Intelligent compression method and system for power grid data and electronic equipment

Abstract

The invention discloses a power grid data intelligent compression method and system and electronic equipment, and the method comprises the steps: determining the global importance information of each piece of equipment state detection data through the correlation information between all pieces of equipment state detection data, and determining the importance type of each piece of equipment state detection data based on the importance information; therefore, data compression of the smart grid is realized based on the importance category, and the whole safe operation of the smart grid is ensured.

Description

Technical Field

[0001] This invention relates to the field of smart grid data compression, and more specifically, to a method, system, and electronic device for intelligent compression of power grid data. Background Technology

[0002] In response to the challenges posed to traditional power systems by global energy shortages and environmental pollution, and to better ensure the safety, reliability, efficiency, flexibility, and environmental friendliness of power systems, the concept of "Smart Grid" has been proposed in recent years. The ultimate goal of Smart Grid is to build a panoramic real-time monitoring system covering the entire production process of the power system.

[0003] In smart grid systems, power equipment condition monitoring data is generated at various stages of the entire power system, such as the condition monitoring data of important equipment like transmission lines, transformers, and circuit breakers. With the Internet of Things (IoT) technology providing significant technical support for comprehensive condition monitoring in power systems, a large amount of condition monitoring data is stored. This condition monitoring data is of great practical significance for analyzing the operating status of power equipment, diagnosing faults, and providing references for equipment maintenance plans.

[0004] As voltage levels continue to rise, the requirements for the stability and security of power systems are also increasing. Power equipment condition monitoring is becoming more comprehensive, and this monitoring data reflects the development trends of power equipment's operating status, possessing immense research value. Furthermore, in acquiring information about the operating status of power equipment, the number of monitoring data collection points is increasing, and the types of equipment condition monitoring methods in power systems are also expanding. The collected monitoring data describes "every detail" of the operating status of each piece of equipment. With the continuous improvement of monitoring equipment systems, the amount of equipment condition monitoring data will grow exponentially. This monitoring data requires high reliability; the volume of condition monitoring data is large, the sources are diverse, and there may be multiple descriptions of the same state characteristic. Moreover, the collected monitoring data often contains uncertainties such as noise and missing data, resulting in a scarcity of valuable, key characteristic data that effectively reflects the equipment's status. Furthermore, the large amount of monitoring data recorded will inevitably increase the difficulty of data processing, and valuable condition monitoring information remains untapped.

[0005] Faced with such complex high-dimensional data, not all equipment condition monitoring data attributes are important or meaningful. Many of these attributes are redundant and repetitive, and not all monitoring data is valuable. Analyzing all condition monitoring data would significantly increase machine processing time and make it more difficult to accurately analyze and judge the condition of power equipment. Therefore, it is of great practical significance to extract important condition indicators from these condition monitoring data, delete useless and repetitive indicators, and directly analyze key condition indicators. This analysis and processing of equipment condition monitoring data, as well as providing useful auxiliary decision-making for the safe operation and production of the entire power system, is crucial.

[0006] Therefore, in order to determine the importance category of each device status monitoring data in order to achieve data compression, an intelligent power grid data compression method is desired. Summary of the Invention

[0007] To address the aforementioned technical problems, this application is proposed. Embodiments of this application provide an intelligent power grid data compression method, system, and electronic device. It calculates the importance index of each device status detection point as a node, and obtains an importance index normalization matrix through logical recursion based on a normalization function in the row and column directions of the importance index. This extracts the importance vectorized information between the nodes. Furthermore, by averaging these values, the probability that all nodes point to the i-th node in a two-dimensional direction can be represented, thus indicating whether the importance of each node is equal at the data level. This can be used to determine the importance category of each device status detection data point, and then achieve data compression based on the importance category to ensure the safe operation and production of the entire power system.

[0008] According to one aspect of this application, a method for intelligent compression of power grid data is provided, comprising: Obtain status monitoring data of various devices in the smart grid; State characterization feature values ​​are extracted from the state monitoring data of each device, and the state characterization feature values ​​of each device are aggregated to obtain a state feature vector; Based on the state feature vector, an importance index matrix is ​​constructed; Based on the importance index matrix, the importance ratio value of the status monitoring data of each device is determined; and Based on the probability value, the compression ratio of the status monitoring data of each device in the smart grid is determined.

[0009] In the intelligent power grid data compression method according to this application, extracting state characterization feature values ​​from the state monitoring data of each device includes: After arranging the state monitoring data of each device into a state input vector according to the sample dimension, the state input vector is input into an encoder model containing a one-dimensional convolutional layer and a fully connected layer to obtain a device state representation vector; and The global mean of the device state representation vector is calculated as the state representation feature value.

[0010] In the intelligent compression method for power grid data according to this application, constructing an importance index matrix based on the state feature vector includes: calculating the importance index between feature values ​​at any two positions in the state feature vector to obtain multiple importance indices; and arranging the multiple importance indices in a matrix to obtain the importance index matrix.

[0011] In the intelligent compression method for power grid data according to this application, the importance index between feature values ​​at any two positions in the state feature vector is calculated to obtain multiple importance indices, including: using the following importance index calculation formula to calculate the importance index between feature values ​​at any two positions in the state feature vector; The importance index is calculated using the following formula: ,in and These are the th elements of the state feature vector. and the The feature values ​​at each position, where... and These are hyperparameters, The value represents the tendency to favor nodes with high feature values, while This indicates a tendency to favor nodes with similar attribute scores.

[0012] In the intelligent compression method for power grid data according to this application, determining the importance ratio value of the status monitoring data of each device based on the importance index matrix includes: processing the importance index matrix using a normalization function to obtain an importance index normalization matrix; calculating the mean of each row and the mean of each column in the importance index normalization matrix to obtain a horizontal correlation feature vector and a vertical correlation feature vector of device status; calculating the weighted sum of the horizontal correlation feature vector and the vertical correlation feature vector of device status to obtain a classification feature vector; inputting the classification feature vector into a classifier to calculate the probability value of each position in the classification feature vector as the importance ratio value of the status monitoring data of the device corresponding to that position; and determining the importance ratio value as the compression ratio of the status monitoring data of each device in the smart grid.

[0013] In the intelligent compression method for power grid data according to this application, inputting the classification feature vector into a classifier to calculate the probability value of each position in the classification feature vector as the importance ratio value of the status monitoring data of the device corresponding to that position includes: inputting the classification feature vector into the class-Softmax classification function of the classifier to calculate the class-Softmax function value of each position in the classification feature vector as the probability value of each position in the classification feature vector.

[0014] According to another aspect of this application, a smart power grid data compression system is provided, comprising: The data acquisition unit is used to obtain status monitoring data of various devices in the smart grid; The encoding unit is used to extract state characterization feature values ​​from the state monitoring data of each device, and aggregate the state characterization feature values ​​of each device to obtain a state feature vector; Importance index measurement unit, used to construct an importance index matrix based on the state feature vector; An importance determination unit is used to determine the importance ratio value of the status monitoring data of each device based on the importance index matrix; and The compression ratio determination unit is used to determine the compression ratio of the status monitoring data of each device in the smart grid based on the probability value.

[0015] In the intelligent power grid data compression system according to this application, the encoding unit is further configured to: arrange the state monitoring data of each device into a state input vector according to the sample dimension, input the state input vector into an encoder model containing a one-dimensional convolutional layer and a fully connected layer to obtain a device state representation vector; and calculate the global mean of the device state representation vector as the state representation feature value.

[0016] In the intelligent power grid data compression system according to this application, the importance index measurement unit is further configured to: calculate the importance index between the feature values ​​at any two positions in the state feature vector to obtain a plurality of importance indices; and arrange the plurality of importance indices in a matrix to obtain the importance index matrix.

[0017] According to another aspect of this application, an electronic device is provided, comprising: a processor; and a memory storing computer program instructions, which, when executed by the processor, cause the processor to perform the intelligent power grid data compression method as described above.

[0018] According to yet another aspect of this application, a computer-readable medium is provided having computer program instructions stored thereon, which, when executed by a processor, cause the processor to perform the intelligent power grid data compression method as described above.

[0019] Compared with existing technologies, the intelligent power grid data compression method, system, and electronic device provided in this application calculates the importance index of each device status detection point as a node, and obtains the importance index normalization matrix through logical recursion based on the normalization function in the row and column directions of the importance index. This extracts the importance vectorization information between the nodes. Furthermore, by averaging these values, the probability that all nodes point to the i-th node in a two-dimensional direction can be represented, thus indicating whether the importance of each node is equal at the data level. This can be used to determine the importance category of each device status detection data, and then achieve data compression based on the importance category to ensure the safe operation and production of the entire power system. Attached Figure Description

[0020] The above and other objects, features, and advantages of this application will become more apparent from the more detailed description of the embodiments of this application in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of this application and form part of the specification. They are used together with the embodiments of this application to explain this application and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same components or steps.

[0021] Figure 1 This is a flowchart of an intelligent power grid data compression method according to an embodiment of this application; Figure 2 This is a schematic diagram of the system architecture of the intelligent power grid data compression method according to an embodiment of this application; Figure 3 This is a flowchart illustrating the process of extracting state characterization feature values ​​from the state monitoring data of each device and aggregating the state characterization feature values ​​of each device to obtain a state feature vector in the intelligent compression method for power grid data according to an embodiment of this application. Figure 4 This is a block diagram of an intelligent power grid data compression system according to an embodiment of this application; Figure 5 This is a block diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0022] Hereinafter, exemplary embodiments according to this application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments of this application. It should be understood that this application is not limited to the exemplary embodiments described herein.

[0023] Scene Overview As mentioned earlier, with the continuous increase in voltage levels, the requirements for the stability and security of power systems are also becoming increasingly stringent. The condition monitoring of power equipment is becoming more comprehensive, and this monitoring data reflects the development trend of the operating status of power equipment, possessing significant research value. Furthermore, in acquiring information about the operating status of power equipment, the number of monitoring data collection points is increasing, and the types of equipment condition monitoring methods in power systems are also expanding. The collected monitoring data describes "every detail" of the operating status of each piece of equipment. With the continuous improvement of monitoring equipment systems, the amount of equipment condition monitoring data will grow exponentially. This monitoring data requires high reliability; the volume of condition monitoring data is large, the sources are diverse, and there may be multiple descriptions of the same state characteristic. Moreover, the collected monitoring data often contains uncertainties such as noise and missing data, resulting in a scarcity of valuable, key characteristic data that effectively reflects the equipment status. Furthermore, the recording of a large amount of monitoring data inevitably leads to a significant increase in the difficulty of monitoring data processing, and valuable condition monitoring information remains untapped.

[0024] Faced with such complex, high-dimensional data, not all equipment condition monitoring data attributes are important or meaningful; many are redundant and repetitive. Not all monitoring data is valuable. Analyzing all condition monitoring data would significantly increase machine processing time and make accurate analysis and judgment of power equipment status more difficult. Therefore, extracting important status indicators from this data, deleting useless and repetitive data, and directly analyzing key status indicators is of great practical significance for the analysis and processing of equipment condition monitoring data and for providing useful auxiliary decision-making for the safe operation and production of the entire power system. Thus, in order to determine the importance category of each equipment condition monitoring data to achieve data compression, an intelligent power grid data compression method is desired.

[0025] Based on this, in the technical solution of this application, firstly, status monitoring data of each device is obtained, and status characterization feature values ​​are extracted from the status monitoring data of each device. Then, the status characterization feature values ​​of each device are aggregated to obtain a status feature vector. Next, each device status detection point is taken as a node, and the feature value x at each position of the status feature vector is... i The associated feature values ​​of the nodes are represented, and the importance index between the feature values ​​at each position of the state feature vector is calculated based on the following formula to obtain the importance index matrix, where the value at each position is... for ,in and These are the th elements of the state feature vector. and the The feature values ​​at each position, where... and These are hyperparameters, The value represents the tendency to favor nodes with high feature values, while This indicates a tendency to favor nodes with similar attribute scores.

[0026] After obtaining the importance index matrix, it is processed using a normalization function to obtain the importance index normalized matrix. This matrix is ​​then used to calculate the score representing the overall utility orientation probability between nodes. ,in This represents the summation of the eigenvalues ​​at all positions in the importance index matrix.

[0027] Then, for the importance index normalized matrix F used to represent the pointing information between nodes... ij The mean of each row is calculated to obtain the horizontal correlation feature vector of the equipment status, and the mean of each column is calculated to obtain the vertical correlation feature vector of the equipment status. Then, the weighted sum of the horizontal correlation feature vector and the vertical correlation feature vector of the equipment status is calculated to obtain the classification feature vector.

[0028] Finally, the classification feature vector is used to calculate the probability value of each location using a classification function, such as a Softmax classification function, which is the importance ratio value of the corresponding equipment status monitoring data.

[0029] It can be understood that by calculating the importance index of each device status detection point as a node, and then obtaining the importance index normalization matrix through logical recursion based on the normalization function in the row and column directions, this importance index normalization matrix essentially represents the directional information between nodes, which can be understood as containing the vectorized importance information between each node. Next, since each row and column of the importance index normalization matrix represents the utility directionality of all other nodes pointing to a certain node, by averaging them, we can represent the probability that all nodes point to the i-th node in the two-dimensional direction. That is, the value at each i position of the device status horizontal association feature vector and the device status vertical association feature vector represents the probability that all nodes point to the i-th node in the two-dimensional direction. Here, this value essentially indicates whether the importance between nodes is equal; that is, if the value is equal, the two nodes can be considered to have equal states, i.e., equal importance, and are therefore used to determine the importance category of each device status detection data.

[0030] Based on this, this application proposes an intelligent compression method for power grid data, which includes: obtaining status monitoring data of various devices in a smart grid; extracting status characterization feature values ​​from the status monitoring data of each device, and aggregating the status characterization feature values ​​of each device to obtain a status feature vector; constructing an importance index matrix based on the status feature vector; determining the importance ratio value of the status monitoring data of each device based on the importance index matrix; and determining the compression ratio of the status monitoring data of each device in the smart grid based on the probability value.

[0031] After introducing the basic principles of this application, various non-limiting embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0032] Exemplary methods Figure 1 The flowchart illustrates an intelligent compression method for power grid data. For example... Figure 1 As shown, the intelligent compression method for power grid data according to an embodiment of this application includes: S110, obtaining state monitoring data of various devices in a smart grid; S120, extracting state characterization feature values ​​from the state monitoring data of each device, and aggregating the state characterization feature values ​​of each device to obtain a state feature vector; S130, calculating the importance index between feature values ​​at each position in the state feature vector to obtain an importance index matrix composed of the importance index, wherein the importance index between feature values ​​is generated based on the difference of the squares of the differences between feature values; S140, using logical recursion based on a normalization function to calculate the fraction of the overall utility orientation probability between the nodes at each position in the importance index matrix to obtain an importance index normalization matrix composed of the fraction, wherein the importance index based on the normalization function is generated based on the squares of the differences between feature values; The normalization function is a natural exponential function that takes the score at position i and j in the importance index matrix and divides it by the sum of the scores in the i-th row and j-th column of the importance index matrix; S150, calculate the mean of each row and the mean of each column in the importance index normalization matrix to obtain the horizontal correlation feature vector and the vertical correlation feature vector of the device status; S160, calculate the weighted sum of the horizontal correlation feature vector and the vertical correlation feature vector of the device status to obtain the classification feature vector; S170, use a classifier to calculate the probability value of each position in the classification feature vector as the importance ratio value of the status monitoring data of the device corresponding to that position; and S180, based on the probability value, determine the compression ratio of the status monitoring data of each device in the smart grid.

[0033] Figure 2 The illustration shows a schematic diagram of the architecture of an intelligent power grid data compression method according to an embodiment of this application. Figure 2 As shown, in the network architecture of the intelligent power grid data compression method, firstly, the acquired status monitoring data of each device in the smart grid (e.g., such as...) is... Figure 2 The P1 shown is obtained through an encoder model containing one-dimensional convolutional layers and fully connected layers (e.g., as shown in the figure). Figure 2 E as shown in the diagram) to obtain a state feature vector containing information about the relationships between data (e.g., such as... Figure 2 The VFS shown in the diagram); then, the importance index between the feature values ​​at each position in the state feature vector is calculated (e.g., as shown in the diagram). Figure 2 (as shown in the US) to obtain an importance index matrix composed of the aforementioned importance indices (e.g., as shown in the US). Figure 2 The MS shown in the diagram); then, using logical recursion based on a normalization function, the scores for the overall utility pointing probability between the nodes at each position in the importance index matrix are calculated (e.g., as shown in the diagram). Figure 2 The FV (as shown) is used to obtain an importance index normalized matrix composed of the said scores (e.g., as shown). Figure 2 (as shown in the MF); then, the mean of each row and the mean of each column in the importance index normalization matrix are calculated to obtain the device state lateral correlation feature vector (e.g., as shown in the MF). Figure 2 The VF1 shown in the diagram and the device state longitudinal correlation feature vector (e.g., as shown in the diagram) Figure 2 (as shown in VF2); then, a weighted sum of the device state lateral correlation feature vector and the device state vertical correlation feature vector is calculated to obtain a classification feature vector (e.g., as shown in VF2). Figure 2 The VFC shown in the diagram); then, the classification feature vector is passed through a classifier (e.g., such as...). Figure 2 The circle S shown in the figure is used to calculate the probability value of each position in the classification feature vector as the importance ratio value of the status monitoring data of the device corresponding to that position; and finally, based on the probability value, the compression ratio of the status monitoring data of each device in the smart grid is determined.

[0034] In steps S110 and S120, status monitoring data of each device in the smart grid is obtained; status characterization feature values ​​are extracted from the status monitoring data of each device, and the status characterization feature values ​​of each device are aggregated to obtain a status feature vector. As mentioned above, considering the correlation between the status monitoring data of each device, if the correlation information between the status monitoring data of each device can be used to determine the importance information of each device status monitoring data to the global context, the importance category of each device status monitoring data can be determined based on the importance information, thereby achieving data compression based on the importance category. Therefore, in the technical solution of this application, status monitoring data of each device is first obtained from the smart grid, wherein the devices in the smart grid include, but are not limited to, generators, transformers, and electrical appliances. Then, the status monitoring data of each device in the smart grid is encoded through an encoder model containing one-dimensional convolutional layers and fully connected layers to obtain a status feature vector containing correlation information between the data.

[0035] Specifically, in this embodiment, the process of obtaining a state feature vector containing the correlation information between data from the state monitoring data of each device in the smart grid through an encoder model containing a one-dimensional convolutional layer and a fully connected layer includes: First, arranging the state monitoring data of each device in the smart grid into a state data vector to facilitate subsequent encoding processing. Then, using the fully connected layer of the encoder model to perform fully connected encoding on the state data vector to extract the high-dimensional latent features of the state monitoring data at each position in the state data vector. Accordingly, in a specific example, the fully connected layer performs fully connected encoding on the state data vector using the following formula; wherein, the formula is: ,in It is the input vector. It is the output vector. It is a weight matrix. It is a bias vector. This represents matrix multiplication. Finally, the state data vector is subjected to one-dimensional convolutional encoding using the one-dimensional convolutional layer of the encoder model to extract high-dimensional implicit correlation features between the state monitoring data at each position of the state data vector, thereby obtaining the state feature vector.

[0036] Figure 3 The illustration shows a flowchart of the process for extracting state characterization feature values ​​from the state monitoring data of each device in the intelligent power grid data compression method according to an embodiment of this application. For example... Figure 3As shown in the embodiment of this application, extracting state representation feature values ​​from the state monitoring data of each device includes: S210, arranging the state monitoring data of each device into a state input vector according to the sample dimension, and inputting the state input vector into an encoder model containing a one-dimensional convolutional layer and a fully connected layer to obtain a device state representation vector; and S220, calculating the global mean of the device state representation vector as the state representation feature value.

[0037] In step S130, the importance index between the eigenvalues ​​at each position in the state feature vector is calculated to obtain an importance index matrix composed of the importance indices. The importance index between eigenvalues ​​is generated based on the squared differences between eigenvalues. That is, in the technical solution of this application, each device state detection point is then taken as a node, and the eigenvalue x at each position of the state feature vector... i The importance index matrix is ​​obtained by further calculating the importance index between the feature values ​​at each position of the state feature vector, representing the associated feature values ​​of the node. In a specific example, the importance index between feature values ​​is generated based on the difference between the squares of the differences between feature values.

[0038] Specifically, in this embodiment of the application, the process of calculating the importance index between the feature values ​​at each position in the state feature vector to obtain an importance index matrix composed of the importance indices includes: calculating the importance index between the feature values ​​at each position in the state feature vector using the following formula to obtain the importance index matrix; wherein, the formula is: ,in and These are the th elements of the state feature vector. and the The feature values ​​at each position, where... and These are hyperparameters, The value represents the tendency to favor nodes with high feature values, while This indicates a tendency to favor nodes with similar attribute scores.

[0039] In step S140, a normalization-based logical recursion is used to calculate the scores representing the overall utility orientation probability between nodes at each position in the importance index matrix, thereby obtaining an importance index normalized matrix composed of these scores. The normalization-based logical recursion represents the natural exponential function, which takes the score at position i and j in the importance index matrix divided by the sum of the scores in the i-th row and j-th column of the importance index matrix. That is, after obtaining the importance index matrix, a normalization-based logical recursion is used to calculate the scores representing the overall utility orientation probability between nodes. In a specific example, this is F... ij =exp(E ij / ∑ i,j E ij ), where ∑ i,j E ij Represents E for the i-th row and j-th column ij The summation, or summation in both row and column directions, is a logical recursion based on a normalization function. It can be understood that by calculating the importance index of each device status detection point as a node, and then obtaining the importance index normalization matrix through logical recursion in the row and column directions based on a normalization function, this importance index normalization matrix essentially represents the directional information between the nodes, and can be understood as containing the vectorized importance information between the nodes.

[0040] Accordingly, in a specific example of this application, the process of using logical recursion based on a normalization function to calculate the score of the overall utility orientation probability between the nodes at each position in the importance index matrix to obtain the importance index normalized matrix composed of the scores includes: using logical recursion based on a normalization function to calculate the score of the overall utility orientation probability between the nodes at each position in the importance index matrix according to the following formula to obtain the importance index normalized matrix; wherein, the formula is: ,in This represents the summation of the eigenvalues ​​at all positions in the importance index matrix.

[0041] In steps S150 and S160, the mean of each row and the mean of each column in the importance index normalization matrix are calculated to obtain the horizontal correlation feature vector and the vertical correlation feature vector of the device state, respectively. A weighted sum of the horizontal correlation feature vector and the vertical correlation feature vector of the device state is then calculated to obtain the classification feature vector. It should be understood that since each row and each column of the importance index normalization matrix represents the utility orientation of all other nodes pointing to a certain node, in the technical solution of this application, further, for the importance index normalization matrix F used to represent the orientation information between the nodes... ij By averaging each row, the probability that all nodes point to the i-th node in the two-dimensional direction can be represented. That is, the value at each i-th position of the device state horizontal association feature vector and the device state vertical association feature vector represents the probability that all nodes point to the i-th node in the two-dimensional direction. It should be understood that this value essentially represents whether the importance of the nodes is equal. That is, if the value is equal, the two nodes can be considered to have equal states, i.e., equal importance, and therefore can be used to determine the importance category of each device state detection data. Therefore, the weighted sum of the device state horizontal association feature vector and the device state vertical association feature vector is then used to obtain a classification feature vector for classification.

[0042] In steps S170 and S180, the classification feature vector is processed by a classifier to calculate the probability value of each position in the classification feature vector, which is used as the importance ratio value of the status monitoring data of the corresponding device. Based on the probability value, the compression ratio of the status monitoring data of each device in the smart grid is determined. That is, in the technical solution of this application, after obtaining the classification feature vector, the classification feature vector is input into the class-based Softmax classification function of the classifier to calculate the class-based Softmax function value of each position in the classification feature vector, which is used as the probability value of each position in the classification feature vector, i.e., the importance ratio value of the corresponding device status monitoring data. In a specific example, the class-based Softmax classification function is the weighted sum of the natural exponential function value (where the negative values ​​of the feature values ​​at each position in the classification feature vector are powers of the natural exponential function value). Then, further, based on the probability value, the compression ratio of the status monitoring data of each device in the smart grid is determined to ensure the safe operation and production of the entire power system.

[0043] In summary, the intelligent power grid data compression method of this application is explained. It calculates the importance index of each device status detection point as a node, and obtains an importance index normalization matrix through logical recursion based on a normalization function in the row and column directions of the importance index. This extracts the importance vectorization information between the nodes. Furthermore, by averaging these values, the probability that all nodes point to the i-th node in a two-dimensional direction can be represented; that is, it essentially represents whether the importance of each node is equal. This can be used to determine the importance category of each device status detection data point, and then achieve data compression based on the importance category to ensure the safe operation and production of the entire power system.

[0044] In another embodiment of this application, a method for intelligent compression of power grid data is also provided, comprising: obtaining status monitoring data of various devices in a smart grid; extracting status characterization feature values ​​from the status monitoring data of the various devices, and aggregating the status characterization feature values ​​of the various devices to obtain a status feature vector; constructing an importance index matrix based on the status feature vector; determining the importance ratio value of the status monitoring data of the various devices based on the importance index matrix; and determining the compression ratio of the status monitoring data of the various devices in the smart grid based on the probability value.

[0045] In the intelligent compression method for power grid data according to this application, extracting state representation feature values ​​from the state monitoring data of each device includes: arranging the state monitoring data of each device into a state input vector according to the sample dimension, inputting the state input vector into an encoder model containing a one-dimensional convolutional layer and a fully connected layer to obtain a device state representation vector; and calculating the global mean of the device state representation vector as the state representation feature value.

[0046] In the intelligent compression method for power grid data according to this application, constructing an importance index matrix based on the state feature vector includes: calculating the importance index between feature values ​​at any two positions in the state feature vector to obtain multiple importance indices; and arranging the multiple importance indices in a matrix to obtain the importance index matrix.

[0047] In the intelligent compression method for power grid data according to this application, the importance index between feature values ​​at any two positions in the state feature vector is calculated to obtain multiple importance indices, including: using the following importance index calculation formula to calculate the importance index between feature values ​​at any two positions in the state feature vector; The importance index is calculated using the following formula: ,in and These are the th elements of the state feature vector. and the The feature values ​​at each position, where... and These are hyperparameters, The value represents the tendency to favor nodes with high feature values, while This indicates a tendency to favor nodes with similar attribute scores.

[0048] In the intelligent compression method for power grid data according to this application, determining the importance ratio value of the status monitoring data of each device based on the importance index matrix includes: processing the importance index matrix using a normalization function to obtain an importance index normalization matrix; calculating the mean of each row and the mean of each column in the importance index normalization matrix to obtain a horizontal correlation feature vector and a vertical correlation feature vector of device status; calculating the weighted sum of the horizontal correlation feature vector and the vertical correlation feature vector of device status to obtain a classification feature vector; inputting the classification feature vector into a classifier to calculate the probability value of each position in the classification feature vector as the importance ratio value of the status monitoring data of the device corresponding to that position; and determining the importance ratio value as the compression ratio of the status monitoring data of each device in the smart grid.

[0049] In the intelligent compression method for power grid data according to this application, inputting the classification feature vector into a classifier to calculate the probability value of each position in the classification feature vector as the importance ratio value of the status monitoring data of the device corresponding to that position includes: inputting the classification feature vector into the class-Softmax classification function of the classifier to calculate the class-Softmax function value of each position in the classification feature vector as the probability value of each position in the classification feature vector.

[0050] Exemplary System Figure 4 The diagram illustrates a block diagram of an intelligent power grid data compression system according to an embodiment of this application. Figure 4As shown, the intelligent power grid data compression system 500 according to an embodiment of this application includes: a data acquisition unit 510, used to acquire status monitoring data of various devices in the smart grid; an encoding unit 520, used to extract status characterization feature values ​​from the status monitoring data of the various devices, and aggregate the status characterization feature values ​​of the various devices to obtain a status feature vector; an importance index measurement unit 530, used to construct an importance index matrix based on the status feature vector; an importance determination unit 540, used to determine the importance ratio value of the status monitoring data of the various devices based on the importance index matrix; and a compression ratio determination unit 550, used to determine the compression ratio of the status monitoring data of the various devices in the smart grid based on the probability value.

[0051] Here, those skilled in the art will understand that the specific functions and operations of each unit and module in the aforementioned intelligent power grid data compression system 500 have been referenced above. Figures 1 to 3 The method for intelligent compression of power grid data has been described in detail, and therefore, its repeated description will be omitted.

[0052] As described above, the intelligent power grid data compression system 500 according to the embodiments of this application can be implemented in various terminal devices, such as servers for intelligent power grid data compression algorithms. In one example, the intelligent power grid data compression system 500 according to the embodiments of this application can be integrated into a terminal device as a software module and / or a hardware module. For example, the intelligent power grid data compression system 500 can be a software module in the operating system of the terminal device, or it can be an application developed for the terminal device; of course, the intelligent power grid data compression system 500 can also be one of many hardware modules of the terminal device.

[0053] Alternatively, in another example, the intelligent power grid data compression system 500 and the terminal device can also be separate devices, and the intelligent power grid data compression system 500 can be connected to the terminal device via wired and / or wireless networks, and transmit interactive information in accordance with the agreed data format.

[0054] Exemplary electronic devices Below, for reference Figure 5 To describe an electronic device according to embodiments of this application. For example... Figure 5 As shown, the electronic device 10 includes one or more processors 11 and memory 12. The processor 11 may be a central processing unit (CPU) or other form of processing unit with data processing capabilities and / or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.

[0055] The memory 12 may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor 11 may execute the program instructions to implement the functions of the intelligent power grid data compression method of the various embodiments of this application described above, and / or other desired functions. Various contents such as importance index matrices and classification feature vectors may also be stored in the computer-readable storage medium.

[0056] In one example, the electronic device 10 may also include an input system 13 and an output system 14, which are interconnected via a bus system and / or other forms of connection mechanism (not shown).

[0057] The input system 13 may include, for example, a keyboard, a mouse, etc.

[0058] The output system 14 can output various information to the outside, including importance ratio values, compression ratios, etc. The output system 14 may include, for example, a display, a speaker, a printer, and a communication network and its connected remote output devices, etc.

[0059] Of course, for the sake of simplicity, Figure 5 Only some of the components of the electronic device 10 relevant to this application are shown in this illustration; components such as buses, input / output interfaces, etc., are omitted. In addition, the electronic device 10 may include any other suitable components depending on the specific application.

[0060] Exemplary computer program products and computer-readable storage media In addition to the methods and devices described above, embodiments of this application may also be computer program products, which include computer program instructions that, when executed by a processor, cause the processor to perform the steps of the intelligent power grid data compression method according to various embodiments of this application as described in the "Exemplary Methods" section above.

[0061] The computer program product can be written in any combination of one or more programming languages ​​to perform the operations of the embodiments of this application. The programming languages ​​include object-oriented programming languages ​​such as Java and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computing device, partially on the user's computing device, as a standalone software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0062] Furthermore, embodiments of this application may also be computer-readable storage media storing computer program instructions thereon, which, when executed by a processor, cause the processor to perform the steps of the intelligent power grid data compression method described in the "Exemplary Methods" section of this specification.

[0063] The computer-readable storage medium may be any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or any combination thereof. More specific examples (a non-exhaustive list) of readable storage media include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.

[0064] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0065] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0066] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0067] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0068] The above description has been given for purposes of illustration and description. Furthermore, this description is not intended to limit the embodiments of this application to the forms disclosed herein. Although numerous exemplary aspects and embodiments have been discussed above, those skilled in the art will recognize certain variations, modifications, alterations, additions, and sub-combinations thereof.

Claims

1. A method for intelligent compression of power grid data, characterized in that, include: Obtain status monitoring data of various devices in the smart grid; State characterization feature values ​​are extracted from the state monitoring data of each device, and the state characterization feature values ​​of each device are aggregated to obtain a state feature vector; Based on the state feature vector, an importance index matrix is ​​constructed; Based on the importance index matrix, the importance ratio value of the status monitoring data of each device is determined; as well as Based on the probability value, the compression ratio of the status monitoring data of each device in the smart grid is determined.

2. The intelligent power grid data compression method according to claim 1, characterized in that, State characterization feature values ​​are extracted from the state monitoring data of each device, including: After arranging the state monitoring data of each device into a state input vector according to the sample dimension, the state input vector is input into an encoder model containing a one-dimensional convolutional layer and a fully connected layer to obtain a device state representation vector; and The global mean of the device state representation vector is calculated as the state representation feature value.

3. The intelligent power grid data compression method according to claim 2, characterized in that, Based on the state feature vector, an importance index matrix is ​​constructed, including: Calculate the importance index between any two feature values ​​in the state feature vector to obtain multiple importance indices; and The multiple importance indices are arranged in a matrix to obtain the importance index matrix.

4. The intelligent power grid data compression method according to claim 3, characterized in that, Calculate the importance index between feature values ​​at any two positions in the state feature vector to obtain multiple importance indices, including: using the following importance index calculation formula to calculate the importance index between feature values ​​at any two positions in the state feature vector; The importance index is calculated using the following formula: ,in and These are the th elements of the state feature vector. and the The feature values ​​at each position, where... and These are hyperparameters, The value represents the tendency to favor nodes with high feature values, while This indicates a tendency to favor nodes with similar attribute scores.

5. The intelligent power grid data compression method according to claim 4, characterized in that, Based on the importance index matrix, the importance ratio value of the status monitoring data of each device is determined, including: The importance index matrix is ​​processed using a normalization function to obtain the importance index normalized matrix; The mean of each row and the mean of each column in the importance index normalization matrix are calculated respectively to obtain the horizontal correlation feature vector and the vertical correlation feature vector of the equipment status. The weighted sum of the horizontal correlation feature vector and the vertical correlation feature vector of the device status is calculated to obtain the classification feature vector; The classification feature vector is input into a classifier to calculate the probability value of each position in the classification feature vector, which is then used as the importance ratio value of the device's status monitoring data corresponding to that position; and The importance ratio is determined as the compression ratio of the status monitoring data of each device in the smart grid.

6. The intelligent power grid data compression method according to claim 5, wherein, Inputting the classification feature vector into a classifier to calculate the probability value of each position in the classification feature vector as the importance ratio value of the status monitoring data of the device corresponding to that position includes: inputting the classification feature vector into the class-softmax classification function of the classifier to calculate the class-softmax function value of each position in the classification feature vector as the probability value of each position in the classification feature vector.

7. A smart power grid data compression system, characterized in that, include: The data acquisition unit is used to obtain status monitoring data of various devices in the smart grid; The encoding unit is used to extract state characterization feature values ​​from the state monitoring data of each device, and aggregate the state characterization feature values ​​of each device to obtain a state feature vector; Importance index measurement unit, used to construct an importance index matrix based on the state feature vector; An importance determination unit is used to determine the importance ratio value of the status monitoring data of each device based on the importance index matrix. as well as The compression ratio determination unit is used to determine the compression ratio of the status monitoring data of each device in the smart grid based on the probability value.

8. The intelligent power grid data compression system according to claim 7, characterized in that, The encoding unit is further configured to: arrange the state monitoring data of each device into a state input vector according to the sample dimension, input the state input vector into an encoder model containing a one-dimensional convolutional layer and a fully connected layer to obtain a device state representation vector; and calculate the global mean of the device state representation vector as the state representation feature value.

9. The intelligent power grid data compression system according to claim 7, characterized in that, The importance index measurement unit is further used to: calculate the importance index between feature values ​​at any two positions in the state feature vector to obtain multiple importance indices; Furthermore, the plurality of importance indices are arranged in a matrix to obtain the importance index matrix.

10. An electronic device, comprising: processor; as well as A memory storing computer program instructions, which, when executed by the processor, cause the processor to perform the intelligent power grid data compression method as described in any one of claims 1-6.

Images