Cooperative control method and device for new energy microgrid

Through deep learning technology, the historical and future data of new energy power generation systems are extracted and integrated to generate a future power generation curve chart, solving the accuracy of the prediction of volatility of new energy power generation and improving the stability and power quality of power grid management.

CN118920576AInactive Publication Date: 2024-11-08STATE GRID HENAN ELECTRIC POWER COMPANY ANYANG POWER SUPPLY +4
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410971580.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-08
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The uncertainty and randomness of new energy power generation lead to fluctuations in grid frequency and voltage, increasing the complexity of grid management. The existing methods based on probability statistics may lead to deviations in prediction results when the assumptions are inconsistent.

Method used

Deep learning technology is used to convolutionally encode the historical and future generation factors data of distributed new energy power generation systems, extract the feature matrix and fuse it to generate a future generation curve chart to achieve accurate prediction.

Benefits of technology

Through accurate prediction of deep learning technology, the prediction accuracy of the volatility of distributed new energy generation is improved, helping grid management to cope with uncertainty and random challenges, and promoting the safety and stability of the power grid and the power quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118920576A_ABST
    Figure CN118920576A_ABST
Patent Text Reader

Abstract

The invention relates to the field of intelligent control of new energy micro-grids, and particularly discloses a cooperative control method and device of a new energy micro-grid, and the method comprises the steps: enabling collected historical power generation data of a distributed new energy power generation system and power generation influence factor data in a period of time in the future to serve as input data; and convolutional coding is carried out on the input data through a deep learning technology. That is, the input data is processed through a deep learning technique to obtain a more accurate feature representation. And then analyzing the feature representations to obtain a generation result, wherein the generation result is used for representing a power generation curve graph of the distributed new energy power generation system in a period of time in the future. In this way, the fluctuation of distributed new energy power generation can be accurately predicted.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of intelligent control of new energy microgrids, and more specifically, to a collaborative control method and device for new energy microgrids. Background Art

[0002] Distributed renewable energy generation, as a renewable and pollution-free energy form, has the potential to improve the insufficient capacity of distribution networks and increase the voltage of end users. However, the uncertainty and randomness of renewable energy generation have brought challenges to the safety, stability and power quality of the power grid. The volatility of this energy may cause fluctuations in the frequency and voltage of the power grid, increasing the complexity of power grid management.

[0003] In order to solve the uncertainty and randomness challenges brought by distributed renewable energy generation, methods based on probability statistics are usually used, such as probability prediction models, random optimization algorithms, etc. This method is usually based on certain assumptions and models. If these assumptions do not match the actual situation, it may lead to deviations in the prediction results.

[0004] Therefore, we look forward to a collaborative control method and device for a new energy microgrid, which can monitor and analyze the distributed new energy power generation system in real time based on deep learning technology, so as to achieve accurate prediction of the volatility of distributed new energy power generation. Summary of the invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a collaborative control method and device for a new energy microgrid, which uses the collected historical power generation data of a distributed new energy power generation system and the data of power generation influencing factors in the future as input data, and performs convolution encoding on these input data respectively through deep learning technology. That is, the input data is processed by deep learning technology to obtain a more accurate feature representation. These feature representations are then analyzed to obtain a generated result, and the generated result is used to represent the power generation curve graph of the distributed new energy power generation system in the future. In this way, it is helpful to achieve accurate prediction of the volatility of distributed new energy power generation.

[0006] According to a first aspect of the present application, a collaborative control method for a new energy microgrid is provided, which includes:

[0007] Collect historical power generation data of distributed renewable energy power generation systems and data on factors affecting power generation in the future;

[0008] Extracting and analyzing the historical power generation data of the distributed renewable energy power generation system and the power generation influencing factor data in the future period of time to obtain a historical power generation time series change feature matrix and a power generation influencing factor time series change feature matrix;

[0009] The historical power generation time series change characteristic matrix and the power generation influencing factor time series change characteristic matrix are subjected to feature fusion to obtain a distributed new energy power generation characteristic matrix;

[0010] A generation result is obtained based on the distributed renewable energy power generation characteristic matrix, and the generation result is used to represent a power generation curve diagram of the distributed renewable energy power generation system in a future period of time.

[0011] In combination with the first aspect of the present application, in a collaborative control method of a new energy microgrid of the first aspect of the present application, feature extraction and analysis are performed on the historical power generation data of the distributed new energy power generation system and the power generation influencing factor data in the future period of time to obtain a historical power generation time series change feature matrix and a power generation influencing factor time series change feature matrix, including: feature encoding the historical power generation data of the distributed new energy power generation system to obtain the historical power generation time series change feature matrix; feature encoding the power generation influencing factor data in the future period of time to obtain the power generation influencing factor time series change feature matrix.

[0012] According to a second aspect of the present application, a collaborative control device for a new energy microgrid is provided, comprising:

[0013] Distributed new energy data collection module, used to collect historical power generation data of distributed new energy power generation systems and data on factors affecting power generation in the future;

[0014] A distributed new energy feature extraction module is used to extract and analyze the historical power generation data of the distributed new energy power generation system and the power generation influencing factor data in the future period of time to obtain a historical power generation time series change feature matrix and a power generation influencing factor time series change feature matrix;

[0015] A distributed new energy feature fusion module is used to fuse the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix to obtain a distributed new energy power generation feature matrix;

[0016] A distributed new energy classification result generation module is used to obtain a generation result based on the distributed new energy power generation feature matrix, and the generation result is used to represent a power generation curve diagram of a distributed new energy power generation system in the future.

[0017] In combination with the second aspect of the present application, in a collaborative control device of a new energy microgrid in the second aspect of the present application, the distributed new energy feature extraction module includes: feature encoding the historical power generation data of the distributed new energy power generation system to obtain the historical power generation time series change feature matrix; feature encoding the power generation influencing factor data in the future period of time to obtain the power generation influencing factor time series change feature matrix.

[0018] Compared with the prior art, the present application provides a method and device for collaborative control of a new energy microgrid, which uses the collected historical power generation data of a distributed new energy power generation system and the data of power generation influencing factors in the future as input data, and performs convolution encoding on these input data respectively through deep learning technology. That is, the input data is processed by deep learning technology to obtain a more accurate feature representation. These feature representations are then analyzed to obtain a generated result, and the generated result is used to represent the power generation curve of the distributed new energy power generation system in the future. In this way, it is helpful to achieve accurate prediction of the volatility of distributed new energy power generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0020] Figure 1 The figure illustrates a flow chart of a collaborative control method of a new energy microgrid according to an embodiment of the present application.

[0021] Figure 2 The figure illustrates a flow chart of step S2 in the collaborative control method of a new energy microgrid according to an embodiment of the present application.

[0022] Figure 3 The figure illustrates a flow chart of step S21 in step S2 of the collaborative control method of the new energy microgrid according to an embodiment of the present application.

[0023] Figure 4 The figure illustrates a flow chart of step S22 in step S2 of the collaborative control method of the new energy microgrid according to an embodiment of the present application.

[0024] Figure 5 The figure illustrates a flow chart of step S4 in the collaborative control method of a new energy microgrid according to an embodiment of the present application.

[0025] Figure 6The figure illustrates a schematic block diagram of a collaborative control device for a new energy microgrid according to an embodiment of the present application. DETAILED DESCRIPTION

[0026] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0027] Figure 1 The figure shows a flow chart of a method for collaborative control of a new energy microgrid according to an embodiment of the present application. Figure 1 As shown, according to the collaborative control method of the new energy microgrid of the embodiment of the present application, the method includes: S1, collecting historical power generation data of the distributed new energy power generation system and data of power generation influencing factors in the future period of time; S2, respectively extracting and analyzing the historical power generation data of the distributed new energy power generation system and the data of power generation influencing factors in the future period of time to obtain a historical power generation time series change feature matrix and a power generation influencing factor time series change feature matrix; S3, feature fusion of the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix to obtain a distributed new energy power generation feature matrix; S4, obtaining a generation result based on the distributed new energy power generation feature matrix, and the generation result is used to represent the power generation curve diagram of the distributed new energy power generation system in the future period of time.

[0028] Distributed renewable energy generation, as an environmentally friendly and renewable energy form, helps alleviate the problem of insufficient distribution network capacity and improve the voltage level of end users. However, due to the uncertainty and randomness of renewable energy generation, it poses challenges to the stability and power quality of the power grid. The volatility of renewable energy generation may cause fluctuations in grid frequency and voltage, increasing the complexity of grid management.

[0029] In view of the uncertainty and randomness challenges brought by distributed renewable energy generation, methods based on probability statistics are usually used to solve them, such as probabilistic prediction models, random optimization algorithms, etc. This method is usually based on certain assumptions and models. If these assumptions do not match the actual situation, it may lead to deviations in the prediction results. Therefore, a collaborative control method and device for a new energy microgrid is expected to monitor and analyze the distributed renewable energy generation system in real time based on deep learning technology, so as to achieve accurate prediction of the volatility of distributed renewable energy generation.

[0030] In step S1, the historical power generation data of the distributed renewable energy power generation system and the power generation influencing factor data in the future are collected. It should be understood that the historical power generation data can be used to train the deep learning model to help the model learn the power generation mode, trend and law of the distributed renewable energy power generation system. That is, the historical data contains the characteristics and patterns of various power generation situations, which can be used as the basis for feature extraction, which helps the model to better understand the system behavior. Therefore, the historical power generation data (including power generation, weather information, load, etc.) of the distributed renewable energy power generation system are first collected as input data. In addition, considering that the power generation influencing factor data (such as weather information, load, etc.) in the future can provide information about the future environment, it is helpful for the model to more accurately predict the power generation of the distributed renewable energy power generation system. Therefore, it is also necessary to collect the power generation influencing factor data (weather information, load, etc.) in the future. In this way, by combining historical data and power generation influencing factor data, it is helpful to improve the prediction accuracy of the volatility of the distributed renewable energy power generation system, thereby effectively responding to the challenges of uncertainty and randomness, and promoting the safety, stability and power quality of the power grid.

[0031] In step S2, feature extraction and analysis are performed on the historical power generation data of the distributed renewable energy power generation system and the power generation influencing factor data in the future period of time to obtain the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix. It should be understood that after obtaining these data, the next step is to perform feature extraction and analysis on these input data.

[0032] Specifically, Figure 2 FIG. 1 is a flowchart of step S2 in the collaborative control method of a new energy microgrid according to an embodiment of the present application. Figure 2 As shown, step S2 includes: S21, feature encoding the historical power generation data of the distributed new energy power generation system to obtain the historical power generation time series change feature matrix; S22, feature encoding the power generation influencing factor data in the future period of time to obtain the power generation influencing factor time series change feature matrix.

[0033] First, feature extraction and analysis are performed on the historical power generation data of the distributed renewable energy power generation system. Specifically, Figure 3 The figure shows a flow chart of step S21 in step S2 of the collaborative control method of the new energy microgrid according to an embodiment of the present application. Figure 3As shown, step S21 includes: step S211, arranging the historical power generation data of the distributed new energy power generation system into multiple historical power generation information sequences according to the time dimension, and then passing them through a context encoder including an embedding layer to obtain multiple historical power generation feature vectors; step S212, arranging the multiple historical power generation feature vectors into a historical power generation time series two-dimensional matrix, and then passing the historical power generation time series change feature extraction module based on a two-dimensional convolution kernel to obtain the historical power generation time series change feature matrix.

[0034] It should be understood that the power generation data of the distributed renewable energy power generation system changes over time, that is, the historical power generation data (including power generation, weather information, load, etc.) of the distributed renewable energy power generation system has certain time sequence characteristics. Therefore, in order to retain the time relationship and order between the data, the historical power generation data of the distributed renewable energy power generation system is first arranged according to the time dimension to obtain the historical power generation information sequence corresponding to each time point. In this way, it can help the model better understand the time dependency between the data, thereby improving the ability to understand and analyze the data. Further, it should be understood that the historical power generation data includes power generation, weather information, load, etc., which contain not only numerical data but also text data. Based on this, considering that the embedding layer can convert discrete text data into dense continuous vector representation, it is helpful to capture the semantic relationship between words; and the context encoder can effectively process sequence data and retain long-term dependencies in the sequence. Therefore, a context encoder containing an embedding layer is used to extract and analyze the text features of the multiple historical power generation information sequences. In this way, it is helpful to better understand and analyze the historical power generation data of the distributed renewable energy power generation system.

[0035] Specifically, step S211 is used to: perform word segmentation processing on each of the multiple historical power generation information sequences to obtain a word sequence; use the embedding layer of the context encoder to map each word in the word sequence into a word embedding vector to obtain a sequence of word embedding vectors; use the converter-based Bert model of the context encoder to perform global context semantic encoding on the sequence of word embedding vectors to obtain multiple feature vectors; and cascade the multiple feature vectors to obtain the multiple historical power generation feature vectors.

[0036] Next, considering the continuous changes in the time dimension of historical power generation data, based on this, in order to better capture the changing laws and trends of historical power generation data in the time series, the historical power generation feature vectors at each time point are arranged into a two-dimensional matrix of the historical power generation time series. Then, a historical power generation time series change feature extraction module based on a two-dimensional convolution kernel is used to extract and analyze the features of the historical power generation time series two-dimensional matrix to obtain a historical power generation time series change feature matrix. Among them, the two-dimensional convolution kernel performs a sliding operation on the two-dimensional matrix, which can effectively extract the spatial features in the historical power generation data. This helps the model capture the correlation and patterns between different time points.

[0037] Specifically, step S212 is used to: use each layer of the historical power generation timing change feature extraction module to perform the following on the input data in the forward transfer of the layer: perform convolution processing based on a two-dimensional convolution kernel on the input data to obtain a convolution feature map; perform pooling processing on the convolution feature map along the channel dimension to obtain a pooling feature map; perform nonlinear activation on the pooling feature map to obtain an activation feature map; wherein the input of the first layer of the historical power generation timing change feature extraction module is the historical power generation timing two-dimensional matrix, and the output of the last layer of the historical power generation timing change feature extraction module is the historical power generation timing change feature matrix.

[0038] Then, the data of the factors affecting power generation in the future period are subjected to feature extraction and analysis. Specifically, Figure 4 The figure shows a flow chart of step S22 in step S2 of the collaborative control method of the new energy microgrid according to an embodiment of the present application. Figure 4 As shown, step S22 includes: step S221, arranging the power generation influencing factor data within the future period of time into multiple power generation influencing factor sequences according to the time dimension, and then passing through the power generation influencing factor feature extraction module to obtain multiple power generation influencing factor feature vectors; step S222, arranging the multiple power generation influencing factor feature vectors into a power generation influencing factor time series two-dimensional matrix, and then passing through the power generation influencing factor time series change feature extraction module based on the feature encoder to obtain the power generation influencing factor time series change feature matrix. Wherein, the power generation influencing factor feature extraction module performs feature encoding operation on the power generation influencing factor data within the future period of time through the context encoder.

[0039] It should be understood that the data of power generation influencing factors in the future period of time has the characteristics of dynamic time series changes. Therefore, in order to better preserve the time relationship between the data of power generation influencing factors in the future period of time, the data of power generation influencing factors in the future period of time are first arranged according to the time dimension to obtain the power generation influencing factor sequence corresponding to each time point. Then, the power generation influencing factor feature extraction module is used to perform semantic understanding and analysis on these power generation influencing factor sequences respectively to obtain multiple power generation influencing factor feature vectors. Among them, the context encoder can convert the power generation influencing factor data into a high-dimensional feature vector, capture the important features in the data, and help improve the prediction performance of the model.

[0040] Specifically, step S221 is used to: perform word segmentation processing on each of the multiple power generation influencing factor sequences to obtain a word sequence; use the embedding layer of the power generation influencing factor feature extraction module to map each word in the word sequence into a word embedding vector to obtain a sequence of word embedding vectors; use the converter-based Bert model of the power generation influencing factor feature extraction module to perform global context semantic encoding on the sequence of word embedding vectors to obtain multiple feature vectors; and cascade the multiple feature vectors to obtain the multiple power generation influencing factor feature vectors.

[0041] Next, considering that through two-dimensional arrangement, the model can better understand and capture the changing laws and trends of different power generation influencing factors in the time series, so as to achieve accurate prediction of the volatility of distributed renewable energy power generation. Therefore, the multiple power generation influencing factor feature vectors are first arranged into a two-dimensional matrix of power generation influencing factors in time series. Further, considering that the feature encoder can process time series data, it helps the model capture the temporal changing characteristics of power generation influencing factors. Therefore, a power generation influencing factor time series change feature extraction module based on a feature encoder is used to perform deep convolution encoding on the power generation influencing factor time series two-dimensional matrix to obtain a power generation influencing factor time series change feature matrix. In this way, it can help the model better understand the time evolution process of power generation influencing factor data, capture the changing characteristics of data in time series, and thus achieve accurate prediction of the volatility of distributed renewable energy power generation.

[0042] Specifically, step S222 is used to: use each layer of the power generation influencing factor time series change feature extraction module to respectively perform the following operations on the input data in the forward transfer of the layer: use the convolution units of each layer of the power generation influencing factor time series change feature extraction module to perform convolution processing based on a two-dimensional convolution kernel on the input data to obtain a convolution feature map; use the pooling units of each layer of the power generation influencing factor time series change feature extraction module to perform pooling processing on the convolution feature map along the channel dimension to obtain a pooling feature map; and use the activation units of each layer of the power generation influencing factor time series change feature extraction module to perform nonlinear activation on the feature values ​​of each position in the pooling feature map to obtain an activation feature map; wherein the output of the last layer of the power generation influencing factor time series change feature extraction module is the power generation influencing factor time series change feature matrix.

[0043] In step S3, the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix are feature fused to obtain a distributed new energy power generation feature matrix. It should be understood that the historical power generation time series change feature matrix contains the feature representation of power generation data in the past time period, which can help the model understand the historical operation, power generation trend and periodic changes of the power generation system. These historical data can provide important information about the past performance of the power generation system. The power generation influencing factor time series change feature matrix contains the feature representation of various influencing factors (such as weather, load, etc.) in the future time period, which can help the model understand the degree of influence of various influencing factors on the power generation system in the future period. By fusing these two feature matrices, the model can comprehensively consider the impact of historical power generation and current power generation influencing factors, so as to more comprehensively analyze the operating state of the power generation system and predict future power generation volatility. Therefore, in order to comprehensively utilize historical data and influencing factor information to achieve more accurate prediction of distributed new energy power generation volatility, the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix are feature fused to obtain a distributed new energy power generation feature matrix.

[0044] Specifically, step S3 is used to: fuse the historical power generation time series change characteristic matrix and the power generation influencing factor time series change characteristic matrix with the following cascade formula to obtain the distributed new energy power generation characteristic matrix; wherein the cascade formula is: a =Concat[M a1 ,M a2 ], where M a1 represents the characteristic matrix of the historical power generation time series change, M a2 represents the time series variation characteristic matrix of the power generation influencing factors, M a represents the distributed renewable energy power generation characteristic matrix, and Concat[·,·] represents the cascade function.

[0045] In step S4, a generation result is obtained based on the distributed renewable energy power generation characteristic matrix, and the generation result is used to represent a power generation curve diagram of the distributed renewable energy power generation system in the future. Specifically, Figure 5 The figure shows a flow chart of step S4 in the collaborative control method of the new energy microgrid according to an embodiment of the present application. Figure 5 As shown, step S4 includes: step S41, passing the distributed renewable energy power generation feature matrix through a distributed renewable energy power generation feature extraction module based on a convolutional neural network model to obtain a distributed renewable energy power generation feature vector; step S42, performing local distribution consistency adaptation on the distributed renewable energy power generation feature vector based on subclass metric correlation prediction to obtain an optimized distributed renewable energy power generation feature vector; step S43, passing the optimized distributed renewable energy power generation feature vector through a generator to obtain the generation result.

[0046] It should be understood that only feature fusion may lead to information loss or oversimplification, and it is impossible to fully explore the complex relationships and patterns in the historical power generation data and power generation influencing factor data. Therefore, in order to capture deeper feature information in the historical power generation data and power generation influencing factor data, the distributed new energy power generation feature matrix is ​​further passed through a distributed new energy power generation feature extraction module based on a convolutional neural network model to obtain a distributed new energy power generation feature vector. Among them, the convolutional neural network model has a strong nonlinear modeling ability, which can better fit the complex power generation system data distribution and improve the accuracy of the prediction. In this way, combining the advantages of feature fusion and convolutional neural network models, the volatility of distributed new energy power generation systems can be analyzed and predicted more comprehensively and accurately.

[0047] Specifically, step S41 is used to: use each layer of the convolutional neural network model to perform the following operations on the input data in the forward pass of the layer: use the convolution units of each layer of the convolutional neural network model to perform convolution processing based on a one-dimensional convolution kernel on the input data to obtain a convolution feature map; use the pooling units of each layer of the convolutional neural network model to perform pooling processing based on a feature matrix on the convolution feature map to obtain a pooling feature map; and use the activation units of each layer of the convolutional neural network model to perform nonlinear activation on the eigenvalues ​​of each position in the pooling feature map to obtain an activation feature map; wherein the output of the last layer of the convolutional neural network model is the distributed renewable energy power generation feature vector.

[0048] In particular, in the technical solution of the present application, it is considered that the historical power generation data and the power generation influencing factor data may have inconsistent data quality due to factors such as sensor accuracy, installation location, and environmental interference, which affects the accuracy of feature extraction. In addition, the power generation of the new energy power generation system is greatly affected by natural conditions. The instability of these conditions may cause the volatility of historical power generation data in different time periods, thereby affecting the local distribution of the feature vector. However, the historical power generation feature extraction module based on the sequence encoder and the power generation influencing factor feature extraction module based on the context encoder may not be able to fully capture all the subtle features related to power generation, especially when the data changes are complex or there is noise. The historical power generation time series change feature extraction module based on the two-dimensional convolution kernel and the power generation influencing factor time series change feature extraction module based on the feature encoder may have limitations in processing long sequence data, resulting in inconsistent local distribution of the time series change feature matrix. At the same time, when fusing the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix, the weights and importance of different features may not be properly processed, which may lead to differences in the local distribution of the distributed new energy power generation feature matrix. In order to solve this problem, the distributed new energy power generation generated feature vector is further adapted for local distribution consistency based on subclass metric correlation prediction.

[0049] Specifically, step S42 includes: determining the weight learning parameter matrix of the generator; extracting each column vector in the weight learning parameter matrix to obtain a set of weight learning parameter column vectors; respectively calculating the metric correlation prediction factor between each weight learning parameter column vector in the set of weight learning parameter column vectors and the distributed renewable energy power generation generation feature vector to obtain a metric correlation prediction factor modulation vector; using the metric correlation prediction factor modulation vector as the weight vector, calculating the point-by-point multiplication between the metric correlation prediction factor modulation vector and the distributed renewable energy power generation generation feature vector to obtain the optimized distributed renewable energy power generation generation feature vector.

[0050] More specifically, respectively calculating the metric correlation prediction factor between each weight learning parameter column vector in the set of the weight learning parameter column vector and the distributed renewable energy power generation feature vector to obtain the metric correlation prediction factor modulation vector, including: respectively calculating the metric correlation prediction factor between each weight learning parameter column vector in the set of the weight learning parameter column vector and the distributed renewable energy power generation feature vector using the following prediction factor calculation formula to obtain the metric correlation prediction factor modulation vector, wherein the prediction factor calculation formula is:

[0051]

[0052] Among them, V irepresents each weight learning parameter column vector in the set of weight learning parameter column vectors, V c represents the distributed renewable energy generation feature vector, It means subtracting by position. is the superposition norm of the two norms of the eigenvectors, ||·|| F is the Frobenius norm of the eigenvector, M p represents the weight learning parameter matrix, exp(·) represents the exponential operation with the natural constant e as the base, V p represents the metric correlation predictor modulation vector.

[0053] Here, considering that there are local distribution differences in the distributed renewable energy power generation feature vector, when it is generated by the generator, there will be a class harmony offset relative to the predetermined class label of the generator, which affects the generation iteration effect. To this end, in the technical solution of the present application, the distributed renewable energy power generation feature vector is adapted for local distribution consistency based on subclass metric correlation prediction, and the superposition paradigm of the two norms between each column vector in the weight learning parameter matrix of the generator and the distributed renewable energy power generation feature vector is used as the subclass label related metric integral factor and the matrix norm of the weight learning parameter of the generator is used as the differential factor, and then the distributed renewable energy power generation feature vector is bidirectionally linearly modulated by the association paradigm between the subclass label related metric integral factor and the differential factor, so as to improve the local feature distribution consistency of the distributed renewable energy power generation feature vector by adaptively stretching and compressing the distributed renewable energy power generation feature vector, thereby reducing the harmony offset of the distributed renewable energy power generation feature vector relative to the predetermined class label of the generator, thereby improving the accuracy of the generation result.

[0054] Next, considering that the generator can help the model generate a power generation curve for a period of time in the future, and provide a prediction of future power generation volatility. This prediction helps the microgrid system to make advance adjustments and plans to cope with future power generation volatility. Therefore, the generator is further used to perform feature analysis on the optimized distributed renewable energy power generation feature vector to obtain a generation result, and the generation result is used to represent the power generation curve of the distributed renewable energy power generation system for a period of time in the future. In this way, it helps to provide more information and decision support for the coordinated control of the microgrid system.

[0055] In summary, according to the collaborative control method of the new energy microgrid according to the embodiment of the present application, the historical power generation data of the distributed new energy power generation system and the power generation influencing factor data in the future period of time collected are used as input data, and these input data are convolutionally encoded respectively through deep learning technology. That is, the input data is processed by deep learning technology to obtain a more accurate feature representation. These feature representations are then analyzed to obtain a generated result, and the generated result is used to represent the power generation curve of the distributed new energy power generation system in the future period of time. In this way, it is helpful to achieve accurate prediction of the volatility of distributed new energy power generation.

[0056] Figure 6 FIG. 1 is a schematic block diagram of a collaborative control device for a new energy microgrid according to an embodiment of the present application. Figure 6 As shown, according to the collaborative control device 100 of the new energy microgrid of the embodiment of the present application, it includes: a distributed new energy data acquisition module 110, which is used to collect the historical power generation data of the distributed new energy power generation system and the power generation influencing factor data in the future period of time; a distributed new energy feature extraction module 120, which is used to extract and analyze the historical power generation data of the distributed new energy power generation system and the power generation influencing factor data in the future period of time respectively to obtain the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix; a distributed new energy feature fusion module 130, which is used to feature fuse the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix to obtain a distributed new energy power generation feature matrix; a distributed new energy classification result generation module 140, which is used to obtain a generation result based on the distributed new energy power generation feature matrix, and the generation result is used to represent the power generation curve diagram of the distributed new energy power generation system in the future period of time.

[0057] As described above, the collaborative control device 100 for a new energy microgrid according to an embodiment of the present application can be implemented in various wireless terminals, such as a server for collaborative control of a new energy microgrid. In one example, the collaborative control device 100 for a new energy microgrid according to an embodiment of the present application can be integrated into a wireless terminal as a software module and / or a hardware module. For example, the collaborative control device 100 for a new energy microgrid can be a software module in the operating system of the wireless terminal, or can be an application developed for the wireless terminal; of course, the collaborative control device 100 for a new energy microgrid can also be one of the many hardware modules of the wireless terminal.

[0058] Alternatively, in another example, the collaborative control device 100 of the new energy microgrid and the wireless terminal may also be separate devices, and the collaborative control device 100 of the new energy microgrid may be connected to the wireless terminal via a wired and / or wireless network, and transmit interactive information in accordance with an agreed data format.

[0059] Here, those skilled in the art can understand that the specific functions and operations of each step in the above-mentioned new energy microgrid collaborative control device have been referred to above. Figure 1 The description of the collaborative control method of the new energy microgrid has been introduced in detail, and therefore, its repeated description will be omitted.

[0060] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0061] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, or it can be an electrical, mechanical or other form of connection.

[0062] The above is only a preferred embodiment of the present application and does not limit the present application in any form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present application without departing from the content of the technical solution of the present application still falls within the protection scope of the technical solution of the present application.

Claims

1. A collaborative control method for a new energy microgrid, characterized in that: include: Collect historical power generation data of distributed renewable energy power generation systems and data on factors affecting power generation in the future; Extracting and analyzing the historical power generation data of the distributed renewable energy power generation system and the power generation influencing factor data in the future period of time to obtain a historical power generation time series change feature matrix and a power generation influencing factor time series change feature matrix; The historical power generation time series change characteristic matrix and the power generation influencing factor time series change characteristic matrix are subjected to feature fusion to obtain a distributed new energy power generation characteristic matrix; A generation result is obtained based on the distributed renewable energy power generation characteristic matrix, and the generation result is used to represent a power generation curve diagram of the distributed renewable energy power generation system in a future period of time.

2. The collaborative control method of the new energy microgrid according to claim 1 is characterized in that: The historical power generation data of the distributed renewable energy power generation system and the power generation influencing factor data in the future period are respectively subjected to feature extraction and analysis to obtain a historical power generation time series change feature matrix and a power generation influencing factor time series change feature matrix, including: Characteristic encoding is performed on the historical power generation data of the distributed renewable energy power generation system to obtain the characteristic matrix of the historical power generation time series change; The power generation influencing factor data within the future period of time is feature-coded to obtain a time-series variation feature matrix of the power generation influencing factor.

3. The collaborative control method of the new energy microgrid according to claim 2 is characterized in that: The historical power generation data of the distributed renewable energy power generation system is feature encoded to obtain the historical power generation time series change feature matrix, including: Arranging the historical power generation data of the distributed renewable energy power generation system into a plurality of historical power generation information sequences according to the time dimension, and then passing them through a context encoder including an embedding layer to obtain a plurality of historical power generation feature vectors; After arranging the multiple historical power generation feature vectors into a historical power generation time series two-dimensional matrix, a historical power generation time series change feature extraction module based on a two-dimensional convolution kernel is used to obtain the historical power generation time series change feature matrix.

4. The collaborative control method of the new energy microgrid according to claim 3 is characterized in that: After arranging the multiple historical power generation feature vectors into a historical power generation time series two-dimensional matrix, the historical power generation time series change feature extraction module based on a two-dimensional convolution kernel is used to obtain the historical power generation time series change feature matrix, which is used to: use each layer of the historical power generation time series change feature extraction module to perform the following operations on the input data in the forward transfer of the layer: Performing convolution processing based on a two-dimensional convolution kernel on the input data to obtain a convolution feature map; Performing pooling processing on the convolutional feature map along the channel dimension to obtain a pooled feature map; Performing nonlinear activation on the pooled feature map to obtain an activated feature map; Among them, the input of the first layer of the historical power generation time series change feature extraction module is the historical power generation time series two-dimensional matrix, and the output of the last layer of the historical power generation time series change feature extraction module is the historical power generation time series change feature matrix.

5. The collaborative control method of the new energy microgrid according to claim 4 is characterized in that: The data of the power generation influencing factors within the future period of time are feature-encoded to obtain a time-series variation feature matrix of the power generation influencing factors, including: Arranging the power generation influencing factor data in the future period of time into multiple power generation influencing factor sequences according to the time dimension and then respectively passing them through the power generation influencing factor feature extraction module to obtain multiple power generation influencing factor feature vectors; After arranging the plurality of power generation influencing factor feature vectors into a power generation influencing factor time series two-dimensional matrix, the power generation influencing factor time series variation feature extraction module based on a feature encoder is used to obtain the power generation influencing factor time series variation feature matrix.

6. The collaborative control method of the new energy microgrid according to claim 5 is characterized in that: The power generation influencing factor data in the future period are arranged into multiple power generation influencing factor sequences according to the time dimension, and then respectively passed through the power generation influencing factor feature extraction module to obtain multiple power generation influencing factor feature vectors for: Performing word segmentation processing on each of the plurality of power generation influencing factor sequences to obtain a word sequence; Using the embedding layer of the power generation influencing factor feature extraction module, each word in the word sequence is mapped into a word embedding vector to obtain a sequence of word embedding vectors; Using the transformer-based Bert model of the power generation influencing factor feature extraction module to perform global context semantic encoding on the sequence of word embedding vectors to obtain multiple feature vectors; as well as The multiple characteristic vectors are cascaded to obtain the multiple power generation influencing factor characteristic vectors.

7. The collaborative control method of the new energy microgrid according to claim 6 is characterized in that: Based on the distributed renewable energy power generation characteristic matrix, a generation result is obtained, and the generation result is used to represent a power generation curve diagram of the distributed renewable energy power generation system in a future period of time, including: The distributed renewable energy power generation feature matrix is ​​passed through a distributed renewable energy power generation feature extraction module based on a convolutional neural network model to obtain a distributed renewable energy power generation feature vector; Performing local distribution consistency adaptation based on subclass metric correlation prediction on the distributed renewable energy power generation feature vector to obtain an optimized distributed renewable energy power generation feature vector; The optimized distributed renewable energy power generation generates a characteristic vector which is passed through a generator to obtain the generated result.

8. The collaborative control method of the new energy microgrid according to claim 7 is characterized in that: The distributed renewable energy power generation feature vector is adapted for local distribution consistency based on subclass metric correlation prediction to obtain an optimized distributed renewable energy power generation feature vector, including: Determining a weight learning parameter matrix of the generator; Extracting each column vector in the weight learning parameter matrix to obtain a set of weight learning parameter column vectors; Respectively calculating the metric correlation prediction factor between each weight learning parameter column vector in the set of weight learning parameter column vectors and the distributed renewable energy power generation feature vector to obtain a metric correlation prediction factor modulation vector; The metric correlation prediction factor modulation vector is used as a weight vector, and the location point multiplication between the metric correlation prediction factor modulation vector and the distributed renewable energy power generation feature vector is calculated to obtain the optimized distributed renewable energy power generation feature vector.

9. The collaborative control method of the new energy microgrid according to claim 8 is characterized in that: Respectively calculating the metric correlation prediction factor between each weight learning parameter column vector in the set of the weight learning parameter column vector and the distributed renewable energy power generation feature vector to obtain a metric correlation prediction factor modulation vector, including: The metric correlation prediction factor between each weight learning parameter column vector in the set of the weight learning parameter column vector and the distributed renewable energy power generation feature vector is calculated using the following prediction factor calculation formula to obtain the metric correlation prediction factor modulation vector, wherein the prediction factor calculation formula is: Among them, V i represents each weight learning parameter column vector in the set of weight learning parameter column vectors, V c represents the distributed renewable energy generation feature vector, It means subtracting by position. is the superposition norm of the two norms of the eigenvectors, ||·|| F is the Frobenius norm of the eigenvector, M p represents the weight learning parameter matrix, exp(·) represents the exponential operation with the natural constant e as the base, V p represents the metric correlation predictor modulation vector.

10. A collaborative control device for a new energy microgrid, characterized in that ,include: Distributed new energy data collection module, used to collect historical power generation data of distributed new energy power generation systems and data on factors affecting power generation in the future; A distributed new energy feature extraction module is used to extract and analyze the historical power generation data of the distributed new energy power generation system and the power generation influencing factor data in the future period of time to obtain a historical power generation time series change feature matrix and a power generation influencing factor time series change feature matrix; A distributed new energy feature fusion module is used to fuse the historical power generation time series change feature matrix and the power generation influencing factor time series change feature matrix to obtain a distributed new energy power generation feature matrix; A distributed new energy classification result generation module is used to obtain a generation result based on the distributed new energy power generation feature matrix, and the generation result is used to represent a power generation curve diagram of a distributed new energy power generation system in the future.