Electric power marketing operation digital twinborn model maintenance method based on consistency constraint

Abstract

The invention relates to a power marketing job digital twin model maintenance method based on consistency constraint. The method comprises the following steps: acquiring a digital twin model corresponding to an electric power marketing job and monitoring data acquired by at least two data sources for the electric power marketing job; based on the maintenance requirement of the digital twin model, determining a consistency constraint dimension matched with the power marketing operation; performing consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimension to obtain consistency constraint data matched with the digital twin model; and maintaining the digital twin model by using the consistency constraint data to obtain an updated model corresponding to the power marketing operation. By adopting the method, the prediction precision of the digital twinborn model corresponding to the power marketing operation can be improved.

Description

Technical Field

[0001] The present application relates to the technical field of electric power, and in particular to a method for maintaining a digital twin model of electric power marketing operations based on consistency constraints. Background Art

[0002] With the gradual liberalization of the power market, power marketing operations are facing unprecedented challenges. Power marketing not only needs to ensure the balance between power supply and demand, but also needs to achieve the optimal allocation of power resources in a market-oriented environment. Therefore, the traditional power marketing method can no longer meet the requirements of refined management. Especially when dealing with rapidly changing power loads, dispatching, and market demands, the traditional method is difficult to make dynamic adjustments in real time and accurately.

[0003] In this context, the digital twin technology has emerged. By constructing a digital twin model based on the digital twin technology, the operating state of the real world can be mapped based on the digital twin model. Thus, various dynamic changes in power marketing operations can be effectively simulated, providing strong support for optimizing decisions, reducing costs, and improving efficiency.

[0004] However, the current digital twin model has the problem of low model prediction accuracy. Summary of the Invention

[0005] Based on this, in view of the above technical problems, it is necessary to provide a method for maintaining a digital twin model of electric power marketing operations based on consistency constraints, which can improve the prediction accuracy of the digital twin model corresponding to electric power marketing operations.

[0006] In a first aspect, the present application provides a method for maintaining a digital twin model of electric power marketing operations based on consistency constraints, including: obtaining the digital twin model corresponding to the electric power marketing operations and the monitoring data collected by at least two data sources for the electric power marketing operations respectively; determining the consistency constraint dimensions matching the electric power marketing operations based on the maintenance requirements of the digital twin model; performing consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimensions to obtain consistency constraint data matching the digital twin model; and using the consistency constraint data to maintain the digital twin model to obtain an updated model corresponding to the electric power marketing operations.

[0007] In one embodiment, the consistency constraint dimension includes at least one sub-constraint dimension; based on the consistency constraint dimension, consistency analysis is performed on the monitoring data respectively corresponding to each data source to obtain consistency constraint data matching the digital twin model, including: for each sub-constraint dimension, based on the sub-constraint dimension, consistency analysis is performed on the monitoring data respectively corresponding to each data source to obtain dimension analysis data corresponding to the sub-constraint dimension; based on the dimension analysis data corresponding to the sub-constraint dimension, the weight coefficient corresponding to the sub-constraint dimension is determined; the weight coefficients respectively corresponding to each sub-constraint dimension and the dimension analysis data respectively corresponding to each sub-constraint dimension are subjected to weighted summation processing to obtain consistency constraint data matching the power marketing operation.

[0008] In one embodiment, based on the sub-constraint dimension, consistency analysis is performed on the monitoring data respectively corresponding to each data source to obtain dimension analysis data corresponding to the sub-constraint dimension, including: selecting a target data source from each data source and determining each remaining data source except the target data source in each data source; based on the difference characteristics between the monitoring data corresponding to the target data source and each remaining data source, obtaining target analysis data corresponding to the target data source under the sub-constraint dimension; taking the next data source adjacent to the target data source in each data source as the new target data source, and returning to the step of determining each remaining data source except the target data source in each data source; according to the order of the target data sources, determining the data set including the target analysis data respectively corresponding to each target data source under the sub-constraint dimension as the dimension analysis data corresponding to the sub-constraint dimension.

[0009] In one embodiment, the sub-constraint dimension includes a data consistency dimension; the method further includes: obtaining the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, and the covariance is used to characterize the difference characteristics between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources; based on the difference characteristics between the monitoring data corresponding to the target data source and each remaining data source, obtaining target analysis data corresponding to the target data source under the sub-constraint dimension, including: obtaining the first standard deviation matching the monitoring data corresponding to the target data source and the second standard deviation matching the monitoring data corresponding to the remaining data sources; based on each covariance, the first standard deviation and each second standard deviation corresponding to the target data source, obtaining target analysis data corresponding to the target data source under the sub-constraint dimension.

[0010] In one embodiment, the sub-constraint dimension includes a time consistency dimension, and the monitoring data includes a plurality of sub-data; the method further includes: dividing the detection time window of the target data source to obtain a plurality of sub-detection times; for each sub-detection time, obtaining the data difference between the sub-data of the remaining data sources at the sub-detection time and the sub-data of the target data source at the sub-detection time; the data difference is used to characterize the difference feature between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources; based on the difference features between the monitoring data corresponding to the target data source and each of the remaining data sources respectively, obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension, including: based on the number of sub-detection times corresponding to the target data source and the respective data differences corresponding to each sub-detection time, obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension.

[0011] In one embodiment, the sub-constraint dimension includes a spatial consistency dimension; the method further includes: obtaining the data distance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, the data distance is used to characterize the difference feature between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources; based on the difference features between the monitoring data corresponding to the target data source and each of the remaining data sources respectively, obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension, including: based on the respective data distances corresponding to the target data source and the number of data sources of at least two data sources, obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension.

[0012] In one embodiment, the dimension analysis data includes the target analysis data corresponding to each target data source respectively; based on the dimension analysis data corresponding to the sub-constraint dimension, determining the weight coefficient corresponding to the sub-constraint dimension, including: performing normalization processing on the respective target analysis data corresponding to the sub-constraint dimension to obtain the target normalization data corresponding to each target analysis data respectively; for each target analysis data, obtaining the entropy value corresponding to the target analysis data based on the target normalization data corresponding to the target analysis data and the number of data sources; based on the maximum entropy value among the entropy values corresponding to the respective target analysis data and the entropy values corresponding to the respective target analysis data, obtaining the sub-coefficients corresponding to the respective target analysis data under the sub-constraint dimension; based on the sub-coefficients corresponding to the respective target analysis data under the sub-constraint dimension, obtaining the weight coefficient corresponding to the sub-constraint dimension.

[0013] Second aspect, the present application provides a device for maintaining a digital twin model for power marketing operations based on consistency constraints. The device includes: an acquisition module, configured to acquire the digital twin model corresponding to the power marketing operation and the monitoring data respectively collected by at least two data sources for the power marketing operation; a determination module, configured to determine the consistency constraint dimension matched with the power marketing operation based on the maintenance requirements of the digital twin model; a processing module, configured to perform consistency analysis on the monitoring data respectively corresponding to each data source based on the consistency constraint dimension to obtain the consistency constraint data matched with the digital twin model; a maintenance module, configured to maintain the digital twin model using the consistency constraint data to obtain an updated model corresponding to the power marketing operation.

[0014] Third aspect, the present application further provides a computer device, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the following steps are implemented: acquiring the digital twin model corresponding to the power marketing operation and the monitoring data respectively collected by at least two data sources for the power marketing operation; determining the consistency constraint dimension matched with the power marketing operation based on the maintenance requirements of the digital twin model; performing consistency analysis on the monitoring data respectively corresponding to each data source based on the consistency constraint dimension to obtain the consistency constraint data matched with the digital twin model; maintaining the digital twin model using the consistency constraint data to obtain an updated model corresponding to the power marketing operation.

[0015] Fourth aspect, the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the following steps are implemented: acquiring the digital twin model corresponding to the power marketing operation and the monitoring data respectively collected by at least two data sources for the power marketing operation; determining the consistency constraint dimension matched with the power marketing operation based on the maintenance requirements of the digital twin model; performing consistency analysis on the monitoring data respectively corresponding to each data source based on the consistency constraint dimension to obtain the consistency constraint data matched with the digital twin model; maintaining the digital twin model using the consistency constraint data to obtain an updated model corresponding to the power marketing operation.

[0016] Fifth aspect, the present application further provides a computer program product, including a computer program. When the computer program is executed by a processor, the following steps are implemented: acquiring the digital twin model corresponding to the power marketing operation and the monitoring data respectively collected by at least two data sources for the power marketing operation; determining the consistency constraint dimension matched with the power marketing operation based on the maintenance requirements of the digital twin model; performing consistency analysis on the monitoring data respectively corresponding to each data source based on the consistency constraint dimension to obtain the consistency constraint data matched with the digital twin model; maintaining the digital twin model using the consistency constraint data to obtain an updated model corresponding to the power marketing operation.

[0017] The above-mentioned method for maintaining the digital twin model of power marketing operations based on consistency constraints obtains the digital twin model corresponding to the power marketing operations and the monitoring data collected by at least two data sources for the power marketing operations respectively. Based on the maintenance requirements of the digital twin model, it determines the consistency constraint dimensions matching the power marketing operations, and performs consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimensions to obtain the consistency constraint data matching the digital twin model. Furthermore, it uses the consistency constraint data to maintain the digital twin model to obtain the updated model corresponding to the power marketing operations. Thus, by adjusting the parameters of the digital twin model corresponding to the power marketing operations through the data feedback corrected based on the consistency constraints, the fidelity between the model and the actual power system state can be ensured, enabling the digital twin model to adapt to the dynamic changes of the power system and improving the prediction accuracy of the digital twin model in power marketing operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] To more clearly illustrate the technical solutions in the embodiments of the present application or related technologies, the following will briefly introduce the drawings required for use in the description of the embodiments of the present application or related technologies. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other related drawings can also be obtained based on these drawings.

[0019] Figure 1 It is a schematic flowchart of the method for maintaining the digital twin model of power marketing operations based on consistency constraints in one embodiment;

[0020] Figure 2 It is a schematic flowchart of performing consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimensions in one embodiment to obtain the consistency constraint data matching the digital twin model;

[0021] Figure 3 It is a schematic flowchart of performing consistency analysis on the monitoring data corresponding to each data source based on the sub-constraint dimensions in one embodiment to obtain the dimension analysis data corresponding to the sub-constraint dimensions;

[0022] Figure 4 It is a schematic flowchart of determining the weight coefficient corresponding to the sub-constraint dimensions based on the dimension analysis data corresponding to the sub-constraint dimensions in one embodiment;

[0023] Figure 5 It is a schematic flowchart of the method for maintaining the digital twin model of power marketing operations based on consistency constraints in another embodiment;

[0024] Figure 6The structural block diagram of the digital twin model maintenance device for power marketing operations based on consistency constraints in an embodiment;

[0025] Figure 7 The internal structure diagram of a computer device in an embodiment. Detailed implementation manners

[0026] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the present application will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0027] The embodiments of the technical solutions of the present application will be described in detail below with reference to the accompanying drawings. The following embodiments are only used to illustrate the technical solutions of the present application more clearly, so they are only examples and cannot be used to limit the protection scope of the present application.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above drawings are intended to cover non-exclusive inclusion.

[0029] In the description of the embodiments of the present application, technical terms such as "first" and "second" are only used to distinguish different objects and cannot be understood as indicating or implying relative importance or implicitly indicating the quantity, specific order or primary-secondary relationship of the indicated technical features. In the description of the embodiments of the present application, "a plurality" means more than two unless otherwise specifically defined.

[0030] Referring to "embodiments" herein means that the specific features, structures or characteristics described in connection with the embodiments can be included in at least one embodiment of the present application. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0031] Existing digital twin models generally perform modeling and prediction based on historical data and fixed algorithms. With the dynamic changes in the power market, existing models often cannot adjust parameters in a timely manner, resulting in a gradual increase in prediction errors. This makes the existing technology lack sufficient flexibility and adaptability when dealing with power load fluctuations and market demand changes.

[0032] In addition, most of the data used in the training of existing digital twin models comes from data from different sources. These data may be inconsistent due to the data collection method, data format, or transmission process, resulting in low accuracy of the model. Since no effective solution is provided for this data inconsistency problem, the model cannot reflect the state of the real world in real time and accurately, affecting the decision-making efficiency of power marketing operations.

[0033] Due to the continuous correction of the accuracy of the digital twin model at present, the accuracy of the model decreases during long-term use. Although the accuracy of the model can be improved through incremental learning, the consistency constraint problem between the data used in model training is not considered, resulting in low stability and reliability of the model in complex environments. Moreover, most of the existing correction methods require complex calculations on a large amount of data, which leads to a long response time of the system and cannot meet the requirements of real-time and high efficiency in power marketing operations. When facing a large-scale power system, the slow update speed and high consumption of computing resources often result in the inability to complete scheduling and prediction tasks in real time.

[0034] In view of this, the present application provides a method for maintaining a digital twin model for power marketing operations based on consistency constraints, and this method can be applied to a server in a maintenance system. Specifically, the server is used to obtain the digital twin model corresponding to the power marketing operation and the monitoring data collected by at least two data sources for the power marketing operation respectively, and based on the maintenance requirements of the digital twin model, determine the consistency constraint dimension matching the power marketing operation, and perform consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimension to obtain the consistency constraint data matching the digital twin model, and then use the consistency constraint data to maintain the digital twin model to obtain the updated model corresponding to the power marketing operation.

[0035] Among them, the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server providing cloud computing services.

[0036] In one embodiment, as Figure 1 shown, a method for maintaining a digital twin model for power marketing operations based on consistency constraints is provided. Taking this method applied to a server in a maintenance system as an example, the method includes the following steps:

[0037] S102, obtain the digital twin model corresponding to the power marketing operation and the monitoring data collected by at least two data sources for the power marketing operation respectively.

[0038] Among them, the digital twin model corresponding to the power marketing operation is used to simulate the power marketing operation. For different power marketing operations, the data content for training the digital twin model corresponding to the power marketing operation is different. For example, if the power marketing operation represents load forecasting, the corresponding data may include historical power load data and other data; if the power marketing operation represents electricity price analysis, the corresponding data may include power market transaction data, user electricity consumption data and other data.

[0039] Among them, the data source refers to the device used to collect the data required for the power marketing operation. The data source includes but is not limited to: sensors, metering devices, telemetry devices, etc. For example, the data required for the power marketing operation can be collected from the power system.

[0040] In some embodiments, the monitoring data may include power load data. Power load data refers to the power demand data of each load point in the power system, including but not limited to: instantaneous load power, peak and valley electricity consumption data, historical load curves, and regional load data, etc.

[0041] In some embodiments, the monitoring data may include power grid status data. The power grid status data is used to reflect the operating status of the power network of the power system, including voltage, frequency, power factor, network topology, etc.

[0042] In one embodiment, the monitoring data may include the device operating status data of the power equipment in the power system. The device operating status data is used to reflect the health status of the power equipment, including but not limited to: device temperature, device vibration data, operating duration, fault alarm information, and maintenance record data, etc.

[0043] S104. Determine the consistency constraint dimension that matches the power marketing operation based on the maintenance requirements of the digital twin model.

[0044] In this embodiment, the maintenance requirements are used to represent the model optimization requirements of the digital twin model. The consistency constraint dimension refers to the dimension considered for model optimization of the digital twin model. The consistency constraint dimension includes at least one sub-constraint dimension, and the sub-constraint dimension includes but is not limited to: format consistency dimension, data consistency dimension, time consistency dimension, and space consistency dimension, etc.

[0045] Among them, the format consistency dimension is used to represent the analysis of the monitoring data from the data format dimension, the data consistency dimension is used to represent the analysis of the monitoring data from the data content dimension, the time consistency dimension is used to represent the analysis of the monitoring data from the data collection time dimension, and the space consistency dimension is used to represent the analysis of the monitoring data from the data space dimension.

[0046] In this embodiment, based on the maintenance requirements of the digital twin model, the method for determining the consistency constraint dimensions matching the power marketing operation is not limited to determine the maintenance requirements of the digital twin model.

[0047] In one implementation, when the usage duration of the digital twin model reaches a preset duration, it is determined that the digital twin model meets the maintenance requirements; based on the mapping relationship between the usage duration and the sub-constraint dimensions, as well as the usage duration of the digital twin model, the consistency constraint dimensions matching the power marketing operation are determined. Among them, the longer the usage duration, the more the number of sub-constraint dimensions.

[0048] In one embodiment, when the error between the input data of the digital twin model and the predicted data of the digital twin model for the input data is greater than a preset threshold, it is determined that the digital twin model meets the maintenance requirements; the preset sub-constraint dimensions configured for the maintenance requirements are determined as the consistency constraint dimensions matching the power marketing operation. For example, the preset sub-constraint dimensions may include data consistency dimension, spatial consistency dimension, and time consistency dimension, and there may be other settings.

[0049] S106. Based on the consistency constraint dimensions, perform consistency analysis on the monitoring data respectively corresponding to each data source to obtain the consistency constraint data matching the digital twin model.

[0050] Among them, the consistency constraint dimensions are used to ensure that the data collected from different data sources can meet certain consistency requirements when input into the digital twin model, prevent model prediction errors caused by inconsistencies between data sources, and can improve the data collaboration between multiple data sources.

[0051] The consistency constraint dimensions may include at least one sub-constraint dimension, and the method for performing consistency analysis on the monitoring data respectively corresponding to each data source based on the consistency constraint dimensions to obtain the consistency constraint data matching the digital twin model is not limited.

[0052] In one implementation, for each sub-constraint dimension, based on the mapping relationship between the sub-constraint dimension and the preset analysis strategy, the preset analysis strategy matching the sub-constraint dimension corresponding to the power marketing operation is determined as the target strategy of the power marketing operation under the sub-constraint dimension; use the target strategy corresponding to the sub-constraint dimension to perform consistency analysis on the monitoring data respectively corresponding to each data source to obtain the dimension analysis results corresponding to the sub-constraint dimension; based on the dimension analysis results respectively corresponding to each sub-constraint dimension, obtain the consistency constraint data matching the digital twin model.

[0053] In some embodiments, the monitoring data corresponding to each data source can be preprocessed to obtain the preprocessed data corresponding to each data source; for each data source, the preprocessed data corresponding to the data source is used as the monitoring data corresponding to the new data source, and consistency analysis is performed on the monitoring data corresponding to each data source based on the consistency constraint dimension to obtain the consistency constraint data matching the digital twin model. That is to say, after the monitoring data corresponding to each data source is preprocessed, the consistency analysis process is performed based on the consistency constraint dimension.

[0054] In some cases, data preprocessing includes noise suppression and data normalization, etc. In some embodiments, a Kalman filter can be used to suppress noise in the monitoring data to filter out the interference of external environmental factors on the acquisition results. In some embodiments, Z-score normalization can be used to normalize the monitoring data to eliminate the dimension difference.

[0055] S108, use the consistency constraint data to maintain the digital twin model to obtain an updated model corresponding to the power marketing operation.

[0056] Among them, the updated model refers to the model after maintaining (i.e., updating) the digital twin model. The function of the updated model is not limited. For example, the updated model can be used for load forecasting and electricity price analysis, etc.

[0057] In some embodiments, an incremental learning method is called to update the parameters of the digital twin model using the consistency constraint data to obtain an updated model corresponding to the power marketing operation.

[0058] For example, the parameters of the digital twin model can be represented as a mapping function , represents each parameter in the digital twin model, X is the consistency constraint data, is the result output by the digital twin model for X, is the true output corresponding to X. The parameters of the digital twin model are dynamically updated through the incremental learning method to gradually adapt to new data.

[0059] Specifically, by adjusting the parameter to minimize the error between the model output and the true output , , and the specific update formula is as follows:

[0060]

[0061] Among them, is the learning rate to control the update step size. For the loss function, the mean squared error can be selected as the loss function, or it can be set to other types of loss functions. For the loss function For the adjustment parameters Gradient.

[0062] For example, taking the mean squared error as the loss function as an example, it satisfies:

[0063]

[0064] Through the loss function, the gradient vector composed of the partial derivatives of all parameters can be obtained, and the gradient vector satisfies:

[0065]

[0066] Thus, all parameters of the digital twin model can be updated synchronously, satisfying:

[0067]

[0068] Based on Figure 1 As shown in the content, by obtaining the digital twin model corresponding to the power marketing operation and the monitoring data collected by at least two data sources for the power marketing operation respectively, and based on the maintenance requirements of the digital twin model, determining the consistency constraint dimensions matching the power marketing operation, and performing consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimensions to obtain the consistency constraint data matching the digital twin model, and then using the consistency constraint data to maintain the digital twin model to obtain the updated model corresponding to the power marketing operation. Thus, by adjusting the parameters of the digital twin model corresponding to the power marketing operation based on the data feedback corrected by the consistency constraint, the fidelity of the model to the actual power system state can be ensured, enabling the digital twin model to adapt to the dynamic changes of the power system and improving the prediction accuracy of the digital twin model in the power marketing operation.

[0069] In one embodiment, the consistency constraint dimension includes at least one sub-constraint dimension. Exemplarily, the implementation manner of performing consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimension to obtain the consistency constraint data matching the digital twin model (i.e., S106) can be as Figure 2 shown, including the following steps:

[0070] S202, for each sub-constraint dimension, perform consistency analysis on the monitoring data corresponding to each data source based on the sub-constraint dimension to obtain the dimension analysis data corresponding to the sub-constraint dimension.

[0071] Exemplarily, based on the mapping relationship between the sub-constraint dimension and the preset analysis strategy, the preset analysis strategy matching the sub-constraint dimension corresponding to the power marketing operation can be determined as the target strategy corresponding to the sub-constraint dimension; based on the target strategy corresponding to the sub-constraint dimension, consistency analysis is performed on the monitoring data corresponding to each data source to obtain the dimension analysis data corresponding to the sub-constraint dimension.

[0072] S204. Determine the weight coefficient corresponding to the sub-constraint dimension based on the dimension analysis data corresponding to the sub-constraint dimension.

[0073] Among them, the method of determining the weight coefficient corresponding to the sub-constraint dimension based on the dimension analysis data corresponding to the sub-constraint dimension is not limited.

[0074] In one implementation, for the dimension analysis data corresponding to the sub-constraint dimension, obtain the first preset coefficient matching the dimension analysis data; based on the mapping relationship between the sub-constraint dimension and the preset coefficient, the preset coefficient matching the sub-constraint dimension corresponding to the power marketing operation can be determined as the second preset coefficient; based on the first preset coefficient and the second preset coefficient, determine the weight coefficient corresponding to the sub-constraint dimension.

[0075] In some cases, the coefficient mean of the first preset coefficient and the second preset coefficient can be determined as the weight coefficient corresponding to the sub-constraint dimension. Or, the maximum value of the first preset coefficient and the second preset coefficient can be determined as the weight coefficient corresponding to the sub-constraint dimension.

[0076] S206. Perform weighted summation processing on the weight coefficients corresponding to each sub-constraint dimension and the dimension analysis data corresponding to each sub-constraint dimension to obtain the consistency constraint data matching the power marketing operation.

[0077] For example, represents the consistency constraint data, represents the monitoring data from different data sources. Taking the consistency constraint dimension including data consistency dimension, time consistency dimension and space consistency dimension as an example, Satisfy:

[0078]

[0079] Among them, represents the weight coefficient corresponding to the data consistency dimension when the sub-constraint dimension is the data consistency dimension, represents the dimension analysis result corresponding to the data consistency dimension when the sub-constraint dimension is the data consistency dimension, represents the weight coefficient corresponding to the time consistency dimension when the sub-constraint dimension is the time consistency dimension, represents the dimension analysis result corresponding to the time consistency dimension when the sub-constraint dimension is the time consistency dimension, Denotes the weight coefficient corresponding to the case where the sub-constraint dimension is the spatial consistency dimension. Denotes the dimensional analysis result corresponding to the case where the sub-constraint dimension is the spatial consistency dimension.

[0080] It should be understood that Is used to characterize the correlation result between data sources. Is used to characterize the temporal consistency result between data sources. Is used to characterize the spatial consistency result between data sources.

[0081] Based on Figure 2 The content shown, by performing weighted summation processing on the weight coefficients corresponding to each sub-constraint dimension and the dimensional analysis data corresponding to each sub-constraint dimension respectively, consistent constraint data matching the power marketing operation is obtained. Thus, by comprehensively considering each constraint dimension, the consistency of data analysis between data sources can be improved, and further, the model prediction accuracy can be enhanced.

[0082] In one embodiment, the implementation manner of obtaining the dimensional analysis data (i.e., S202) corresponding to the sub-constraint dimension by performing consistency analysis on the monitoring data corresponding to each data source based on the sub-constraint dimension can be as Figure 3 Shown, including the following steps:

[0083] S302, Select a target data source from each data source and determine each remaining data source among the data sources except the target data source.

[0084] Exemplarily, the target data source can be selected from each data source in the order of the data sources. For example, in the order of the data sources, the first data source is determined as the target data source, and each data source other than the first data source is determined as each remaining data source; the second data source is determined as the target data source, and each data source other than the second data source is determined as each remaining data source, and so on.

[0085] S304, Based on the difference characteristics between the monitoring data corresponding to the target data source and each remaining data source respectively, obtain the target analysis data corresponding to the target data source under the sub-constraint dimension.

[0086] Among them, the manner of obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension based on the difference characteristics between the monitoring data corresponding to the target data source and each remaining data source respectively is not limited.

[0087] In one implementation, for each remaining data source, the data difference between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data source is used as the difference feature between the monitoring data corresponding to the target data source and each of the remaining data sources; the data difference set representing each difference feature is determined as the target analysis data corresponding to the target data source under the sub-constraint dimension.

[0088] S306. Use the next data source adjacent to the target data source among all data sources as the new target data source, and return to the step of determining each remaining data source except the target data source among all data sources.

[0089] S308. According to the order of the target data sources, determine the data set containing the target analysis data corresponding to each target data source under the sub-constraint dimension as the dimension analysis data corresponding to the sub-constraint dimension.

[0090] Based on Figure 3 As shown in the content, by comparing the differences between the target data source and other remaining data sources, when obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension, the analysis accuracy can be improved, and further the model training accuracy can be improved.

[0091] In one embodiment, the sub-constraint dimension includes a data consistency dimension. Exemplarily, obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension based on the difference features between the monitoring data corresponding to the target data source and each of the remaining data sources (i.e., S304) may include the following steps:

[0092] S11. Obtain the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data source. The covariance is used to represent the difference feature between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data source.

[0093] Exemplarily, if the monitoring data includes multiple sub-data, then based on the multiple sub-data corresponding to the target data source and the multiple sub-data corresponding to the remaining data source, the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data source can be obtained.

[0094] S12. Obtain the first standard deviation matching the monitoring data corresponding to the target data source and the second standard deviation matching the monitoring data corresponding to the remaining data source.

[0095] Exemplarily, if the monitoring data includes multiple sub-data, then based on the multiple sub-data corresponding to the target data source, the first standard deviation matching the monitoring data corresponding to the target data source can be obtained; similarly, the second standard deviation matching the monitoring data corresponding to the remaining data source can be obtained.

[0096] S13. Obtain the target analysis data corresponding to the target data source in the sub-constraint dimension based on the respective covariances, the first standard deviation, and the respective second standard deviations corresponding to the target data source.

[0097] Exemplarily, the sub-constraint dimension includes the data consistency dimension, and the target analysis data corresponding to the target data source in the data consistency dimension satisfies:

[0098]

[0099] where N represents the number of data sources, represents the monitoring data corresponding to the target data source, represents the target analysis data corresponding to the target data source in the data consistency dimension, represents the monitoring data corresponding to the remaining data sources, represents the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, represents the first standard deviation matching the monitoring data corresponding to the target data source, represents the second standard deviation matching the monitoring data corresponding to the remaining data sources.

[0100] Furthermore, for each target data source, by combining the target analysis data respectively corresponding to each target data source, the dimension analysis data corresponding to the data consistency dimension can be obtained.

[0101] For example, represents the dimension analysis result corresponding to the data consistency dimension, then satisfies:

[0102]

[0103] Based on the content shown in S11 - S13, by considering the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, as well as their respective standard deviations, thus, when analyzing data from the perspective of data consistency, the analysis accuracy can be improved, and further the model maintenance accuracy can be improved.

[0104] In one embodiment, the sub-constraint dimension includes the time consistency dimension. Exemplarily, based on the difference features between the monitoring data corresponding to the target data source and each of the remaining data sources, obtaining the target analysis data (i.e., S304) corresponding to the target data source in the sub-constraint dimension may include the following steps:

[0105] S21. Divide the detection time window of the target data source to obtain multiple sub-detection moments.

[0106] Among them, the detection time windows of different target data sources may be the same or different.

[0107] Exemplarily, if the monitoring data includes multiple sub - data, then based on the collection times of the multiple sub - data corresponding to the target data source, the latest collection time and the earliest collection time are determined; based on the time interval between the latest collection time and the earliest collection time, the detection time window is determined.

[0108] In some cases, the collection time window composed of the time interval between the latest collection time and the earliest collection time can be determined as the detection time window. Or, a partial interval in the collection time window can be determined as the detection time window, that is, the collection time window includes the detection time window.

[0109] S22. For each sub - detection time, obtain the data difference between the sub - data of the remaining data sources at the sub - detection time and the sub - data of the target data source at the sub - detection time; the data difference is used to characterize the difference feature between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources.

[0110] Exemplarily, the data difference value can be characterized as the data difference.

[0111] S23. Based on the number of sub - detection times corresponding to the target data source and the respective data differences corresponding to each sub - detection time, obtain the target analysis data corresponding to the target data source under the sub - constraint dimension.

[0112] Exemplarily, the sub - constraint dimension includes the time - consistency dimension, and the target analysis data corresponding to the target data source under the time - consistency dimension satisfies:

[0113]

[0114] where T represents the number of sub - detection times corresponding to the target data source, represents the target analysis data corresponding to the target data source under the time - consistency dimension, represents the data difference between the sub - data of the remaining data sources at the sub - detection time t and the sub - data of the target data source at the sub - detection time t.

[0115] Furthermore, for each target data source, by combining the target analysis data respectively corresponding to each target data source, the dimension analysis data corresponding to the time - consistency dimension can be obtained.

[0116] For example, represents the dimension analysis result corresponding to the time - consistency dimension, then satisfies:

[0117]

[0118] Based on the content shown in S21 - S23, by analyzing the data differences between the sub - data of the remaining data sources and the sub - data of the target data source at the sub - detection moment, when analyzing data from the perspective of time consistency, the analysis accuracy can be improved, and further the model maintenance accuracy can be improved.

[0119] In one embodiment, the sub - constraint dimension includes the spatial consistency dimension. Exemplarily, based on the difference features between the monitoring data corresponding to the target data source and each of the remaining data sources, obtaining the target analysis data corresponding to the target data source under the sub - constraint dimension (i.e., S304) may include the following steps:

[0120] S31, obtain the data distance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, where the data distance is used to characterize the difference features between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources.

[0121] Among them, the data distance can refer to the Euclidean distance or other types of spatial distances.

[0122] S32, based on each data distance corresponding to the target data source and the number of data sources of at least two data sources, obtain the target analysis data corresponding to the target data source under the sub - constraint dimension.

[0123] Exemplarily, taking the data distance as the Euclidean distance as an example, the sub - constraint dimension includes the spatial consistency dimension, and the target analysis data corresponding to the target data source under the spatial consistency dimension satisfies:

[0124]

[0125] Among them, represents the target analysis data corresponding to the target data source under the spatial consistency dimension, represents the Euclidean distance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources.

[0126] Further, for each target data source, by combining the target analysis data respectively corresponding to each target data source, the dimension analysis data corresponding to the spatial consistency dimension can be obtained.

[0127] For example, represents the dimension analysis result corresponding to the spatial consistency dimension, then satisfies:

[0128]

[0129] Based on the content shown in S31 - S32, by calculating the data distance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, when analyzing data from the perspective of spatial consistency, the analysis accuracy can be improved, and further, the model maintenance accuracy can be enhanced.

[0130] It should be understood that in the collaborative work of multiple data sources, the importance and reliability of different data sources may change dynamically. Therefore, it is necessary to dynamically adjust the weights of data correlation, time consistency, and spatial consistency, that is, the weight coefficients corresponding to different sub - constraint dimensions, to ensure coordination in the model.

[0131] Specifically, the present application can use the entropy weight method for dynamic weight adjustment. The entropy weight method can effectively quantify the amount of information of each data source and dynamically adjust the weights according to the amount of information. Information entropy is an index to measure information uncertainty. The larger the information entropy, the more uncertain the amount of information of the data source, and its weight should be smaller; conversely, the data source with a smaller information entropy contains more certain information, and its weight should be larger.

[0132] In one embodiment, the dimension analysis data includes the target analysis data corresponding to each target data source. Exemplarily, the implementation method for determining the weight coefficient corresponding to the sub - constraint dimension (i.e., S204) based on the dimension analysis data corresponding to the sub - constraint dimension can be as Figure 4 shown, including the following steps:

[0133] S402, perform normalization processing on each target analysis data corresponding to the sub - constraint dimension to obtain the target normalized data corresponding to each target analysis data.

[0134] Exemplarily, taking the sub - constraint dimension including the data consistency dimension as an example, represents the target analysis data corresponding to the target data source under the data consistency dimension, then the target normalized data corresponding to the target analysis data under the data consistency dimension satisfies:

[0135]

[0136] Exemplarily, taking the sub - constraint dimension including the time consistency dimension as an example, represents the target analysis data corresponding to the target data source under the time consistency dimension, then the target normalized data corresponding to the target analysis data under the time consistency dimension satisfies:

[0137]

[0138] Exemplarily, taking the sub - constraint dimension including the spatial consistency dimension as an example, represents the target data source under the spatial consistency dimension For the corresponding target analysis data, the target normalized data corresponding to the target analysis data in the spatial consistency dimension satisfies:

[0139]

[0140] S404. For each target analysis data, obtain the entropy value corresponding to the target analysis data based on the target normalized data corresponding to the target analysis data and the number of data sources.

[0141] Exemplarily, taking the sub-constraint dimension including the data consistency dimension as an example, represents the target analysis data corresponding to the target data source in the data consistency dimension corresponding, represents the target normalized data corresponding to the target analysis data in the data consistency dimension. Then, the entropy value corresponding to the target analysis data in the data consistency dimension satisfies:

[0142]

[0143] where k is a constant, and k satisfies: , to ensure the normalization of entropy.

[0144] Exemplarily, taking the sub-constraint dimension including the time consistency dimension as an example, represents the target analysis data corresponding to the target data source in the time consistency dimension corresponding, represents the target normalized data corresponding to the target analysis data in the time consistency dimension. Then, the target normalized data corresponding to the target analysis data in the time consistency dimension satisfies:

[0145]

[0146] Exemplarily, taking the sub-constraint dimension including the spatial consistency dimension as an example, represents the target analysis data corresponding to the target data source in the spatial consistency dimension corresponding, represents the target normalized data corresponding to the target analysis data in the spatial consistency dimension. Then, the target normalized data corresponding to the target analysis data in the spatial consistency dimension satisfies:

[0147]

[0148] S406. Based on the maximum entropy value among the entropy values corresponding to each target analysis data and the entropy values corresponding to each target analysis data, obtain the sub-coefficients corresponding to each target analysis data in the sub-constraint dimension.

[0149] Exemplarily, taking the sub-constraint dimension including the data consistency dimension as an example, the sub-coefficient corresponding to the target analysis data under the data consistency dimension satisfies:

[0150]

[0151] Wherein, represents the target analysis data corresponding to the target data source under the data consistency dimension ; represents the entropy value corresponding to the target analysis data under the data consistency dimension ; represents the maximum entropy value among the entropy values respectively corresponding to each target analysis data under the data consistency dimension ; represents the sub-coefficient corresponding to the target analysis data under the data consistency dimension

[0152] Exemplarily, taking the sub-constraint dimension including the time consistency dimension as an example, the sub-coefficient corresponding to the target analysis data under the time consistency dimension satisfies:

[0153]

[0154] Wherein, represents the target analysis data corresponding to the target data source under the time consistency dimension ; represents the entropy value corresponding to the target analysis data under the time consistency dimension ; represents the maximum entropy value among the entropy values respectively corresponding to each target analysis data under the time consistency dimension ; represents the sub-coefficient corresponding to the target analysis data under the time consistency dimension

[0155] Exemplarily, taking the sub-constraint dimension including the space consistency dimension as an example, the sub-coefficient corresponding to the target analysis data under the space consistency dimension satisfies:

[0156]

[0157] Wherein, represents the target analysis data corresponding to the target data source under the space consistency dimension ; represents the entropy value corresponding to the target analysis data under the space consistency dimension ; represents the maximum entropy value among the entropy values respectively corresponding to each target analysis data under the space consistency dimension ; represents the sub-coefficient corresponding to the target analysis data under the space consistency dimension

[0158] S408. Obtain the weight coefficient corresponding to the sub-constraint dimension based on the sub-coefficients corresponding to the respective target analysis data under the sub-constraint dimension.

[0159] Exemplarily, combine the sub-coefficients corresponding to the respective target analysis data under the sub-constraint dimension to obtain the weight coefficient corresponding to the sub-constraint dimension.

[0160] Exemplarily, perform weighted processing on the sub-coefficients corresponding to the respective target analysis data under the sub-constraint dimension and the respective target analysis data under the sub-constraint dimension to obtain the data processing result under the sub-constraint dimension; perform a summation process on the data processing results respectively corresponding to each sub-constraint dimension to obtain the consistency constraint data matching the digital twin model.

[0161] Based on Figure 4 the content shown, by determining the weight coefficient corresponding to the sub-constraint dimension from the perspective of entropy, the determination accuracy of the weight coefficient can be improved, and further the maintenance accuracy of the digital twin model can be enhanced.

[0162] Combining the above content, in one embodiment, as Figure 5 shown, a method for maintaining a digital twin model of power marketing operations based on consistency constraints is provided. Taking the example that this method is applied to a server in a maintenance system for illustration, the following steps may be included:

[0163] S502. Obtain the digital twin model corresponding to the power marketing operation and the monitoring data collected by at least two data sources respectively for the power marketing operation.

[0164] S504. Determine the consistency constraint dimension that matches the power marketing operation based on the maintenance requirements of the digital twin model.

[0165] Exemplarily, the consistency constraint dimension includes data consistency dimension, time consistency dimension, and spatial consistency.

[0166] S506. Select a target data source from each data source and determine each remaining data source except the target data source in each data source.

[0167] S508. Based on the difference characteristics between the monitoring data corresponding to the target data source and each remaining data source, obtain the target analysis data corresponding to the target data source under the data consistency dimension, the target analysis data corresponding to the target data source under the time consistency dimension, and the target analysis data corresponding to the target data source under the spatial consistency dimension.

[0168] Exemplarily, for each remaining data source, obtain the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data source; obtain the first standard deviation matching the monitoring data corresponding to the target data source and the second standard deviation matching the monitoring data corresponding to the remaining data source; based on the covariances, the first standard deviation, and the second standard deviations corresponding to the target data source, obtain the target analysis data corresponding to the target data source in the data consistency dimension.

[0169] Exemplarily, divide the detection time window of the target data source to obtain multiple sub-detection moments; for each sub-detection moment, obtain the data difference between the sub-data of the remaining data source at the sub-detection moment and the sub-data of the target data source at the sub-detection moment; based on the number of sub-detection moments corresponding to the target data source and the respective data differences corresponding to the sub-detection moments, obtain the target analysis data corresponding to the target data source in the time consistency dimension.

[0170] Exemplarily, for each remaining data source, obtain the data distance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data source; based on the data distances corresponding to the target data source and the number of data sources of at least two data sources, obtain the target analysis data corresponding to the target data source in the space consistency dimension.

[0171] S510, take the next data source adjacent to the target data source among the data sources as the new target data source, and return the step of determining each remaining data source except the target data source among the data sources.

[0172] S512, based on the target analysis data corresponding to each target data source in the data consistency dimension, obtain the dimension analysis data corresponding to the data consistency dimension.

[0173] S514, based on the dimension analysis data corresponding to the data consistency dimension, determine the weight coefficient corresponding to the data consistency dimension.

[0174] S516, based on the target analysis data corresponding to each target data source in the time consistency dimension, obtain the dimension analysis data corresponding to the time consistency dimension.

[0175] S518, based on the dimension analysis data corresponding to the time consistency dimension, determine the weight coefficient corresponding to the time consistency dimension.

[0176] S520, based on the target analysis data corresponding to each target data source in the space consistency dimension, obtain the dimension analysis data corresponding to the space consistency dimension.

[0177] S522, based on the dimension analysis data corresponding to the space consistency dimension, determine the weight coefficient corresponding to the space consistency dimension.

[0178] S524, perform weighted summation on the dimension analysis data corresponding to the data consistency dimension, the dimension analysis data corresponding to the time consistency dimension, and the dimension analysis data corresponding to the space consistency dimension, respectively, with the weight coefficients corresponding to the data consistency dimension, the weight coefficients corresponding to the time consistency dimension, and the weight coefficients corresponding to the space consistency dimension, to obtain the consistency constraint data matching the digital twin model.

[0179] S526, use the consistency constraint data to maintain the digital twin model to obtain the updated model corresponding to the power marketing operation.

[0180] Among them, the specific content of S502 - S526 can refer to the foregoing content for adaptation description.

[0181] Combined with the above content, it can be seen that the method provided by this application solves the problems in the prior art in power marketing operations by introducing consistency constraints, ensures the consistency and synergy of information from different data sources during the model calibration process, and avoids the increase in model errors caused by data inconsistency. Thus, it can effectively solve the contradiction between diverse data sources and real-time feedback, optimize the accuracy and stability of the model while ensuring data consistency, thereby improving the efficiency and effectiveness of power marketing operations, and providing more reliable technical support for the efficient operation of power marketing operations.

[0182] Moreover, by calibrating the model based on the monitoring data collected separately for power marketing operations from at least two data sources, the parameters of the digital twin model can be automatically adjusted according to the real-time changes in power marketing operations. This flexible dynamic adjustment mechanism enables the model to continuously adapt to external changes such as power load and market demand, maintaining the continuous effectiveness and accuracy of the model. Moreover, it can also enable the digital twin model in power marketing operations to perform accurate calibration when facing complex power networks, thereby improving the prediction accuracy of the model. Compared with the prior art, this application can maintain a higher model accuracy and stability in a more complex and dynamic environment.

[0183] In summary, compared with the digital twin models in existing power marketing operations, the method provided in this application has the following advantages. On the one hand, in the existing technology, due to the inconsistency or error of multi-source data, the accuracy of the digital twin model decreases, and it cannot accurately reflect the operating state of the actual power system. However, in this application, by designing a consistency constraint mechanism, it is ensured that the data from different data sources (such as power load, grid status, equipment status, etc.) can meet the consistency requirements when input into the digital twin model, avoiding the problem of model accuracy degradation caused by data inconsistency between data sources. This consistency constraint mechanism enables effective coordination of multiple data sources in the model, solves the problem of model distortion caused by data inconsistency, and provides the stability and accuracy of the model. Moreover, by introducing the consistency constraint dimension and the entropy weight method, the influence of different data sources can be effectively adjusted, ensuring the consistency of the model input data and improving the prediction accuracy of the model in power marketing operations.

[0184] On the second hand, the digital twin models in the existing technology often cannot adapt to the changes of the power system in real time. However, in this application, by introducing an incremental learning strategy, the parameters of the digital twin model can be corrected according to the real-time feedback of errors, so as to ensure the fidelity of the model to the actual power system state, keep it consistent with the actual system state, effectively reduce the prediction error, make the model adapt to the dynamic changes of the power system, and improve the prediction accuracy of the digital twin model in power marketing operations and the accuracy and reliability of power dispatching.

[0185] It should be understood that although the steps in the flowcharts involved in the above-described embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, the execution of these steps has no strict order limit, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-described embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or steps or stages in other steps.

[0186] Based on the same inventive concept, an embodiment of the present application further provides a digital twin model maintenance device for a power marketing operation based on consistency constraints, which is used to implement the above-mentioned digital twin model maintenance method for a power marketing operation based on consistency constraints. The solution provided by this device to solve the problem is similar to the solution described in the above method. Therefore, the specific limitations in one or more embodiments of the digital twin model maintenance device for a power marketing operation based on consistency constraints provided below can refer to the limitations on the digital twin model maintenance method for a power marketing operation based on consistency constraints in the above text, and will not be elaborated here.

[0187] In an exemplary embodiment, as Figure 6 shown, a digital twin model maintenance device for a power marketing operation based on consistency constraints is provided, including: an acquisition module 602, a determination module 604, a processing module 606, and a maintenance module 608. Among them, the acquisition module 602 is used to acquire the digital twin model corresponding to the power marketing operation and the monitoring data collected by at least two data sources for the power marketing operation respectively; the determination module 604 is used to determine the consistency constraint dimension matching the power marketing operation based on the maintenance requirements of the digital twin model; the processing module 606 is used to perform consistency analysis on the monitoring data corresponding to each data source based on the consistency constraint dimension to obtain consistency constraint data matching the digital twin model; the maintenance module 608 is used to use the consistency constraint data to maintain the digital twin model to obtain an updated model corresponding to the power marketing operation.

[0188] In one of the embodiments, the consistency constraint dimension includes at least one sub-constraint dimension; the processing module 606 is further used to: for each sub-constraint dimension, perform consistency analysis on the monitoring data corresponding to each data source based on the sub-constraint dimension to obtain dimension analysis data corresponding to the sub-constraint dimension; determine the weight coefficient corresponding to the sub-constraint dimension based on the dimension analysis data corresponding to the sub-constraint dimension; perform weighted summation processing on the weight coefficients corresponding to each sub-constraint dimension and the dimension analysis data corresponding to each sub-constraint dimension to obtain consistency constraint data matching the power marketing operation.

[0189] In one embodiment, the processing module 606 is further configured to: select a target data source from each data source, and determine each remaining data source except the target data source among the data sources; obtain target analysis data corresponding to the target data source in the sub-constraint dimension based on the difference characteristics between the monitoring data corresponding to the target data source and each remaining data source; use the next data source adjacent to the target data source among the data sources as the new target data source, and return to the step of determining each remaining data source except the target data source among the data sources; determine, in the order of the target data sources, a data set including the target analysis data corresponding to each target data source in the sub-constraint dimension as the dimension analysis data corresponding to the sub-constraint dimension.

[0190] In one embodiment, the sub-constraint dimension includes a data consistency dimension; the processing module 606 is further configured to: obtain the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, where the covariance is used to characterize the difference characteristics between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources; obtain a first standard deviation matching the monitoring data corresponding to the target data source and a second standard deviation matching the monitoring data corresponding to the remaining data sources; obtain the target analysis data corresponding to the target data source in the sub-constraint dimension based on the covariances, the first standard deviation, and the second standard deviations corresponding to the target data source.

[0191] In one embodiment, the sub-constraint dimension includes a time consistency dimension, and the monitoring data includes multiple sub-data; the processing module 606 is further configured to: divide the detection time window of the target data source to obtain multiple sub-detection times; for each sub-detection time, obtain the data difference between the sub-data of the remaining data sources at the sub-detection time and the sub-data of the target data source at the sub-detection time, where the data difference is used to characterize the difference characteristics between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources; obtain the target analysis data corresponding to the target data source in the sub-constraint dimension based on the number of sub-detection times corresponding to the target data source and the data differences corresponding to each sub-detection time.

[0192] In one embodiment, the sub-constraint dimension includes a spatial consistency dimension; the processing module 606 is further configured to: obtain the data distance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, where the data distance is used to characterize the difference characteristics between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources; obtain the target analysis data corresponding to the target data source in the sub-constraint dimension based on the data distances corresponding to the target data source and the number of data sources of at least two data sources.

[0193] In one embodiment, the dimensional analysis data includes target analysis data corresponding to each target data source; the processing module 606 is further configured to: perform normalization processing on each target analysis data corresponding to the sub-constraint dimension to obtain target normalization data corresponding to each target analysis data; for each target analysis data, obtain the entropy value corresponding to the target analysis data based on the target normalization data corresponding to the target analysis data and the number of data sources; based on the maximum entropy value among the entropy values corresponding to each target analysis data and the entropy values corresponding to each target analysis data, obtain sub-coefficients corresponding to each target analysis data under the sub-constraint dimension; and obtain the weight coefficient corresponding to the sub-constraint dimension based on the sub-coefficients corresponding to each target analysis data under the sub-constraint dimension.

[0194] Each module in the above-described digital twin model maintenance device for power marketing operations based on consistency constraints can be implemented in whole or in part by software, hardware, and their combinations. The above-mentioned modules can be embedded in the processor in the computer device in hardware form or be independent of it, or can be stored in the memory in the computer device in software form, so that the processor can call and execute the operations corresponding to the above-mentioned modules.

[0195] In an exemplary embodiment, a computer device is provided. The computer device can be a server, and its internal structure diagram can be as Figure 7 shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, abbreviated as I / O), and a communication interface. Among them, the processor, the memory, and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as monitoring data corresponding to each data source. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with external terminals through a network connection. When the computer program is executed by the processor, it implements a method for maintaining a digital twin model for power marketing operations based on consistency constraints.

[0196] Those skilled in the art can understand that Figure 7 the structure shown in

[0197] In one embodiment, a computer device is further provided, including a memory and a processor. A computer program is stored in the memory, and when the processor executes the computer program, the steps in the above method embodiments are implemented.

[0198] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0199] In one embodiment, a computer program product is provided, including a computer program. When the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0200] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use, and processing of relevant data need to comply with relevant regulations.

[0201] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above methods. Among them, any reference to a memory, database, or other medium used in the embodiments provided in the present application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc. The databases involved in the embodiments provided in the present application can include at least one of relational databases and non-relational databases. Non-relational databases can include distributed databases based on blockchain, etc., without limitation. The processors involved in the embodiments provided in the present application can be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, data processing logics based on quantum computing, artificial intelligence (AI) processors, etc., without limitation.

[0202] The technical features of the above embodiments can be combined arbitrarily. For the sake of concise description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope recorded in the present application.

[0203] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.

Claims

1. A method for maintaining a digital twin model of power marketing operations based on consistency constraints, characterized in that, The method includes: Obtaining a digital twin model corresponding to a power marketing operation, and monitoring data collected by at least two data sources for the power marketing operation respectively; Determining a consistency constraint dimension matching the power marketing operation based on the maintenance requirements of the digital twin model; Performing consistency analysis on the monitoring data corresponding to each of the data sources based on the consistency constraint dimension to obtain consistency constraint data matching the digital twin model; Using the consistency constraint data to maintain the digital twin model to obtain an updated model corresponding to the power marketing operation.

2. The method according to claim 1, characterized in that The consistency constraint dimension includes at least one sub-constraint dimension; the performing consistency analysis on the monitoring data corresponding to each of the data sources based on the consistency constraint dimension to obtain consistency constraint data matching the digital twin model includes: For each of the sub-constraint dimensions, performing consistency analysis on the monitoring data corresponding to each of the data sources based on the sub-constraint dimension to obtain dimension analysis data corresponding to the sub-constraint dimension; Determining a weight coefficient corresponding to the sub-constraint dimension based on the dimension analysis data corresponding to the sub-constraint dimension; Performing weighted summation processing on the weight coefficients corresponding to each of the sub-constraint dimensions and the dimension analysis data corresponding to each of the sub-constraint dimensions to obtain consistency constraint data matching the power marketing operation.

3. The method according to claim 2, wherein The performing consistency analysis on the monitoring data corresponding to each of the data sources based on the sub-constraint dimension to obtain dimension analysis data corresponding to the sub-constraint dimension includes: Selecting a target data source from each of the data sources, and determining each remaining data source among the data sources except the target data source; Obtaining target analysis data corresponding to the target data source under the sub-constraint dimension based on the difference characteristics between the monitoring data corresponding to the target data source and each of the remaining data sources; Taking the next data source adjacent to the target data source among each of the data sources as a new target data source, and returning to the step of determining each remaining data source among the data sources except the target data source; According to the order of the target data sources, determining a data set including the target analysis data corresponding to each of the target data sources under the sub-constraint dimension as the dimension analysis data corresponding to the sub-constraint dimension.

4. The method according to claim 3, wherein The sub-constraint dimension includes a data consistency dimension; the method further includes: Obtaining the covariance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, where the covariance is used to characterize the difference characteristics between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources; The obtaining target analysis data corresponding to the target data source under the sub-constraint dimension based on the difference characteristics between the monitoring data corresponding to the target data source and each of the remaining data sources includes: Obtaining a first standard deviation matching the monitoring data corresponding to the target data source and a second standard deviation matching the monitoring data corresponding to the remaining data sources; Based on the covariances corresponding to the target data source, the first standard deviation, and the second standard deviations, obtain the target analysis data corresponding to the target data source under the sub-constraint dimension.

5. The method according to claim 3, characterized in that, The sub-constraint dimension includes a time consistency dimension, and the monitoring data includes multiple sub-data. The method further includes: Divide the detection time window of the target data source to obtain multiple sub-detection times. For each sub-detection time, obtain the data difference between the sub-data of the remaining data sources at the sub-detection time and the sub-data of the target data source at the sub-detection time; the data difference is used to characterize the difference feature between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources. The obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension based on the difference features between the monitoring data corresponding to the target data source and each of the remaining data sources includes: Based on the number of sub-detection times corresponding to the target data source and the data differences corresponding to each sub-detection time, obtain the target analysis data corresponding to the target data source under the sub-constraint dimension.

6. The method according to claim 3, characterized in that, The sub-constraint dimension includes a space consistency dimension; the method further includes: Obtain the data distance between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources, where the data distance is used to characterize the difference feature between the monitoring data corresponding to the target data source and the monitoring data corresponding to the remaining data sources. The obtaining the target analysis data corresponding to the target data source under the sub-constraint dimension based on the difference features between the monitoring data corresponding to the target data source and each of the remaining data sources includes: Based on the data distances corresponding to the target data source and the number of data sources of the at least two data sources, obtain the target analysis data corresponding to the target data source under the sub-constraint dimension.

7. The method according to claim 3, wherein The dimension analysis data includes the target analysis data corresponding to each target data source. The determining the weight coefficient corresponding to the sub-constraint dimension based on the dimension analysis data corresponding to the sub-constraint dimension includes: Perform normalization processing on each target analysis data corresponding to the sub-constraint dimension to obtain the target normalized data corresponding to each target analysis data. For each target analysis data, obtain the entropy value corresponding to the target analysis data based on the target normalized data corresponding to the target analysis data and the number of data sources. Based on the maximum entropy value among the entropy values corresponding to each target analysis data and the entropy values corresponding to each target analysis data, obtain the sub-coefficients corresponding to each target analysis data under the sub-constraint dimension. Based on the sub-coefficients corresponding to each target analysis data under the sub-constraint dimension, obtain the weight coefficient corresponding to the sub-constraint dimension.

8. A digital twin model maintenance device for power marketing operations based on consistency constraints, characterized in that, The device includes: An acquisition module, configured to acquire a digital twin model corresponding to an electric power marketing operation and monitoring data respectively collected by at least two data sources for the electric power marketing operation. A determination module, configured to determine a consistency constraint dimension matching the power marketing operation based on the maintenance requirements of the digital twin model; A processing module, configured to perform consistency analysis on the monitoring data respectively corresponding to each data source based on the consistency constraint dimension, and obtain consistency constraint data matching the digital twin model; A maintenance module, configured to maintain the digital twin model by using the consistency constraint data, and obtain an updated model corresponding to the power marketing operation.

9. A computer device, comprising a memory and a processor, the memory storing a computer program, characterized in that, When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images