Real-time operation monitoring method and system for new energy power generation system

Abstract

The invention relates to the field of real-time monitoring, and particularly discloses a real-time operation monitoring method and system for a new energy power generation system. The method comprises the following steps: firstly, acquiring photovoltaic panel energy collection values at a plurality of preset time points acquired by a database, photovoltaic panel historical energy collection text data acquired from the database and photovoltaic panel inclination angle values at a plurality of preset time points acquired by an inclination angle measuring device; and finally, a classification result is obtained through a classifier to judge whether the inclination angle of the new energy photovoltaic panel needs to be adjusted, so that the photovoltaic system is kept in an optimal working state, and the stability and the reliability of the system are improved.

Description

Technical Field

[0001] This application relates to the field of real-time monitoring, and more specifically, to a method and system for real-time monitoring of the operation of a new energy power generation system. Background Art

[0002] New energy refers to clean and green energy mainly represented by renewable energy sources such as solar energy, wind energy, water energy, and geothermal energy. It has the characteristics of being inexhaustible and environmentally friendly, and does not produce greenhouse gas emissions. The development of new energy helps to reduce the dependence on traditional fossil fuels and promote the transformation of the energy structure.

[0003] In the design of solar energy collection, that is, the design of photovoltaic power stations, the optimal tilt angle is usually selected based on historical data of the annual cumulative radiation amount at different tilt angles. However, fixing at the angle with the highest annual cumulative radiation amount may not be the best choice because the sunlight irradiation angle changes every day, resulting in a reduction in the direct radiation amount received by the photovoltaic modules, thereby affecting the power generation efficiency. Therefore, dynamically adjusting the tilt angle of the photovoltaic panels to maximize the energy collection efficiency is an important consideration in optimizing the design of photovoltaic power stations.

[0004] Therefore, there is a need for a method and system for real-time monitoring of the operation of a new energy power generation system. Summary of the Invention

[0005] To solve the above technical problems, this application is proposed. Embodiments of this application provide a method and system for real-time monitoring of the operation of a new energy power generation system. First, it obtains the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panels at multiple predetermined time points collected by the tilt angle measuring device. Then, using deep learning technology, feature extraction and correlation analysis are performed on the three, and finally, a classification result is obtained through a classifier to determine whether the tilt angle of the new energy photovoltaic panel needs to be adjusted, so as to keep the photovoltaic system in the best working state and improve the stability and reliability of the system.

[0006] According to one aspect of this application, there is provided a method for real-time monitoring of the operation of a new energy power generation system, which includes:

[0007] Obtain the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panels at multiple predetermined time points collected by the tilt angle measuring device;

[0008] Extract a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected from the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device;

[0009] Based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector, determine whether the tilt angle of the new energy photovoltaic panel needs to be adjusted.

[0010] According to another aspect of the present application, there is provided a real-time operation monitoring system for a new energy power generation system, which includes:

[0011] A photovoltaic energy data acquisition module, configured to acquire the photovoltaic panel energy collection values at multiple predetermined time points collected from the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device;

[0012] A photovoltaic energy data extraction module, configured to extract a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected from the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device;

[0013] A photovoltaic panel tilt adjustment judgment module, configured to determine whether the tilt angle of the new energy photovoltaic panel needs to be adjusted based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector.

[0014] Compared with the prior art, a real-time operation monitoring method and system for a new energy power generation system provided by the present application first acquires the photovoltaic panel energy collection values at multiple predetermined time points collected from the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device, then uses deep learning technology to perform feature extraction and correlation analysis on the three, and finally obtains a classification result through a classifier to determine whether the tilt angle of the new energy photovoltaic panel needs to be adjusted, so as to keep the photovoltaic system in the best working state and improve the stability and reliability of the system. Description of the Drawings

[0015] The above and other objects, features, and advantages of the present application will become more apparent by describing the embodiments of the present application in more detail with reference to the accompanying drawings. 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 to the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0016] Figure 1 It is a flowchart of a method for real-time monitoring of the operation of a new energy power generation system according to an embodiment of the present application.

[0017] Figure 2 It is a flowchart of feature extraction for the photovoltaic panel energy collection values at multiple predetermined time points collected by the database and the photovoltaic panel historical energy collection text data collected from the database in the method for real-time monitoring of the operation of a new energy power generation system according to an embodiment of the present application.

[0018] Figure 3 It is a flowchart of feature encoding for the photovoltaic panel energy collection values at multiple predetermined time points collected by the database in the method for real-time monitoring of the operation of a new energy power generation system according to an embodiment of the present application.

[0019] Figure 4 It is a flowchart of semantic encoding for the photovoltaic panel historical energy collection text data collected from the database in the method for real-time monitoring of the operation of a new energy power generation system according to an embodiment of the present application.

[0020] Figure 5 It is a flowchart of feature extraction for the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device in the method for real-time monitoring of the operation of a new energy power generation system according to an embodiment of the present application.

[0021] Figure 6 It is a block diagram of a real-time monitoring system for the operation of a new energy power generation system according to an embodiment of the present application.

[0022] Figure 7 It is a block diagram of an electronic device according to an embodiment of the present application. Detailed Embodiments

[0023] Hereinafter, exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited by the exemplary embodiments described herein.

[0024] Exemplary method

[0025] Figure 1The flowchart of the real-time operation monitoring method for a new energy power generation system according to an embodiment of the present application. As Figure 1 shown, the real-time operation monitoring method for a new energy power generation system according to an embodiment of the present application includes: S110, obtaining the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the inclination angle values of the photovoltaic panel at multiple predetermined time points collected by the inclination angle measurement device; S120, extracting a photovoltaic panel energy correlation feature vector and a photovoltaic panel inclination angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the inclination angle values of the photovoltaic panel at multiple predetermined time points collected by the inclination angle measurement device; S130, based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel inclination angle feature vector, determining whether the inclination angle of the new energy photovoltaic panel needs to be adjusted.

[0026] In the above real-time monitoring method for the operation of a new energy power generation system, in step S110, obtain the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the historical energy collection text data of the photovoltaic panel collected from the database, and the inclination angle values of the photovoltaic panel at multiple predetermined time points collected by the inclination angle measuring device. It should be understood that new energy refers to an energy form that uses renewable resources or energy technologies to replace traditional fossil fuels. Among them, solar energy, as one of the important new energy sources, converts the radiant energy of sunlight into electrical energy or heat energy, and has the characteristics of being clean, renewable, and pollution-free. In the energy collection design of solar energy, that is, in the design of a photovoltaic power station, the optimal inclination angle is generally selected based on the historical data of the annual cumulative radiation amount at different inclination angles. However, simply fixing at the angle with the highest annual cumulative radiation amount may not be the most ideal choice. This is because the irradiation angle of sunlight changes every day, which will cause a reduction in the direct radiation received by the photovoltaic module, thereby affecting the power generation efficiency. Therefore, dynamically adjusting the inclination angle of the photovoltaic panel to maximize the energy collection efficiency is an important consideration in the design optimization of a photovoltaic power station. However, simply detecting and only monitoring the inclination angle without monitoring the energy collection value will result in an inability to understand the actual energy collection efficiency of the photovoltaic panel. Although the inclination angle is important, the actual energy collection value is the key indicator for evaluating the performance of the photovoltaic panel. The lack of energy collection value data makes it impossible to comprehensively evaluate and compare the performance of the photovoltaic panel. Among them, without the historical energy collection text data of the photovoltaic panel, the analysis of the long-term performance changes of the photovoltaic panel will be restricted. And historical data is crucial for understanding the performance and trends of the photovoltaic panel under different conditions. The lack of a complete data set will affect the optimization decision-making of the photovoltaic panel system. Therefore, the technical solution of this application combines data and uses deep learning technology to achieve real-time monitoring and optimized operation of the new energy power generation system. Specifically, real-time monitoring of the photovoltaic panel energy collection value can help evaluate the current energy generation situation of the photovoltaic panel, timely discover the energy generation situation of the photovoltaic panel at different time points, contribute to adjusting the system operation to optimize energy utilization, and improve the performance and efficiency of the monitoring system. Analyzing historical energy collection data can help identify the long-term energy collection trend of the photovoltaic panel and provide a reference for future system optimization. Historical energy collection data can be used to compare the energy collection situations in different time periods and help identify potential problems or improvement points. And the inclination angle of the photovoltaic panel affects the light reception situation, thereby affecting the energy collection efficiency. Monitoring the inclination angle of the photovoltaic panel can help determine whether the angle of the photovoltaic panel needs to be adjusted to receive sunlight to the greatest extent and improve the energy collection efficiency. Generally speaking, obtaining these data is to comprehensively understand the energy collection situation of the photovoltaic panel, including real-time data, historical data, and the inclination angle of the photovoltaic panel, so as to achieve real-time monitoring and optimized operation of the new energy power generation system. By analyzing these data, the system parameters can be adjusted in a timely manner to improve the energy utilization efficiency, ensure the stable operation of the system, and maximize the energy collection.

[0027] In the real-time monitoring method for the operation of the above new energy power generation system, in step S120, a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector are extracted from the photovoltaic panel energy collection values at multiple predetermined time points collected from the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device. It should be understood that by extracting the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector, the original data can be converted into a more informative and interpretable feature representation form. Feature extraction helps to reduce the dimension of the data, extract the key information in the data, and provide a basis for subsequent analysis and decision-making. Conducting correlation analysis on the photovoltaic panel energy collection values and the tilt angle values can help understand the relationship between them, possibly discover the correlation between the two, and further guide whether it is necessary to adjust the tilt angle of the photovoltaic panel to optimize the energy collection efficiency. Based on these feature vectors, the system can use machine learning or other algorithms for analysis and decision-making, helping the operation and maintenance personnel or the system to automatically adjust the tilt angle of the photovoltaic panel to improve the energy collection efficiency and ensure the stable operation of the system.

[0028] In a specific embodiment of the present application, step S120 of extracting a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected from the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device includes: performing feature extraction on the photovoltaic panel energy collection values at multiple predetermined time points collected from the database and the photovoltaic panel historical energy collection text data collected from the database to obtain the photovoltaic panel energy correlation feature vector; performing feature extraction on the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device to obtain the photovoltaic panel tilt angle feature vector. It should be understood that the original photovoltaic panel energy collection data may contain a large number of dimensions and information, and feature extraction can help reduce the data dimension, extract the most representative and key features, and reduce the data complexity. Among them, the photovoltaic panel energy collection values and the historical energy collection text data may contain a certain correlation relationship. By feature extraction, these correlation features can be extracted, which helps to analyze the operation status and performance of the photovoltaic panel system, convert the data into a form that is easier to analyze and understand, so as to discover the patterns, trends, and rules in the data, and help identify the correlation features of the photovoltaic panel energy collection.

[0029] Furthermore, by extracting features from the tilt angle data, the original data can be transformed into more representative and interpretable feature vectors to better describe the tilt angle characteristics of the photovoltaic panel. Through feature extraction, key features in the tilt angle data, such as periodic variations and outliers, can be identified, thus better understanding the variation law of the tilt angle of the photovoltaic panel. The extraction of the tilt angle feature vector can provide support for system optimization. For example, the angle of the photovoltaic panel can be adjusted according to the change of the tilt angle to maximize the energy collection efficiency.

[0030] Figure 2 The flowchart for feature extraction of the photovoltaic panel energy collection values at multiple predetermined time points collected from the database and the photovoltaic panel historical energy collection text data collected from the database in the operation real-time monitoring method of the new energy power generation system according to the embodiment of the present application. As Figure 2 shown, in a specific embodiment of the present application, feature extraction of the photovoltaic panel energy collection values at multiple predetermined time points collected from the database and the photovoltaic panel historical energy collection text data collected from the database to obtain the photovoltaic panel energy correlation feature vector includes: S210, performing feature encoding on the photovoltaic panel energy collection values at multiple predetermined time points collected from the database to obtain a photovoltaic panel energy collection feature vector; S220, performing semantic encoding on the photovoltaic panel historical energy collection text data collected from the database to obtain a historical energy collection feature vector; S230, associating the photovoltaic panel energy collection feature vector and the historical energy collection feature vector to obtain the photovoltaic panel energy correlation feature vector. It should be understood that through feature encoding, the original photovoltaic panel energy collection values can be converted into a more compact feature vector representation, thereby reducing the dimension and complexity of the data for easy storage and processing. Moreover, feature encoding is an important step in data mining and machine learning, which can convert the original data into a format that can be processed by machine learning algorithms, thus realizing the analysis and prediction of the photovoltaic panel energy collection behavior.

[0031] Furthermore, the extraction of the historical energy collection feature vector can help the system identify patterns and trends in the text data, thus better understanding the evolution and law of the historical energy collection behavior. The historical energy collection feature vector obtained through semantic encoding can provide support for decision-making, helping the system make predictions, optimizations, and adjustments based on historical data to improve the energy collection efficiency. Through semantic encoding, the text data can be converted into a semantic representation to help the system understand the meaning and information in the text data, thus better analyzing and utilizing the historical energy collection data. Semantic encoding helps extract key information, themes, and associations from the text data, converting the unstructured text data into a structured feature vector for subsequent analysis and modeling.

[0032] Furthermore, by combining the photovoltaic panel energy collection feature vector and the historical energy collection feature vector, the energy collection situation at the current moment and the historical energy collection data can be comprehensively considered, providing more comprehensive information for the system. Among them, correlating different types of feature vectors can enrich the feature space, provide more information for data analysis and modeling, and help the system conduct a more comprehensive evaluation of the photovoltaic panel energy collection behavior.

[0033] Figure 3 It is a flowchart for feature encoding of the photovoltaic panel energy collection values at multiple predetermined time points collected by the database in the real-time monitoring method for the operation of the new energy power generation system according to the embodiments of the present application. As Figure 3 shown, in a specific embodiment of the present application, feature encoding of the photovoltaic panel energy collection values at multiple predetermined time points collected by the database to obtain a photovoltaic panel energy collection feature vector includes: S211, arranging the photovoltaic panel energy collection values at multiple predetermined time points collected by the database in the time dimension to obtain a photovoltaic panel energy collection input vector; S212, passing the photovoltaic panel energy collection input vector through a photovoltaic panel energy collection time series encoder to obtain the photovoltaic panel energy collection feature vector. It should be understood that the photovoltaic panel energy collection values change with time, and arranging them in the time dimension can form time series data, facilitating time series analysis, such as trend analysis or periodic analysis. Arranging the energy collection values at multiple predetermined time points in the time dimension can maintain the continuity and time series relationship of the data, helping the system better understand the time correlation between the data. At the same time, the photovoltaic panel energy collection values arranged in the time dimension can be used as input features to help the system extract time-related feature information for modeling and predicting the photovoltaic panel energy collection behavior.

[0034] Furthermore, the time series encoder can effectively capture the time series information in the time series data, and since the photovoltaic panel energy collection values change with time, the patterns and rules hidden in the time series of the photovoltaic panel energy collection values can be extracted. Among them, the time series encoder can convert the time series data into a more representative feature vector, thereby extracting more valuable feature information, which helps the system better understand and analyze the photovoltaic panel energy collection data. The feature vector obtained by conversion through the time series encoder usually has a lower dimension, which can help reduce the complexity of the data, improve the calculation efficiency, and at the same time retain important feature information. Specifically, use the fully connected layer of the photovoltaic panel energy collection time series encoder to perform fully connected encoding on the photovoltaic panel energy collection input vector to extract the high-dimensional hidden features of the feature values at each position in the photovoltaic panel energy collection input vector; and, use the one-dimensional convolutional layer of the photovoltaic panel energy collection time series encoder to perform one-dimensional encoding on the photovoltaic panel energy collection input vector to extract the high-dimensional hidden correlation features of the correlations between the feature values at each position in the photovoltaic panel energy collection input vector.

[0035] Figure 4 A flowchart for semantic encoding of the historical energy collection text data of a photovoltaic panel collected from a database in a real-time monitoring method for an operation of a new energy power generation system according to an embodiment of the present application. As Figure 4 shown, in a specific embodiment of the present application, semantic encoding of the historical energy collection text data of the photovoltaic panel collected from the database to obtain a historical energy collection feature vector includes: S221, performing word segmentation on the historical energy collection text data of the photovoltaic panel collected from the database to obtain a plurality of historical energy collection data items; S222, passing the plurality of historical energy collection data items through a historical energy collection context encoder based on a convolutional neural network to obtain the historical energy collection feature vector. It should be understood that word segmentation can help clean and process the original text data, split the text data into smaller data items, help remove noise and unnecessary information, and improve data quality. Among them, word segmentation can divide text data according to words or phrases, make the data more standardized and organized, and facilitate statistical analysis, visualization, and modeling. Performing word segmentation on text data can convert complex text information into structured data items, making the data easier to understand and process, and being beneficial to subsequent data analysis and applications. The data items after word segmentation can be used as the basic units of features, helping to extract key information and features in the data, and providing a basis for subsequent feature engineering and model training.

[0036] Furthermore, the context encoder based on a convolutional neural network can effectively capture the context information between historical energy collection data items, helping the system better understand the association and order between data items. The convolutional neural network can learn the local features and patterns between data items, extract the features of the data through convolutional operations and pooling operations, and help improve the representation ability of the data and the performance of the model. Moreover, the convolutional neural network has the characteristic of parameter sharing, which can reduce the number of parameters of the model, improve the training efficiency of the model, and can effectively process data with local correlation. Specifically, using the embedding layer of the historical energy collection context encoder based on a convolutional neural network to map each data item in the plurality of historical energy collection data items into a word embedding vector to obtain a sequence of word embedding vectors; using the Transformer-based Bert model of the historical energy collection context encoder based on a convolutional neural network to perform global context semantic encoding on the sequence of word embedding vectors to obtain a plurality of feature vectors; and, cascading the plurality of feature vectors to obtain the historical energy collection feature vector.

[0037] Figure 5It is a flowchart for feature extraction of the tilt angle values of the photovoltaic panels at multiple predetermined time points collected by the tilt angle measurement device in the real-time operation monitoring method of the new energy power generation system according to the embodiments of the present application. As Figure 5 shown, in a specific embodiment of the present application, feature extraction of the tilt angle values of the photovoltaic panels at multiple predetermined time points collected by the tilt angle measurement device to obtain a photovoltaic panel tilt angle feature vector includes: S310, arranging the tilt angle values of the photovoltaic panels at multiple predetermined time points collected by the tilt angle measurement device in the time dimension to obtain a photovoltaic panel tilt angle input vector; S320, passing the photovoltaic panel tilt angle input vector through a photovoltaic panel tilt angle feature extractor based on a multi-scale neighborhood to obtain the photovoltaic panel tilt angle feature vector. It should be understood that the tilt angle values of the photovoltaic panels have a certain temporal sequence as they change over time. Arranging them in the time dimension can better retain the information in the time series, which is beneficial for capturing time-related features. Arranging the tilt angle values in time can make full use of historical data, enabling the model to consider the previous tilt angle change trend and helping to predict future tilt angle values. Many time series prediction models require input data arranged in chronological order to correctly capture time-related patterns and trends. Therefore, arranging the data in the time dimension helps to be compatible with these models.

[0038] Furthermore, the multi-scale neighborhood-based feature extractor can capture the changing patterns of tilt angles at different scales, thereby providing a more comprehensive understanding of the characteristics of the tilt angles of photovoltaic panels. The information fusion at different scales can improve the richness and representational ability of the features. The multi-scale neighborhood-based feature extractor helps to model the spatial relationships between the tilt angles of photovoltaic panels, including local and global information, and enhances the understanding and analysis capabilities of the tilt angle data of photovoltaic panels. Moreover, the multi-scale feature extractor can enhance the feature representation ability of the tilt angle data of photovoltaic panels, making the extracted features more discriminative and representative, which helps to improve the accuracy and performance of subsequent tasks. Combining the multi-scale neighborhood-based feature extractor can improve the performance of the model, enabling the model to better understand the characteristics of the tilt angle data of photovoltaic panels. Specifically, the photovoltaic panel tilt angle feature extractor based on multi-scale neighborhood includes: a first convolutional layer, a second convolutional layer parallel to the first convolutional layer, and a concatenation layer connected to the first convolutional layer and the second convolutional layer, where the first convolutional layer uses a one-dimensional convolutional kernel with a first scale, and the second convolutional layer uses a one-dimensional convolutional kernel with a second scale. More specifically, use the first convolutional layer of the photovoltaic panel tilt angle feature extractor based on multi-scale neighborhood to perform one-dimensional convolutional encoding on the photovoltaic panel tilt angle input vector to obtain a first-scale feature vector; use the second convolutional layer of the photovoltaic panel tilt angle feature extractor based on multi-scale neighborhood to perform one-dimensional convolutional encoding on the photovoltaic panel tilt angle input vector to obtain a second-scale feature vector; and concatenate the first-scale feature vector and the second-scale feature vector to obtain the photovoltaic panel tilt angle feature vector.

[0039] In the above real-time operation monitoring method of the new energy power generation system, in step S130, based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector, it is determined whether the tilt angle of the new energy photovoltaic panel needs to be adjusted. It should be understood that by analyzing the photovoltaic panel energy correlation feature vector and the tilt angle feature vector, the energy output situation of the photovoltaic panel can be evaluated. If the tilt angle has a significant impact on the energy output, adjusting the tilt angle of the photovoltaic panel may improve the energy efficiency. Adjusting the tilt angle of the photovoltaic panel can help the photovoltaic panel obtain the maximum solar radiation within a specific time period, thereby increasing the power output of the photovoltaic panel. By analyzing the tilt angle feature vector, it can be determined whether the current tilt angle affects the maximum power output.

[0040] In a specific embodiment of the present application, in step S130, the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector are fused to obtain a photovoltaic panel tilt angle judgment feature vector; the photovoltaic panel tilt angle judgment feature vector is subjected to feature space reconstruction based on the intrinsic mapping to obtain an optimized photovoltaic panel tilt angle judgment feature vector; the optimized photovoltaic panel tilt angle judgment feature vector is passed through a classifier to obtain a classification result, and the classification result is used to judge whether the tilt angle of the new energy photovoltaic panel needs to be adjusted. It should be understood that the energy correlation feature vector and the tilt angle feature vector provide information in different aspects. Fusing these two types of features can make the feature vector more abundant, reduce the limitations brought by a single feature, thereby including more information about the state of the photovoltaic panel, and helping to improve the reliability of the judgment. In one embodiment, the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector are fused through a cascade function to obtain a photovoltaic panel tilt angle judgment feature vector.

[0041] Specifically, although the photovoltaic panel energy collection input vector and the tilt angle input vector can reflect the energy output situation and physical posture of the photovoltaic panel at multiple predetermined time points, this method of arranging based on the time dimension mainly focuses on the numerical change trend and fails to deeply explore the more complex patterns and interactions hidden behind. For example, the energy output of the photovoltaic panel is not only affected by the solar radiation intensity and weather conditions, but is also closely related to its installation location, surrounding environment occlusion and other factors, and there are non-linear and dynamically changing relationships among these factors. Similarly, although the historical energy collection text data provides rich background information, it may be difficult to comprehensively capture the details such as historical experience and performance in special situations contained in the text only by relying on word segmentation processing and convolutional neural network encoding. In addition, simply associating the photovoltaic panel energy collection feature vector with the historical energy collection feature vector, and subsequent fusion with the tilt angle feature vector, although integrating different types of data resources to a certain extent, this fusion method may more be a combination of features at the surface level, and does not fully consider the potential deep connections among the features and their performance differences under different conditions. Therefore, for those cases that rely on the internal structure information of the features for accurate analysis, this method may cause some subtle but key feature information to be ignored or not fully utilized, thus affecting the accuracy and reliability of the final classification result, and further causing deviation in the judgment of whether the tilt angle of the photovoltaic panel needs to be adjusted. Therefore, further, the photovoltaic panel tilt angle judgment feature vector is subjected to feature space reconstruction based on the intrinsic mapping to obtain an optimized photovoltaic panel tilt angle judgment feature vector.

[0042] Specifically, perform feature space reconstruction based on intrinsic mapping on the photovoltaic panel tilt angle judgment eigenvector to obtain an optimized photovoltaic panel tilt angle judgment eigenvector, including: First, construct a fine-grained association topology matrix of the photovoltaic panel tilt angle judgment eigenvector, which is expressed by the formula:

[0043]

[0044] where V represents the photovoltaic panel tilt angle judgment eigenvector, v i and v j represent the eigenvalues at the i-th and j-th positions of the photovoltaic panel tilt angle judgment eigenvector respectively, d(v i , v j ) represents calculating the Euclidean distance, and D i,j represents the eigenvalue at the (i, j) position of the fine-grained association topology matrix.

[0045] That is, by constructing a fine-grained association topology matrix, the association strength or similarity between the feature information at different positions in the photovoltaic panel tilt angle judgment eigenvector can be explicitly quantified, such as the dynamic coupling relationship between the tilt angle change and the energy collection efficiency fluctuation within the same time period. Specifically, by encoding the complex dependence relationships within the features, the judgment accuracy of the model for the photovoltaic panel tilt angle adjustment requirements can be improved, while enhancing the adaptability of the model to dynamic environmental changes, ensuring that the system can still achieve optimal angle decision-making based on data-driven association topology under complex lighting conditions.

[0046] Second, use a convolution operator to perform deep feature extraction on the fine-grained association topology matrix to obtain a photovoltaic panel tilt angle kernel space dynamic response matrix, which is expressed by the formula:

[0047] M = Conv(D)

[0048] where D represents the fine-grained association topology matrix, Conv represents the convolutional layer, and M represents the photovoltaic panel tilt angle kernel space dynamic response matrix.

[0049] That is, through the convolution kernel, the discrete fine-grained association topology matrix is transformed into a task-oriented dynamic response representation, revealing the potential causal chain or dynamic balance mechanism between the angle adjustment decision and the energy efficiency fluctuation. Specifically, through the data-driven learned photovoltaic panel tilt angle kernel space dynamic response matrix, the differential association laws of the angle adjustment strategies of the photovoltaic panel under different environmental conditions (such as seasonal alternation, cloud occlusion) can be characterized, such as identifying the response feature differences between small-amplitude high-frequency adjustments under sunny conditions and large-amplitude low-frequency adjustments under rainy conditions, and avoiding misjudgment caused by local association noise.

[0050] Then, perform tensor domain eigen - decomposition on the photovoltaic panel tilt - angle judgment eigen - vector to obtain a set of photovoltaic panel tilt - angle intrinsic eigen - decomposition vectors, which is expressed by the formula:

[0051]

[0052] where \(T\) represents the transpose of the vector, \(\Lambda\) represents the diagonal matrix, \(\lambda_1\) and \(\lambda\) m respectively represent the first and the \(m\) - th eigenvalues on the diagonal of the diagonal matrix, \(U\) represents the set of photovoltaic panel tilt - angle intrinsic eigen - decomposition vectors, and \(x_1\), \(x_2\), \(x\) m respectively represent the first, the \(i\) - th and the \(m\) - th photovoltaic panel tilt - angle intrinsic eigen - decomposition vectors.

[0053] That is, through tensor domain eigen - decomposition, the high - dimensional feature space can be mapped to the latent subspace spanned by the orthogonal basis vectors. The generated set of photovoltaic panel tilt - angle intrinsic eigen - decomposition vectors can suppress the interference of noise components on the classification result, and at the same time enhance the generalization ability of the model to the implicit rules in historical data, ensuring that the photovoltaic panel can still achieve precise angle - adjustment triggering based on the decomposed core feature patterns in a highly dynamic environment.

[0054] Next, input each photovoltaic panel tilt - angle intrinsic eigen - decomposition vector in the set of photovoltaic panel tilt - angle intrinsic eigen - decomposition vectors into the dynamic weight - assignment module based on the self - attention mechanism to obtain a set of photovoltaic panel tilt - angle modulation feature response vectors, which is expressed by the formula:

[0055] \(Y=\text{Transformer}\{[x_1,x_2,\cdots,x\}\) m \(=[y_1,y_2,\cdots,y\}\) m

[0056] where \(\text{Transformer}\) represents the transformer model, \(Y\) represents the set of photovoltaic panel tilt - angle modulation feature response vectors, and \(y_1\), \(y_2\), \(y\) m respectively represent the first, the \(i\) - th and the \(m\) - th photovoltaic panel tilt - angle modulation feature response vectors.

[0057] It should be understood that the self - attention mechanism can dynamically identify context - aware weight - assignment requirements such as "the enhanced correlation between instantaneous fluctuation components and historical compensation strategy components under cloudy weather" or "the long - term trend component dominates the decision - making during the high - radiation period at noon" by calculating the mutual - dependence relationships between each photovoltaic panel tilt - angle intrinsic eigen - decomposition vector. This enables the modulated photovoltaic panel tilt - angle modulation feature response vectors to focus on the currently most relevant angle - adjustment driving factors, thereby improving the decision - making sensitivity and anti - interference ability of the classifier in a complex dynamic environment, and at the same time enhancing the generalization ability to capture low - frequency but high - value patterns in historical data. ​

[0058] Then, project each photovoltaic panel tilt angle modulation feature response vector in the set of photovoltaic panel tilt angle modulation feature response vectors onto the photovoltaic panel tilt angle kernel space dynamic response matrix to obtain a set of photovoltaic panel tilt angle feature mask coding vectors, which is expressed by the formula:

[0059]

[0060] Wherein, represents matrix multiplication, S represents the characteristic scale of the photovoltaic panel tilt angle kernel space dynamic response matrix, y i represents the i-th photovoltaic panel tilt angle modulation feature response vector, L represents the length of the photovoltaic panel tilt angle modulation feature response vector, z i represents the i-th photovoltaic panel tilt angle feature mask coding vector.

[0061] That is, by projecting the set of photovoltaic panel tilt angle modulation feature response vectors onto the photovoltaic panel tilt angle kernel space dynamic response matrix, a non-linear interaction framework of global structure constraint and local dynamic response is constructed. Specifically, through the context awareness of the feature kernel domain and the dynamic adaptation of the decoupled structure, the global stability constraint and the local dynamic sensitivity are mapped to a unified feature space, so that the generated set of photovoltaic panel tilt angle feature mask coding vectors not only retains the physical interpretability of the tilt angle adjustment, but also dynamically integrates the non-linear correction of the environmental variables on the angle sensitivity, thereby providing a physically robust high-dimensional feature representation for subsequent tilt angle optimization decisions.

[0062] Finally, fuse the set of photovoltaic panel tilt angle feature mask coding vectors to obtain an optimized photovoltaic panel tilt angle judgment feature vector, which is expressed by the formula:

[0063] V' = Concat{z1, z2, …, z m}

[0064] Wherein, Concat represents the concatenation function, z1, z2, z m respectively represent the first, the i-th and the m-th photovoltaic panel tilt angle feature mask coding vectors, and V' represents the optimized photovoltaic panel tilt angle judgment feature vector.

[0065] That is, by fusing the set of photovoltaic panel tilt angle feature mask coding vectors, an optimization framework of multi-perspective feature complementarity and dynamic weight adaptation is constructed. Through the non-linear context interaction mechanism, the global structure constraint and the real-time dynamic response are mapped to a unified feature space, so that the generated optimized photovoltaic panel tilt angle judgment feature vector not only retains the physical boundary constraint of the tilt angle adjustment, but also dynamically integrates the quantitative influence of the environmental variables on the angle sensitivity, thereby providing a high-dimensional feature representation with both physical interpretability and environmental self-adaptability for subsequent tilt angle optimization decisions.

[0066] Furthermore, the classifier can learn the patterns and rules in the training data, so as to more accurately judge whether the tilt angle of the new energy photovoltaic panel needs to be adjusted, avoiding the interference of subjective factors. With the help of the classifier, the tilt angle of the photovoltaic panel can be quickly judged, and it can be timely found whether adjustment is needed, so as to take actions for optimization more quickly.

[0067] In summary, the embodiment of the present application first obtains the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device. Then, using deep learning technology, feature extraction and correlation analysis are performed on the three. Finally, a classification result is obtained through the classifier to judge whether the tilt angle of the new energy photovoltaic panel needs to be adjusted, so as to keep the photovoltaic system in the best working state and improve the stability and reliability of the system.

[0068] Exemplary system

[0069] Figure 6 It is a block diagram of an operation real-time monitoring system for a new energy power generation system according to an embodiment of the present application. As Figure 6 shown, the operation real-time monitoring system 100 for a new energy power generation system according to an embodiment of the present application includes: a photovoltaic energy data acquisition module 110, configured to obtain the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device; a photovoltaic energy data extraction module 120, configured to extract a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device; a photovoltaic panel tilt adjustment judgment module 130, configured to judge whether the tilt angle of the new energy photovoltaic panel needs to be adjusted based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector.

[0070] As described above, the real-time operation monitoring system 100 of the new energy power generation system according to the embodiments of the present application can be implemented in various terminal devices, such as a server deployed with a real-time operation monitoring algorithm for the new energy power generation system. In one example, the real-time operation monitoring system 100 of the new energy power generation system can be integrated into the terminal device as a software module and / or a hardware module. For example, the real-time operation monitoring system 100 of the new energy power generation system can be a software module in the operating system of the terminal device, or can be an application program developed for the terminal device; of course, the real-time operation monitoring system 100 of the new energy power generation system can also be one of the numerous hardware modules of the terminal device.

[0071] Alternatively, in another example, the real-time operation monitoring system 100 of the new energy power generation system and the terminal device can also be separate devices, and the real-time operation monitoring system 100 of the new energy power generation system can be connected to the terminal device through a wired and / or wireless network, and transmit and interact information according to a predefined data format.

[0072] Here, those skilled in the art can understand that the specific operations of each step in the above real-time operation monitoring system of the new energy power generation system have been introduced in detail in the description of the real-time operation monitoring method of the new energy power generation system referred to above, and therefore, the repeated description thereof will be omitted. Figures 1 to 5 The real-time operation monitoring method of the new energy power generation system has been introduced in detail, and therefore, the repeated description thereof will be omitted.

[0073] Exemplary electronic device

[0074] Next, reference will be made to Figure 7 to describe the electronic device according to the embodiments of the present application.

[0075] As Figure 7 shown, the electronic device 10 includes an input device 11, an input interface 12, a central processing unit 13, a memory 14, an output interface 15, an output device 16, and a bus 17. Among them, the input interface 12, the central processing unit 13, the memory 14, and the output interface 15 are connected to each other through the bus 17, and the input device 11 and the output device 16 are respectively connected to the bus 17 through the input interface 12 and the output interface 15, and then connected to other components of the electronic device 10.

[0076] Specifically, the input device 11 receives external input information and transmits the input information to the central processing unit 13 through the input interface 12; the central processing unit 13 processes the input information based on the computer-executable instructions stored in the memory 14 to generate output information, temporarily or permanently stores the output information in the memory 14, and then transmits the output information to the output device 16 through the output interface 15; the output device 16 outputs the output information to the outside of the electronic device 10 for the user to use.

[0077] In one embodiment, Figure 7 The illustrated electronic device 10 may be implemented as a network device, which may include: a memory configured to store a program; and a processor configured to run the program stored in the memory to execute the real-time operation monitoring method of any one of the new energy power generation systems described in the above embodiments.

[0078] According to an embodiment of the present application, the process described above with reference to the flowchart may be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program tangibly embodied on a machine-readable medium, and the computer program includes program code for performing the method shown in the flowchart. In such an embodiment, the computer program may be downloaded and installed from a network, and / or installed from a removable storage medium.

[0079] Those of ordinary skill in the art can understand that all or some of the steps in the methods disclosed above, and the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, and appropriate combinations thereof. In a hardware implementation, the division of the functional modules / units mentioned above does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be executed by several physical components in cooperation. Some or all of the physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or transitory medium). As is well known to those of ordinary skill in the art, the term computer storage medium includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, as is well known to those of ordinary skill in the art, communication media typically includes computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transmission mechanism, and may include any information delivery medium.

[0080] It is understandable that the above embodiments are merely exemplary embodiments adopted to illustrate the principle of the present application. However, the present application is not limited thereto. For those of ordinary skill in the art, various modifications and improvements can be made without departing from the spirit and essence of the present application, and these modifications and improvements are also regarded as the protection scope of the present application.

Claims

1. A real-time monitoring method for the operation of a new energy power generation system, characterized in that, Including: Obtaining the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device; Extracting a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device; Based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector, determining whether the tilt angle of the new energy photovoltaic panel needs to be adjusted.

2. The real-time monitoring method for the operation of the new energy power generation system according to claim 1, characterized in that Extracting a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device, including: Performing feature extraction on the photovoltaic panel energy collection values at multiple predetermined time points collected by the database and the photovoltaic panel historical energy collection text data collected from the database to obtain the photovoltaic panel energy correlation feature vector; Performing feature extraction on the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device to obtain the photovoltaic panel tilt angle feature vector.

3. The real-time monitoring method for the operation of the new energy power generation system according to claim 2, characterized in that Performing feature extraction on the photovoltaic panel energy collection values at multiple predetermined time points collected by the database and the photovoltaic panel historical energy collection text data collected from the database to obtain the photovoltaic panel energy correlation feature vector, including: Performing feature encoding on the photovoltaic panel energy collection values at multiple predetermined time points collected by the database to obtain a photovoltaic panel energy collection feature vector; Performing semantic encoding on the photovoltaic panel historical energy collection text data collected from the database to obtain a historical energy collection feature vector; Associating the photovoltaic panel energy collection feature vector and the historical energy collection feature vector to obtain the photovoltaic panel energy correlation feature vector.

4. The real-time monitoring method for the operation of the new energy power generation system according to claim 3, characterized in that, Performing feature encoding on the photovoltaic panel energy collection values at multiple predetermined time points collected by the database to obtain a photovoltaic panel energy collection feature vector, including: Arranging the photovoltaic panel energy collection values at multiple predetermined time points collected by the database in the time dimension to obtain a photovoltaic panel energy collection input vector; Passing the photovoltaic panel energy collection input vector through a photovoltaic panel energy collection time series encoder to obtain the photovoltaic panel energy collection feature vector.

5. The real-time monitoring method for the operation of the new energy power generation system according to claim 4, characterized in that, Performing semantic encoding on the photovoltaic panel historical energy collection text data collected from the database to obtain a historical energy collection feature vector, including: Performing word segmentation on the photovoltaic panel historical energy collection text data collected from the database to obtain multiple historical energy collection data items; Passing the multiple historical energy collection data items through a historical energy collection context encoder based on a convolutional neural network to obtain the historical energy collection feature vector.

6. The real-time monitoring method for the operation of the new energy power generation system according to claim 5, characterized in that Feature extraction is performed on the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device to obtain a photovoltaic panel tilt angle feature vector, including: Arrange the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device in the time dimension to obtain a photovoltaic panel tilt angle input vector; Pass the photovoltaic panel tilt angle input vector through a photovoltaic panel tilt angle feature extractor based on a multi-scale neighborhood to obtain the photovoltaic panel tilt angle feature vector.

7. The real-time monitoring method for the operation of the new energy power generation system according to claim 6, characterized in that, Based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector, determine whether the tilt angle of the new energy photovoltaic panel needs to be adjusted, including: Fuse the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector to obtain a photovoltaic panel tilt angle judgment feature vector; Perform feature space reconstruction based on the intrinsic mapping on the photovoltaic panel tilt angle judgment feature vector to obtain an optimized photovoltaic panel tilt angle judgment feature vector; Pass the optimized photovoltaic panel tilt angle judgment feature vector through a classifier to obtain a classification result, and the classification result is used to determine whether the tilt angle of the new energy photovoltaic panel needs to be adjusted.

8. The real-time monitoring method for the operation of the new energy power generation system according to claim 7, characterized in that, Perform feature space reconstruction based on the intrinsic mapping on the photovoltaic panel tilt angle judgment feature vector to obtain an optimized photovoltaic panel tilt angle judgment feature vector, including: Construct a per-granularity association topology matrix of the photovoltaic panel tilt angle judgment feature vector; Use a convolution operator to perform deep feature extraction on the per-granularity association topology matrix to obtain a photovoltaic panel tilt angle kernel space dynamic response matrix; Perform tensor domain feature decomposition on the photovoltaic panel tilt angle judgment feature vector to obtain a set of photovoltaic panel tilt angle intrinsic feature decomposition vectors; Input each photovoltaic panel tilt angle intrinsic feature decomposition vector in the set of photovoltaic panel tilt angle intrinsic feature decomposition vectors into a dynamic weight allocation module based on a self-attention mechanism to obtain a set of photovoltaic panel tilt angle modulation feature response vectors; Project each photovoltaic panel tilt angle modulation feature response vector in the set of photovoltaic panel tilt angle modulation feature response vectors onto the photovoltaic panel tilt angle kernel space dynamic response matrix to obtain a set of photovoltaic panel tilt angle feature mask coding vectors; Fuse the set of photovoltaic panel tilt angle feature mask coding vectors to obtain an optimized photovoltaic panel tilt angle judgment feature vector.

9. An operation real-time monitoring system for a new energy power generation system, characterized in that, Including: A photovoltaic energy data acquisition module, configured to acquire the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device; A photovoltaic energy data extraction module, configured to extract a photovoltaic panel energy correlation feature vector and a photovoltaic panel tilt angle feature vector from the photovoltaic panel energy collection values at multiple predetermined time points collected by the database, the photovoltaic panel historical energy collection text data collected from the database, and the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measurement device; A photovoltaic panel tilt adjustment judgment module, which is used to judge whether the tilt angle of the new energy photovoltaic panel needs to be adjusted based on the photovoltaic panel energy correlation feature vector and the photovoltaic panel tilt angle feature vector.

10. The real-time operation monitoring system of the new energy power generation system according to claim 9, characterized in that, The photovoltaic energy data extraction module includes: Performing feature extraction on the photovoltaic panel energy collection values at multiple predetermined time points collected from the database and the photovoltaic panel historical energy collection text data collected from the database to obtain the photovoltaic panel energy correlation feature vector; Performing feature extraction on the tilt angle values of the photovoltaic panel at multiple predetermined time points collected by the tilt angle measuring device to obtain the photovoltaic panel tilt angle feature vector.

Images