, and on the basis of the power load data and the power supply capability data in the training sample set and the plurality of pieces of supply capability sub-data of each power supply capability, training the power load prediction and control model to obtain a power load control model; and inputting power load data of a virtual power plant to be controlled into the power load control model to obtain a power load control strategy for said virtual power plant. Therefore, the reliability and flexibility of power supply are improved.

Description

A Virtual Power Plant Power Load Control Method and System Technical Field

[0001] This invention relates to the field of power system management, and in particular to a method and system for controlling power load in a virtual power plant. Background Technology

[0002] In existing power systems, virtual power plants, as a novel power resource management approach, integrate distributed generation resources, energy storage devices, and controllable loads to achieve optimized allocation and dispatch of power resources. Virtual power plants can effectively improve the operational efficiency and reliability of power systems, reduce operating costs, and promote the utilization of renewable energy. However, with the continuous development of the electricity market and the increasing complexity of electricity demand, traditional power load control methods are no longer sufficient to meet the requirements of modern power systems for flexibility and response speed.

[0003] Existing power load control methods typically rely on static load forecasting models and simple control strategies. These methods often fail to accurately predict dynamic changes in electricity demand and are ill-suited to adapting to fluctuations in power supply capacity. Furthermore, traditional control strategies lack in-depth analysis of power system stability, making it difficult to optimize economic and environmental impacts while ensuring power supply reliability. In terms of power supply capacity analysis, existing technologies often neglect the segmentation and dynamic adjustment of power supply capacity, resulting in the inability to fully utilize the potential of power resources and a failure to cope with rapid changes in the electricity market.

[0004] In the process of implementing the embodiments of the present invention, the inventors have discovered that the prior art has at least the following problems or defects: when processing power load and supply capacity data, the existing methods lack effective data preprocessing and feature extraction methods, resulting in low model training efficiency and limited prediction accuracy; at the same time, the existing technology fails to fully consider the stability of the power system and the multi-objective optimization requirements in power load control, resulting in insufficient adaptability and comprehensive performance of the control strategy. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] The purpose of this invention is to provide a virtual power plant power load control method and system, which achieves efficient, stable and economical operation of the power system through a series of innovative training methods and technical strategies, while reducing environmental impact and improving the reliability and flexibility of power supply.

[0007] In a first aspect, embodiments of the present invention relate to a virtual power plant power load control method, the method comprising:

[0008] Obtain the power load data and corresponding power supply capacity data of the virtual power plant, and generate a control strategy training sample set based on the power load data and corresponding power supply capacity data;

[0009] Based on the stability of the power system, the data of each power supply capacity in the training sample set is analyzed to obtain multiple sub-supply capacity data for each power supply capacity;

[0010] A power load prediction and control model is constructed. The power load prediction and control model is trained based on the power load data, power supply capacity data, and multiple sub-supply capacity data of each power supply capacity in the training sample set to obtain the power load control model.

[0011] The power load data of the virtual power plant to be controlled is input into the power load control model to obtain the power load control strategy of the virtual power plant to be controlled.

[0012] Furthermore, based on the stability of the power system, the data for each power supply capacity in the training sample set are analyzed to obtain multiple sub-supply capacity data for each power supply capacity, including:

[0013] Take the initial state of each power supply capacity as the first type and calculate the stability of the state; iterate through each state of the power supply capacity in turn and calculate the stability of the current state. If the stability of the current state is the same as the stability of the previous state, then the current state and the previous state are of the same type; otherwise, the current state and the previous state are of different types.

[0014] Based on the type of each state in the power supply capacity, multiple sub-supply capacity data of the power supply capacity are obtained.

[0015] Furthermore, based on the sub-supply capacity data of power supply capacity, a composite loss function with load forecasting error penalty is used to calculate the training loss of the power load forecasting and control model.

[0016] Furthermore, the load forecasting error penalty matrix corresponding to the power supply matrix is ​​calculated through the following steps:

[0017] Calculate the Euclidean distance matrix (EDM) within the m-th sub-mask matrix based on the electricity demand matrix. m Based on the Euclidean distance matrix EDM within the mask m Calculate the in-mask penalty term of the m-th sub-mask matrix. int,m ;

[0018] The external Euclidean distance matrix (EDM) of the m-th sub-mask matrix is ​​calculated based on the power supply matrix. ext,m Based on the Euclidean distance matrix EDM outside the mask ext,m Calculate the out-of-mask penalty term of the m-th sub-mask matrix. ext,m ;

[0019] Penalty within the mask int,m and the penalty item outside the mask ext,m Normalizing and summing the results yields the distance penalty matrix (penal) corresponding to the m-th sub-mask matrix. m ;

[0020] The distance penalty matrix corresponding to each sub-mask matrix is ​​penal m The load forecasting error penalty matrix Penal is obtained by splicing the power supply matrix.

[0021] Furthermore, the Euclidean distance matrix EDM within the mask of the m-th sub-mask matrix m It has the same dimension as the m-th sub-mask matrix of the electricity demand matrix;

[0022] Calculate the in-mask penalty term of the m-th submask matrix.

[0023] Calculate the out-of-mask penalty term for the m-th sub-mask matrix.

[0024] Furthermore, the external Euclidean distance matrix EDM of the m-th sub-mask matrix ext,m It has the same dimension as the m-th sub-mask matrix of the power supply matrix;

[0025] Secondly, embodiments of the present invention relate to a virtual power plant power load control system, comprising:

[0026] The feature extraction module is used to extract features from power load data and power supply capacity data using a deep learning network;

[0027] The prediction module is used to predict future power load based on extracted features using a time series prediction algorithm;

[0028] The optimization module is used to generate power load control strategies based on prediction results and power supply capacity data using optimization algorithms;

[0029] The evaluation module is used to assess the performance of power load control strategies, including economic efficiency, reliability, and environmental impact.

[0030] Furthermore, a control strategy training sample set is generated based on the power load data and the corresponding power supply capacity data, including:

[0031] Gaussian process regression was used to interpolate the power load data and the corresponding power supply capacity data to handle missing data;

[0032] Wavelet transform was performed on power load data and power supply capacity data to extract time-frequency features;

[0033] For each sample data, principal component analysis is performed on the power load data and power supply capacity data respectively to reduce data dimensionality and improve model training efficiency.

[0034] Based on the dimensionality-reduced data, a clustering algorithm is used to classify the power load type and power supply capacity type to generate training samples.

[0035] The embodiments disclosed above have the following beneficial effects: The virtual power plant power load control method of the present invention can improve the operating efficiency and stability of the power system. Through feature extraction using deep learning networks, combined with time series prediction algorithms, future power load can be predicted more accurately, thereby making the power load control strategy more precise and timely. Furthermore, the power load control strategy generated by the optimized algorithm can optimize economic efficiency, reliability, and environmental impact while ensuring the reliability of power supply, achieving efficient utilization of power resources.

[0036] Furthermore, by introducing a composite loss function with load prediction error penalty, this invention can effectively reduce prediction errors during model training and improve the accuracy of power load prediction. Simultaneously, by calculating the load prediction error penalty matrix corresponding to the power supply matrix, it can better handle the uncertainty and variability in power supply capacity data, making power load control strategies more flexible and adaptable. The application of these techniques can not only improve the response speed and regulation capacity of the power system but also reduce operating costs, enhancing the economic efficiency and environmental friendliness of the power system.

[0037] Furthermore, the above summary does not enumerate all the features required for embodiments of the present invention, and other combinations of these feature groups may also constitute embodiments of the present invention. Attached Figure Description

[0038] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. Several embodiments of the invention are illustrated in the drawings by way of example and not limitation, wherein:

[0039] Figure 1 is a flowchart illustrating a virtual power plant power load control method according to an embodiment of the present invention.

[0040] Figure 2 is a schematic diagram of the structure of a virtual power plant power load control system provided in one embodiment of the present invention;

[0041] Figure 3 schematically illustrates the structure of a computing device according to an embodiment of the present invention;

[0042] In the accompanying drawings, the same or corresponding reference numerals indicate the same or corresponding parts. Detailed Implementation

[0043] The principles and spirit of the invention will now be described with reference to several exemplary embodiments. It should be understood that these embodiments are given merely to enable those skilled in the art to better understand and implement the invention, and are not intended to limit the scope of the invention in any way. Rather, these embodiments are provided to make this disclosure more thorough and complete, and to fully convey the scope of this disclosure to those skilled in the art.

[0044] Those skilled in the art will recognize that embodiments of the present invention can be implemented as a system, apparatus, device, method, or computer program product. Therefore, this disclosure can be specifically implemented in the following forms: entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0045] According to an embodiment of the present invention, a method and system for controlling the power load of a virtual power plant are proposed.

[0046] It should be noted that the number of any elements in the accompanying drawings is for illustrative purposes only and not as a limitation, and any naming is for distinction only and has no limiting meaning.

[0047] The principles and spirit of the present invention will be explained in detail below with reference to several representative embodiments.

[0048] Referring to Figure 1 below, Figure 1 is a schematic flowchart of a virtual power plant power load control method provided by an embodiment of the present invention. It should be noted that the embodiments of the present invention can be applied to any applicable scenario.

[0049] Figure 1 shows a virtual power plant power load control method provided by an embodiment of the present invention. The method 100 includes:

[0050] Step 101: Obtain the power load data and corresponding power supply capacity data of the virtual power plant, and generate a control strategy training sample set based on the power load data and corresponding power supply capacity data;

[0051] Step 102: Analyze the power supply capacity data of each power supply capacity in the training sample set based on the stability of the power system to obtain multiple sub-supply capacity data for each power supply capacity;

[0052] Step 103: Construct an electricity load prediction and control model. The electricity load prediction and control model is trained based on the electricity load data, electricity supply capacity data, and multiple sub-supply capacity data of each electricity supply capacity in the training sample set to obtain an electricity load control model.

[0053] Step 104: Input the power load data of the virtual power plant to be controlled into the power load control model to obtain the power load control strategy of the virtual power plant to be controlled.

[0054] It should be noted that this embodiment relates to a virtual power plant power load control method. This method acquires the power load data and corresponding power supply capacity data of the virtual power plant, and generates a control strategy training sample set based on this data. This allows the system to accurately match power supply and demand, optimizing the allocation and use of power resources.

[0055] Specifically, acquiring power load and power supply capacity data involves collecting real-time data from various nodes of the virtual power plant, including but not limited to power generation, power consumption, and energy storage status. This data is collected through smart meters, sensors, and data acquisition systems. The generation of the control strategy training sample set relies on data preprocessing, feature extraction, and label definition to ensure that the model can learn effective control strategies from historical data.

[0056] Preferably, to improve data quality and availability, data cleaning and missing value handling techniques, such as interpolation and time series forecasting, can be employed to ensure the integrity and accuracy of the training sample set. Furthermore, machine learning algorithms, such as random forests, support vector machines, or neural networks, can be used to construct power load forecasting and control models, and methods such as cross-validation can be used to optimize model parameters and improve the model's generalization ability.

[0057] In some embodiments, the power supply capacity data for each power supply capacity in the training sample set is analyzed based on the stability of the power system to obtain multiple sub-supply capacity data for each power supply capacity, including:

[0058] Take the initial state of each power supply capacity as the first type and calculate the stability of the state; iterate through each state of the power supply capacity in turn and calculate the stability of the current state. If the stability of the current state is the same as the stability of the previous state, then the current state and the previous state are of the same type; otherwise, the current state and the previous state are of different types.

[0059] Based on the type of each state in the power supply capacity, multiple sub-supply capacity data of the power supply capacity are obtained.

[0060] Specifically, stability analysis of power supply capacity involves assessing the initial state of each power supply capacity and calculating its stability index. Stability can be measured in various ways, such as based on power supply volatility, frequency response characteristics, or supply-demand balance. In practical applications, a series of parameters can be set to quantify stability, such as supply fluctuation thresholds and frequency deviation tolerance. These parameters determine the stability of each state and, accordingly, classify the power supply capacity into different sub-supply capacities.

[0061] Preferably, to conduct more accurate stability analysis of power supply capacity, advanced data analysis techniques, such as machine learning algorithms, can be employed to automatically identify and classify the stability state of power supply capacity. Furthermore, adaptive algorithms can be introduced to dynamically adjust the parameters of stability analysis based on real-time data to adapt to changes in power system operating conditions. When classifying power supply capacity, in addition to stability, other factors such as cost-effectiveness and environmental impact can also be considered to achieve more comprehensive power supply capacity management.

[0062] In some embodiments, the training loss of the power load forecasting and control model is calculated using a composite loss function with load forecasting error penalty, based on sub-supply capacity data of power supply capacity.

[0063] The formula for calculating the composite loss function is as follows:

[0064] Where L represents the total loss value, N represents the total amount of power supply capacity data, and w i PD represents the weight of the i-th power supply capacity. i Let PS represent the i-th electricity demand matrix. i Denotes the i-th power supply matrix, penal i Let represent the load forecasting error penalty matrix corresponding to the i-th power supply matrix, λ represent the regularization parameter, and ||·|| represent the Frobenius norm of the matrix.

[0065] Specifically, the calculation of the composite loss function involves multiple parameters and concepts. The total loss value, the objective function to be minimized during model training, consists of several parts, including the weights of the power supply capacity data, the power demand matrix, the power supply matrix, and the load forecasting error penalty matrix. These parameters can be set using historical data and expert knowledge. For example, the weights can be allocated based on the importance of power supply capacity, while the forecasting error penalty matrix can be adjusted according to the magnitude of the forecasting error.

[0066] Preferably, to further improve the training performance of the model, various methods can be used to refine the calculation of the composite loss function. For example, a regularization parameter can be introduced to prevent overfitting, or different distance metrics can be used to calculate the prediction error. Furthermore, different optimization algorithms can be considered to solve for the minimum value of the loss function, such as gradient descent, stochastic gradient descent, or the Adam optimizer. These methods can be selected based on the complexity of the model and the characteristics of the training data to achieve the best training results.

[0067] In some embodiments, the load forecasting error penalty matrix corresponding to the power supply matrix is ​​calculated through the following steps:

[0068] Calculate the Euclidean distance matrix (EDM) within the m-th sub-mask matrix based on the electricity demand matrix. m Based on the Euclidean distance matrix EDM within the mask m Calculate the in-mask penalty term of the m-th sub-mask matrix. int,m ;

[0069] The external Euclidean distance matrix (EDM) of the m-th sub-mask matrix is ​​calculated based on the power supply matrix. ext,m Based on the Euclidean distance matrix EDM outside the mask ext,m Calculate the out-of-mask penalty term of the m-th sub-mask matrix. ext,m ;

[0070] Penalty within the mask int,m and the penalty item outside the mask ext,m Normalizing and summing the results yields the distance penalty matrix (penal) corresponding to the m-th sub-mask matrix. m ;

[0071] The distance penalty matrix corresponding to each sub-mask matrix is ​​penal m The load forecasting error penalty matrix Penal is obtained by splicing the power supply matrix.

[0072] It should be noted that this implementation describes in detail how to calculate the load forecasting error penalty matrix corresponding to the power supply matrix. This step involves analyzing the sub-mask matrix of the power demand matrix, calculating the Euclidean distance matrix inside and outside the mask, generating penalty terms accordingly, and finally integrating these penalty terms into a single matrix for use in the model training process.

[0073] Specifically, the sub-mask matrix of the power supply matrix refers to dividing the power supply matrix into multiple regions or subsets, with each subset corresponding to a sub-mask matrix. The in-mask Euclidean distance matrix calculates the shortest distance from each point within the sub-mask matrix to the mask boundary. These distances can be determined using computational geometry methods, such as Dijkstra's algorithm or the Floyd-Warshall algorithm. The in-mask penalty term is calculated based on these distances and can be defined using different mathematical functions, such as exponential or linear functions. The calculation of the out-of-mask Euclidean distance matrix and the out-of-mask penalty term is similar to that of the in-mask matrix, but focuses on the region outside the sub-mask matrix.

[0074] Preferably, to more accurately calculate the penalty terms inside and outside the mask, more advanced numerical analysis methods, such as Gaussian blurring or machine learning algorithms, can be employed to optimize distance calculation and penalty term generation. Furthermore, an adaptive mechanism can be introduced to dynamically adjust the weights of the penalty terms based on the model's performance during training, thereby improving the model's generalization ability and prediction accuracy. In practical applications, real-time data and historical trends of the power system can also be considered to adjust the penalty terms inside and outside the mask in real time to adapt to changes in the power system's operating state. These methods can further improve the model's accuracy and robustness in predicting power load.

[0075] In some embodiments, the in-mask Euclidean distance matrix EDM of the m-th sub-mask matrix m The m-th sub-mask matrix has the same dimension as the power demand matrix; the in-mask Euclidean distance matrix EDM of the m-th sub-mask matrix is... m The value of the element at position (i,j,k) is:

[0076] Where dist(i,j,k) represents the shortest Euclidean distance from point (i,j,k) to the mask boundary of the submask matrix, Mask m This represents the mask region of the m-th sub-mask matrix.

[0077] Specifically, the Euclidean distance matrix within the submask matrix is ​​a matrix with the same dimensions as the submask matrix of the electricity demand matrix. Each element represents the shortest Euclidean distance from the corresponding point to the boundary of the submask region. This distance can be determined using computational geometry methods, such as the Bresenham algorithm or Dijkstra's algorithm. In practical applications, the calculation of this distance needs to consider the topology of the power network and the characteristics of power flow.

[0078] Preferably, to improve the efficiency and accuracy of calculating the Euclidean distance matrix within the mask, parallel computing techniques can be employed to process large-scale electricity demand matrices. Furthermore, adaptive algorithms can be introduced to optimize the distance calculation process, for example, automatically adjusting calculation parameters based on dynamic changes in electricity demand. In some cases, machine learning methods can also be used to predict and estimate distances within the mask, especially when data is sparse or computational resources are limited.

[0079] In some embodiments, according to the formula penal int,m =(max(EDM) m )-EDM m )⊙GT m

[0080] Calculate the in-mask penalty term of the m-th submask matrix, where EDM m Let GT represent the Euclidean distance matrix within the mask of the m-th sub-mask matrix. m Let m represent the m-th submask matrix, max(·) means replacing all elements of the matrix with the maximum element value in the matrix, and ⊙ means multiplying corresponding elements of the matrix.

[0081] It should be noted that this implementation describes in detail how to calculate the in-mask penalty terms of the sub-mask matrix. This step involves combining the in-mask Euclidean distance matrix with the sub-mask matrix and using specific mathematical operations to generate penalty terms, which will be used in the subsequent construction of the load prediction error penalty matrix.

[0082] Specifically, the calculation of the in-mask penalty term involves the in-mask Euclidean distance matrix and the sub-mask matrix. The in-mask Euclidean distance matrix is ​​a matrix containing the distance of each point to the nearest mask boundary, while the sub-mask matrix is ​​a binary matrix used to identify points inside and outside the mask area.

[0083] More specifically, when calculating the penalty term within the mask, a suitable mathematical function needs to be determined to process the two matrices. For example, a maximum value replacement function can be used to replace all elements in the Euclidean distance matrix within the mask with the maximum value, and then element-wise multiplied with the sub-mask matrix. This operation can highlight points within the mask region, providing more accurate error penalty information for load prediction.

[0084] Preferably, to improve the flexibility and adaptability of the penalty term calculation within the mask, a parameterized method can be introduced to adjust the penalty term calculation. For example, a penalty coefficient can be set to adjust the penalty intensity of points within the mask.

[0085] Furthermore, different mathematical operations, such as weighted averages or exponential functions, can be employed to enhance the model's sensitivity to prediction errors for different types of loads. In practical applications, these parameters can be optimized using methods such as cross-validation based on the specific needs of the power system and historical data to achieve optimal prediction performance. These improvements enable the model to more accurately reflect the dynamic changes in power load, thereby improving the operating efficiency and stability of the power system.

[0086] In some embodiments, according to the formula penal ext,m =(max(EDM) ext,m )-EDM ext,m )⊙(1-GT m )

[0087] Calculate the out-of-mask penalty term of the m-th sub-mask matrix, where EDM ext,m Let GT represent the Euclidean distance matrix outside the mask of the m-th sub-mask matrix. m Let m represent the m-th submask matrix, max(·) means replacing all elements of the matrix with the maximum element value in the matrix, and ⊙ means multiplying corresponding elements of the matrix.

[0088] Specifically, the calculation of the out-of-mask penalty involves the out-of-mask Euclidean distance matrix and the sub-mask matrix. The out-of-mask Euclidean distance matrix is ​​a matrix that records the distance of each point to the nearest mask boundary, but only for points located outside the mask region. The sub-mask matrix is ​​an identifier matrix used to distinguish points inside and outside the mask region.

[0089] Furthermore, when calculating the out-of-mask penalty term, a specific mathematical function can be used. For example, all elements in the out-of-mask Euclidean distance matrix can be replaced with the maximum value using a maximum value replacement function, and then multiplied element-wise with the complement of the sub-mask matrix, i.e., 1 minus the sub-mask matrix. This operation can effectively identify and process points outside the mask region, providing the model with the necessary error penalty information.

[0090] Preferably, to improve the computational efficiency and accuracy of the penalty term outside the mask, various strategies can be employed to optimize this process. For example, a dynamically adjusted penalty coefficient can be introduced, which can be adjusted based on the real-time operating status and historical data of the power system.

[0091] Furthermore, advanced numerical analysis methods, such as machine learning algorithms, can be considered to predict and estimate distances outside the mask, especially in situations where data is sparse or computational resources are limited. These methods can further improve the model's response speed and prediction accuracy to changes in power load, thereby optimizing the operating efficiency and stability of the power system.

[0092] In some embodiments, the external Euclidean distance matrix EDM of the m-th sub-mask matrix ext,m The dimension of the m-th sub-mask matrix is ​​the same as that of the power supply matrix; the outer Euclidean distance matrix EDM is... ext,m The value of the element at position (i,j,k) is:

[0093] Where dist(i,j,k) represents the shortest Euclidean distance from point (i,j,k) to the mask boundary of the submask matrix, Mask m This represents the mask region of the m-th sub-mask matrix.

[0094] It should be noted that this embodiment describes in detail how to calculate the Euclidean distance matrix outside the submask matrix. This matrix is ​​a key component of the power supply matrix, used to determine the shortest distance from each point to the boundary of the submask region, especially for points located outside the mask region.

[0095] Specifically, calculating the Euclidean distance matrix outside the mask involves analyzing each point in the power supply matrix to determine their distance to the nearest mask boundary. This typically involves complex geometric calculations that can be implemented using various algorithms, such as Dijkstra's algorithm or the A* algorithm.

[0096] Furthermore, in practical applications, the mask region can be defined by the topology of the power network, the path of power flow, or other relevant factors. Each element of the external Euclidean distance matrix is ​​determined by the shortest distance from a point to the mask boundary; if the point is not within the mask region, the value is 0.

[0097] Preferably, to improve the efficiency and accuracy of calculating the Euclidean distance matrix outside the mask, parallel computing techniques can be used to process large-scale power supply matrices. Furthermore, adaptive algorithms can be introduced to optimize the distance calculation process, for example, automatically adjusting calculation parameters based on dynamic changes in power demand. In some cases, machine learning methods can also be used to predict and estimate the distance outside the mask, especially when data is sparse or computational resources are limited.

[0098] The embodiments disclosed above have the following beneficial effects: The virtual power plant power load control method of the present invention can construct a power load prediction and control model by acquiring power load data and power supply capacity data, thereby achieving precise control and optimization of power load. This method can effectively improve the operating efficiency and stability of the power system, while reducing the risks caused by load fluctuations. By subdividing and dynamically adjusting power supply capacity, it can respond more flexibly to rapid changes in the power market and achieve efficient utilization of power resources.

[0099] Furthermore, the control method employs a composite loss function with load prediction error penalty, which improves prediction accuracy during model training. This approach reduces power load control errors caused by prediction mistakes, thereby enhancing power system reliability. Simultaneously, in-depth analysis of sub-supply capacity data allows for more precise adjustments to the power load, enabling finer control of the power system and further improving its economic efficiency and environmental friendliness.

[0100] Referring further to Figure 2, as an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a virtual power plant power load control system. These device embodiments correspond to the method embodiments shown in Figure 2. This real-time image recognition model training system can be specifically applied to various electronic devices. As shown in Figure 2, one embodiment of a virtual power plant power load control system 200 includes:

[0101] Feature extraction module 201 is used to extract features from power load data and power supply capacity data using a deep learning network;

[0102] Prediction module 202 is used to predict future power load based on extracted features using a time series prediction algorithm;

[0103] Optimization module 203 is used to generate power load control strategies based on prediction results and power supply capacity data using optimization algorithms;

[0104] Evaluation module 204 is used to evaluate the performance of power load control strategies, including economic efficiency, reliability, and environmental impact.

[0105] It is understood that the modules described in the virtual power plant load control system 200 correspond to the steps in the virtual power plant load control method described with reference to Figure 1. Therefore, the operations, features, and beneficial effects described above for the virtual power plant load control method also apply to the virtual power plant load control system 200 and its included modules, and will not be repeated here.

[0106] In some embodiments, a control strategy training sample set is generated based on the power load data and the corresponding power supply capacity data, including:

[0107] Gaussian process regression was used to interpolate the power load data and the corresponding power supply capacity data to handle missing data;

[0108] Wavelet transform was performed on power load data and power supply capacity data to extract time-frequency features;

[0109] For each sample data, principal component analysis is performed on the power load data and power supply capacity data respectively to reduce data dimensionality and improve model training efficiency.

[0110] Based on the dimensionality-reduced data, a clustering algorithm is used to classify the power load type and power supply capacity type to generate training samples.

[0111] Specifically, the virtual power plant load control system includes a feature extraction module, a prediction module, an optimization module, and an evaluation module. The feature extraction module uses deep learning networks to analyze and extract key features from load and power supply capacity data. The prediction module, based on the extracted features, uses time series forecasting algorithms to predict future load. The optimization module, based on the prediction results and power supply capacity data, uses optimization algorithms to generate a load control strategy. The evaluation module is responsible for evaluating the performance of the control strategy, including its economic efficiency, reliability, and environmental impact.

[0112] Preferably, to improve the accuracy and efficiency of the system, advanced deep learning techniques such as Convolutional Neural Networks (CNNs) or Recurrent Neural Networks (RNNs) can be employed in the feature extraction module to better capture the spatiotemporal characteristics of the data. In the prediction module, algorithms such as Long Short-Term Memory Networks (LSTMs) or Gated Recurrent Units (GRUs) can be used to improve the accuracy of time series prediction. The optimization module can employ genetic algorithms, Particle Swarm Optimization (PSO), or simulated annealing to find the optimal power load strategy. The evaluation module can incorporate multi-objective optimization and Life Cycle Assessment (LCA) methods to comprehensively evaluate the overall performance of the control strategy. Furthermore, the system can integrate real-time data monitoring and adaptive learning mechanisms to dynamically adjust the control strategy and adapt to the continuous changes in the power system.

[0113] Referring now to FIG3, a schematic diagram of the structure of an electronic device 300 suitable for implementing some embodiments of the present disclosure is shown. The electronic devices in some embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptops, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. The terminal device shown in FIG3 is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present disclosure.

[0114] As shown in Figure 3, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0115] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 shows electronic device 300 with various devices, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Each box shown in Figure 3 may represent one device, or multiple devices may be represented as needed.

[0116] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0117] The storage medium in this application embodiment stores program instructions capable of implementing all the above methods. These program instructions can be stored in the storage medium in the form of a software product, including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks, or terminal devices such as computers, servers, mobile phones, and tablets.

[0118] It should be noted that, in some embodiments of this disclosure, the computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, the computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination thereof.

[0119] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol), and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0120] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol), and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0121] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs that, when executed by the electronic device, enable the electronic device to write computer program code in one or more programming languages ​​or combinations thereof for performing operations of some embodiments of this disclosure. The programming languages ​​include object-oriented programming languages—such as Java, Smalltalk, and C++—and conventional procedural programming languages—such as the "C" language or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0122] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0123] The units described in some embodiments of this disclosure can be implemented in software or in hardware. The described units can also be located in a processor, and the functions described above can be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: Field Programmable Gate Arrays (FPGAs), Application-Specific Integrated Circuits (ASICs), Application Standard Products (ASSPs), System-on-Chip (SoCs), Complex Programmable Logic Devices (CPLDs), and so on.

[0124] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for controlling the power load of a virtual power plant, characterized in that, Includes the following steps: Obtain the power load data and corresponding power supply capacity data of the virtual power plant, and generate a control strategy training sample set based on the power load data and corresponding power supply capacity data; Based on the stability of the power system, the power supply capacity data of each corresponding power supply capacity data in the control strategy training sample set is analyzed to obtain multiple sub-supply capacity data of each corresponding power supply capacity data. A power load prediction and control model is constructed. The power load prediction and control model is trained based on the power load data, power supply capacity data, and multiple sub-supply capacity data of each corresponding power supply capacity data in the training sample set to obtain the power load control model. The power load data of the virtual power plant to be controlled is input into the power load control model to obtain the power load control strategy of the virtual power plant to be controlled. Based on multiple sub-supply capacity data of power supply capacity, the training loss of the power load forecasting and control model is calculated using a load forecasting error penalty matrix with load forecasting error penalty. The load forecasting error penalty matrix corresponding to the power supply matrix is ​​calculated using the following steps: Calculate the Euclidean distance matrix (EDM) within the m-th sub-mask matrix based on the electricity demand matrix. m Based on the Euclidean distance matrix EDM within the mask m Calculate the in-mask penalty term of the m-th sub-mask matrix. int,m ; The external Euclidean distance matrix (EDM) of the m-th sub-mask matrix is ​​calculated based on the power supply matrix. ext,m Based on the Euclidean distance matrix EDM outside the mask ext,m Calculate the out-of-mask penalty term of the m-th sub-mask matrix. ext,m ; Penalty within the mask int,m and the penalty item outside the mask ext,m Normalizing and summing the results yields the distance penalty matrix (penal) corresponding to the m-th sub-mask matrix. m ; The distance penalty matrix corresponding to each sub-mask matrix is ​​penal m The load forecasting error penalty matrix Penal is obtained by splicing the power supply matrix.

2. The virtual power plant power load control method according to claim 1, characterized in that, Based on the stability of the power system, the data for each power supply capacity in the training sample set are analyzed to obtain multiple sub-supply capacity data for each power supply capacity, including: Take the initial state of each power supply capacity as the first type and calculate the stability of the initial state; iterate through each state of the power supply capacity in turn and calculate the stability of the current state. If the stability of the current state is the same as the stability of the previous state, then the current state and the previous state are of the same type; otherwise, the current state and the previous state are of different types. Multiple sub-supply capacity data of the power supply capacity are obtained based on the type of each state in the power supply capacity.

3. The virtual power plant power load control method according to claim 1, characterized in that, The Euclidean distance matrix EDM within the mask of the m-th sub-mask matrix m It has the same dimension as the m-th sub-mask matrix of the electricity demand matrix.

4. The virtual power plant power load control method according to claim 1, characterized in that, Calculate the in-mask penalty term of the m-th submask matrix.

5. The virtual power plant power load control method according to claim 1, characterized in that, Calculate the out-of-mask penalty term for the m-th sub-mask matrix.

6. The virtual power plant power load control method according to claim 5, characterized in that, The Euclidean distance matrix (EDM) outside the mask of the m-th sub-mask matrix ext,m It has the same dimension as the m-th sub-mask matrix of the power supply matrix.

7. A virtual power plant power load control system, wherein the system implements the method as described in claim 1, characterized in that, The system includes: The feature extraction module is used to extract features from power load data and power supply capacity data using a deep learning network; The prediction module is used to predict future power load based on extracted features using a time series prediction algorithm; The optimization module is used to generate power load control strategies based on prediction results and power supply capacity data using optimization algorithms; The evaluation module is used to assess the performance of power load control strategies, including economic efficiency, reliability, and environmental impact.

8. The virtual power plant power load control system according to claim 7, characterized in that, A control strategy training sample set is generated based on the power load data and the corresponding power supply capacity data, including: Gaussian process regression was used to interpolate the power load data and the corresponding power supply capacity data to handle missing data; Wavelet transform was performed on power load data and power supply capacity data to extract time-frequency features; For each sample data, principal component analysis is performed on the power load data and power supply capacity data respectively to reduce data dimensionality and improve model training efficiency. Based on the dimensionality-reduced data, a clustering algorithm is used to classify the power load type and power supply capacity type to generate training samples.

Images