A power digital modeling sharing service platform based on digital twin technology

Abstract

The present invention relates to the technical field of digitalization and intelligentization of electric power systems, and discloses an electric power digital modeling sharing service platform based on digital twin technology, comprising a data acquisition system, a data analysis system, a data prediction system and a data optimization system. Data acquisition is performed in the data acquisition system, and a real-time index #imgabs0#, a dynamic model error value #imgabs1#, a simulation model interval credibility #imgabs2# and an intelligent decision accuracy rate #imgabs3# are calculated in the data analysis system. The data prediction system predicts in advance according to the simulation model interval credibility #imgabs4#. The data prediction system and the data optimization system are connected through a network. The data optimization system optimizes the parameters of a synchronization data subunit, a dynamic model update subunit and an intelligent decision support subunit according to the real-time index #imgabs5#, a dynamic model error value #imgabs6# and an intelligent decision accuracy rate #imgabs7#. The above process can effectively solve the problem of the difficulty in realizing the integration and symbiosis of the virtual and the real.

Description

Technical Field

[0001] The present invention relates to the technical field of digitalization and intelligentization of power systems, and specifically to a power digital modeling sharing service platform based on digital twin technology. Background Art

[0002] The power digital modeling sharing service platform based on digital twin technology is a broad and in-depth technical field involving the digital and intelligent construction of power grids, the implementation and optimization of smart energy systems, and the integrated application of supporting technologies. The comprehensive application of this technology is expected to promote the development of the power industry towards intelligence and efficiency, and provide strong technical support for the reliable operation of power systems. However, there is a major flaw in the existing technology, that is, the difficulty in achieving the symbiosis of virtual and real. Although digital twin technology can construct an accurate digital mapping of the physical power grid, in practice, due to the simplification of the model and the introduction of assumptions, it may not be possible to fully replicate all the characteristics and behaviors of the physical power grid. In addition, even if the model can be accurately simulated, the difference between the simulation results and the real power grid may affect the effectiveness of the decision. This difficulty in virtual-real integration limits the potential application of digital twin technology in power systems, and further technological development and innovation are needed to address this challenge. Summary of the invention

[0003] 1. Technical issues to be resolved

[0004] In view of the shortcomings of the prior art, the present invention provides an electric power digital modeling sharing service platform based on digital twin technology, which has the advantage of simple implementation of the integration and symbiosis of the virtual and the real, and solves the problem of the difficulty in implementing the integration and symbiosis of the virtual and the real.

[0005] (II) Technical solution

[0006] To achieve the above-mentioned object, the present invention provides the following technical solutions: an electric power digital modeling sharing service platform based on digital twin technology, comprising a data acquisition system, a data analysis system, a data prediction system and a data optimization system;

[0007] The data acquisition system includes a real-time synchronization data subunit, a dynamic model update subunit, a simulation enhancement subunit and an intelligent decision support subunit. The real-time synchronization data subunit obtains real-time synchronization data through real-time data integration tools, message queues and data bus technologies, and after numbering, is connected to the data analysis system through a network. The dynamic model update subunit obtains dynamic model update data through sensors, and after numbering, is connected to the data analysis system through a network. The simulation enhancement subunit obtains simulation enhancement data through multi-dimensional sensors and monitoring equipment, and after numbering, is connected to the data analysis system through a network. The intelligent decision support subunit obtains intelligent decision support data through the network and numbers it. The data acquisition system is connected to the data analysis system through a network.

[0008] The data analysis system includes a data real-time subunit, a dynamic model error statistics subunit, a simulation model subunit and an intelligent decision evaluation subunit. The data real-time subunit calculates a real-time index Hr according to real-time synchronization data, the dynamic model error statistics subunit calculates a dynamic model error value Ij according to dynamic model update data, the simulation model subunit calculates a simulation model interval credibility Fl according to simulation enhancement data, and the intelligent decision evaluation subunit calculates an intelligent decision accuracy Kz according to intelligent decision support data. The data analysis system is connected to a data prediction system via a network;

[0009] The data prediction system makes predictions in advance according to the interval credibility Fl of the simulation model, and the data prediction system and the data optimization system are connected via a network;

[0010] The data optimization system optimizes the parameters of the synchronization data subunit, the dynamic model update subunit and the intelligent decision support subunit according to the real-time index Hr, the dynamic model error value Ij and the intelligent decision accuracy Kz.

[0011] Preferably, the real-time synchronization data subunit numbers the data sending time and the data receiving time according to the real-time synchronization data characteristics, and the data sending time is numbered as T 1 、T 2 、T 3 ,…T n , the data receiving time number is Ts 1 , Ts 2 , Ts 3 ,…Ts n , n represents the total number of registered real-time synchronization data.

[0012] Preferably, the data real-time subunit calculates the real-time index Hr according to the real-time synchronization data, and the calculation formula is:

[0013]

[0014] In the formula, Hr represents the real-time index, T 1 , T 2 , T 3 ,…T n Indicates the data sending time, T i Indicates the sending time of the i-th item of registered data, Ts 1 , Ts 2 , Ts 3 ,…Ts n Indicates data receiving time, Ts i Indicates the time when the i-th item of data is received. Indicates the registered n data transmission delays.

[0015] Preferably, the dynamic model updating subunit numbers the actual observed value of the power sample and the predicted value of the power sample according to the dynamic model updating data characteristics, and the actual observed value of the power sample is numbered D 1 , D 2 , D 3 ,…D n , the predicted value number of the power sample is Dj 1 ,Dj 2 ,Dj 3 ,…Dj n , n represents the total number of power samples.

[0016] Preferably, the dynamic model error statistics subunit calculates the dynamic model error value Ij according to the dynamic model update data, and the calculation formula is:

[0017]

[0018] In the formula, Ij represents the dynamic model error value, D 1 , D 2 , D 3 ,…D n represents the actual observed value of the power sample, D i represents the i-th actual observation value, Dj 1 ,Dj 2 ,Dj 3 ,…Dj n represents the predicted value of the power sample, Dji represents the i-th predicted value, and n represents the number of samples.

[0019] Preferably, the simulation enhancement subunit numbers the platform power, equipment and temperature data according to the simulation enhancement data characteristics, and the platform power, equipment and temperature data are numbered as L 1 , L 2 , L 3 ,…L n, n represents the total number of types of simulation enhancement data registration.

[0020] Preferably, the simulation model subunit calculates the simulation model interval credibility Fl according to the simulation enhancement data, and the calculation formula is:

[0021]

[0022] In the formula, Fl represents the interval credibility of the simulation model, L 1 , L 2 , L 3 ,…L n represents platform power, equipment and temperature data, n represents the total number of types of simulation enhancement data registration, L i represents the i-th simulation enhancement data registration type, Represents the sample mean of platform power, equipment and temperature data. Represents the square of the difference between the data point and the sample mean of the platform power, equipment and temperature data. represents the critical value of the t distribution, Indicates the sample standard deviation of platform power, equipment, and temperature data.

[0023] Preferably, the intelligent decision support subunit numbers the correctly predicted number of power grid failures, the incorrectly predicted number of power grid failures, the correctly predicted number of normal system operation states, and the number of actual failures that cannot be predicted according to the intelligent decision support data characteristics, and the correctly predicted number of power grid failures, the incorrectly predicted number of power grid failures, the correctly predicted number of normal system operation states, and the number of actual failures that cannot be predicted are numbered Z. 1 , Z 2 , Z 3 , Z 4 .

[0024] Preferably, the intelligent decision evaluation subunit calculates the intelligent decision accuracy Kz according to the intelligent decision support data, and the calculation formula is:

[0025]

[0026] In the formula, Kz represents the accuracy of intelligent decision-making, Z 1 , Z 2 , Z 3 , Z 4 It represents the number of correctly predicted grid faults, the number of incorrectly predicted grid faults, the number of correctly predicted normal system operation states, and the number of unpredicted actual faults. 1 +Z 3 represents the number of events predicted correctly, Z 1 +Z 2 +Z3 +Z 4 Represents the total number of predicted events.

[0027] Preferably, when the interval credibility of the simulation model Fl<0.1, the data prediction system will increase the real-time monitoring of the power simulation model in advance, and adjust the basic parameters of the safety performance of the power simulation model. When the interval credibility of the simulation model Fl≥0.1, the data prediction system will support the long-term trend analysis of the existing model, and establish a connection between the actual system and the virtual model; at the same time, the data transmission frequency and data update frequency of the synchronous data subunit are optimized according to the real-time index Hr;

[0028] The data optimization system optimizes the model parameters and model structure of the dynamic model update subunit according to the dynamic model error value Ij, and optimizes the learning rate range, regularization coefficient and batch size of the intelligent decision support subunit according to the intelligent decision accuracy rate Kz.

[0029] Compared with the prior art, the present invention provides a power digital modeling sharing service platform based on digital twin technology, which has the following beneficial effects:

[0030] 1. The data prediction system of the present invention predicts the actual parameters of the simulation model in advance by calculating the simulation model interval credibility Fl. When the simulation model interval credibility Fl<0.1, the data prediction system will increase the real-time monitoring of the power simulation model in advance, and adjust the basic parameters of the power simulation model's safety performance. When the simulation model interval credibility Fl≥0.1, the data prediction system will support the long-term trend analysis of the existing model, and use the optimized simulation model to establish a connection between the actual system and the virtual model to predict and analyze the behavior in the actual system, and substitute the actual monitored parameters into the simulation model to continuously improve and refine the model. By calculating the simulation model interval credibility Fl, the difficulty of realizing the integration of the virtual and the real can be effectively solved, and the application value and practicality of digital twin technology can be improved.

[0031] 2. The present invention calculates the real-time index Hr as the basis for optimizing the platform data transmission and processing flow. When the real-time index Hr value is at 1 / 3 of the preset threshold, the data prediction system will increase the data transmission frequency and data update frequency by 10% on the original basis. When the real-time index Hr is at 2 / 3 of the preset threshold, the data prediction system can also issue a fault warning, and indicate the area where the data transmission problem is located according to the position of the indicator light, so as to take measures in advance to avoid platform failure. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a schematic diagram of the structure of the present invention. DETAILED DESCRIPTION

[0033] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0034] See also Figure 1 , a power digital modeling sharing service platform based on digital twin technology, including data acquisition system, data analysis system, data prediction system and data optimization system;

[0035] The data acquisition system includes a real-time synchronization data subunit, a dynamic model update subunit, a simulation enhancement subunit and an intelligent decision support subunit. The real-time synchronization data subunit obtains real-time synchronization data through real-time data integration tools, message queues and data bus technologies, and after numbering, connects with the data analysis system through the network. The dynamic model update subunit obtains dynamic model update data through sensors, and after numbering, connects with the data analysis system through the network. The simulation enhancement subunit obtains simulation enhancement data through multi-dimensional sensors and monitoring equipment, and after numbering, connects with the data analysis system through the network. The intelligent decision support subunit obtains intelligent decision support data through the network and numbers it. The data acquisition system is connected with the data analysis system through the network.

[0036] The data analysis system includes a data real-time subunit, a dynamic model error statistics subunit, a simulation model subunit and an intelligent decision evaluation subunit. The data real-time subunit calculates a real-time index Hr according to real-time synchronization data, the dynamic model error statistics subunit calculates a dynamic model error value Ij according to dynamic model update data, the simulation model subunit calculates a simulation model interval credibility Fl according to simulation enhancement data, and the intelligent decision evaluation subunit calculates an intelligent decision accuracy Kz according to intelligent decision support data. The data analysis system is connected to the data prediction system through a network;

[0037] The data prediction system makes predictions in advance according to the interval credibility Fl of the simulation model, and the data prediction system and the data optimization system are connected through a network;

[0038] The data optimization system optimizes the parameters of the synchronization data subunit, dynamic model update subunit and intelligent decision support subunit according to the real-time index Hr, dynamic model error value Ij and intelligent decision accuracy Kz.

[0039] The real-time synchronization data subunit numbers the data sending time and data receiving time according to the real-time synchronization data characteristics. The data sending time is numbered as T 1 、T2 、T 3 ,…T n , the data receiving time number is Ts 1 , Ts 2 , Ts 3 ,…Ts n , n represents the total number of registered real-time synchronization data.

[0040] The data real-time subunit calculates the real-time index Hr based on the real-time synchronization data. The calculation formula is:

[0041]

[0042] In the formula, Hr represents the real-time index, T 1 、T 2 、T 3 ,…T n Indicates the data sending time, T i Indicates the sending time of the i-th item of registered data, Ts 1 , Ts 2 , Ts 3 ,…Ts n Indicates data receiving time, Ts i Indicates the time when the i-th item of data is received. Indicates the registered n data transmission delays.

[0043] The advantages are: by calculating the real-time index Hr as the basis for optimizing the platform data transmission and processing flow, when the real-time index Hr value is at 1 / 3 of the preset threshold, the data prediction system will increase the data transmission frequency and data update frequency by 10% on the original basis; when the real-time index Hr is at 2 / 3 of the preset threshold, the data prediction system can also issue a fault warning, and indicate the area where the data transmission problem is located according to the position of the indicator light, so as to take measures in advance to avoid platform failure.

[0044] The dynamic model update subunit numbers the actual observation value of the power sample and the predicted value of the power sample according to the dynamic model update data characteristics. The actual observation value of the power sample is numbered D 1 , D 2 , D 3 ,…D n , the predicted value number of the power sample is Dj 1 ,Dj 2 ,Dj 3 ,…Dj n , n represents the total number of power samples.

[0045] The dynamic model error statistics subunit calculates the dynamic model error value Ij according to the dynamic model update data, and its calculation formula is:

[0046]

[0047] In the formula, Ij represents the dynamic model error value, D 1 , D 2 , D 3 ,…D n represents the actual observed value of the power sample, D i represents the i-th actual observation value, Dj 1 ,Dj 2 ,Dj 3 ,…Dj n represents the predicted value of the power sample, Dji represents the i-th predicted value, and n represents the number of samples.

[0048] The advantages are: by continuously monitoring the dynamic model error value Ij, the data optimization system can promptly discover the changing trends of the model parameters and model structure in the dynamic model update subunit, and take corresponding measures to increase or decrease the model parameters in the model update subunit, while changing the complexity of the model structure while maintaining flexibility. This measure helps to achieve a high degree of consistency and accuracy between the virtual model and the actual system.

[0049] The simulation enhancement subunit numbers the platform power, equipment and temperature data according to the simulation enhancement data characteristics. The platform power, equipment and temperature data are numbered as L 1 , L 2 , L 3 ,…L n , n represents the total number of types of simulation enhancement data registration.

[0050] The simulation model subunit calculates the simulation model interval credibility Fl according to the simulation enhancement data, and its calculation formula is:

[0051]

[0052] In the formula, Fl represents the interval credibility of the simulation model, L 1 , L 2 , L 3 ,…L n represents platform power, equipment and temperature data, n represents the total number of types of simulation enhancement data registration, L i represents the i-th simulation enhancement data registration type, Represents the sample mean of platform power, equipment and temperature data. Represents the square of the difference between the data point and the sample mean of the platform power, equipment and temperature data. represents the critical value of the t distribution, Indicates the sample standard deviation of platform power, equipment, and temperature data.

[0053] The advantages are: the data prediction system predicts the actual parameters of the simulation model in advance by calculating the simulation model interval credibility Fl. When the simulation model interval credibility Fl<0.1, the data prediction system will increase the real-time monitoring of the power simulation model in advance, and adjust the basic parameters of the power simulation model's safety performance. When the simulation model interval credibility Fl≥0.1, the data prediction system will support the long-term trend analysis of the existing model, and use the optimized simulation model to establish a connection between the actual system and the virtual model to predict and analyze the behavior in the actual system, and substitute the actual monitored parameters into the simulation model to continuously improve and refine the model. By calculating the simulation model interval credibility Fl, the difficulty of realizing the integration of virtual and real can be effectively solved, and the application value and practicality of digital twin technology can be improved.

[0054] The intelligent decision support subunit numbers the correctly predicted number of power grid failures, the incorrectly predicted number of power grid failures, the correctly predicted number of normal system operation states, and the number of actual failures that were not predicted according to the characteristics of the intelligent decision support data. The correctly predicted number of power grid failures, the incorrectly predicted number of power grid failures, the correctly predicted number of normal system operation states, and the number of actual failures that were not predicted are numbered Z. 1 , Z 2 , Z 3 , Z 4 .

[0055] The intelligent decision evaluation subunit calculates the intelligent decision accuracy Kz based on the intelligent decision support data, and the calculation formula is:

[0056]

[0057] In the formula, Kz represents the accuracy of intelligent decision-making, Z 1 , Z 2 , Z 3 , Z 4 It represents the number of correctly predicted grid faults, the number of incorrectly predicted grid faults, the number of correctly predicted normal system operation states, and the number of unpredicted actual faults. 1 +Z 3 represents the number of events predicted correctly, Z 1 +Z 2 +Z 3 +Z 4 Represents the total number of predicted events.

[0058] The advantages are: the accuracy of intelligent decision is evaluated by calculating the intelligent decision accuracy Kz through the intelligent decision evaluation subunit. The calculation of the intelligent decision accuracy Kz includes not only the correctly predicted number of power grid failures, but also the correctly predicted number of normal system operation states. The number of correctly predicted events includes the correctly predicted number of power grid failures and the correctly predicted number of normal system operation states, and the number of all predicted events includes all predicted number of power grid failures and the number of normal system operation states. By continuously monitoring the intelligent decision accuracy Kz, the data optimization system can predict the development trend of model parameters and model structure in the intelligent decision support subunit. When the value of the intelligent decision accuracy Kz exceeds the preset value of 3 / 4, corresponding measures are taken immediately to expand the learning rate range of the intelligent decision support subunit, increase the regularization coefficient or reduce the batch size, thereby realizing the virtual and real symbiosis of platform power data.

[0059] When the interval credibility of the simulation model Fl<0.1, the data prediction system will increase the real-time monitoring of the power simulation model in advance, and adjust the basic parameters of the safety performance of the power simulation model. When the interval credibility of the simulation model Fl≥0.1, the data prediction system will support the long-term trend analysis of the existing model, and establish a connection between the actual system and the virtual model; at the same time, the data transmission frequency and data update frequency of the synchronous data sub-unit will be optimized according to the real-time index Hr;

[0060] The data optimization system optimizes the model parameters and model structure of the dynamic model update subunit according to the dynamic model error value Ij, and optimizes the learning rate range, regularization coefficient and batch size of the intelligent decision support subunit according to the intelligent decision accuracy Kz.

[0061] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A power digital modeling sharing service platform based on digital twin technology, characterized in that: Including data collection system, data analysis system, data prediction system and data optimization system; The data acquisition system includes a real-time synchronization data subunit, a dynamic model update subunit, a simulation enhancement subunit and an intelligent decision support subunit. The real-time synchronization data subunit obtains real-time synchronization data through real-time data integration tools, message queues and data bus technologies, and after numbering, is connected to the data analysis system through a network. The dynamic model update subunit obtains dynamic model update data through sensors, and after numbering, is connected to the data analysis system through a network. The simulation enhancement subunit obtains simulation enhancement data through multi-dimensional sensors and monitoring equipment, and after numbering, is connected to the data analysis system through a network. The intelligent decision support subunit obtains intelligent decision support data through the network and numbers it. The data acquisition system is connected to the data analysis system through a network. The data analysis system includes a data real-time subunit, a dynamic model error statistics subunit, a simulation model subunit and an intelligent decision evaluation subunit. The data real-time subunit calculates a real-time index Hr according to real-time synchronization data, the dynamic model error statistics subunit calculates a dynamic model error value Ij according to dynamic model update data, the simulation model subunit calculates a simulation model interval credibility Fl according to simulation enhancement data, and the intelligent decision evaluation subunit calculates an intelligent decision accuracy Kz according to intelligent decision support data. The data analysis system is connected to a data prediction system via a network; The data prediction system makes predictions in advance according to the interval credibility Fl of the simulation model, and the data prediction system and the data optimization system are connected via a network; The data optimization system optimizes the parameters of the synchronization data subunit, the dynamic model update subunit and the intelligent decision support subunit according to the real-time index Hr, the dynamic model error value Ij and the intelligent decision accuracy Kz.

2. According to claim 1, a power digital modeling sharing service platform based on digital twin technology is characterized by: The real-time synchronous data subunit numbers the data sending time and the data receiving time according to the real-time synchronous data characteristics, and the data sending time is numbered as T1, T2, T3, ... T n , the data receiving time numbers are Ts1, Ts2, Ts3, ...Ts n , n represents the total number of registered real-time synchronization data.

3. According to claim 2, a power digital modeling sharing service platform based on digital twin technology is characterized by: The data real-time subunit calculates the real-time index Hr according to the real-time synchronization data, and the calculation formula is: In the formula, Hr represents the real-time index, T1, T2, T3, ...T n Indicates the data sending time, T i Indicates the sending time of the i-th data item registered, Ts1, Ts2, Ts3, ...Ts n Indicates data receiving time, Ts i Indicates the time when the i-th item of data is received. Indicates the registered n data transmission delays.

4. According to claim 1, a power digital modeling sharing service platform based on digital twin technology is characterized by: The dynamic model updating subunit numbers the actual observed values ​​of the power samples and the predicted values ​​of the power samples according to the dynamic model updating data characteristics, and the actual observed values ​​of the power samples are numbered as D1, D2, D3, ... D n , the predicted values ​​of the power samples are numbered Dj1, Dj2, Dj3, ...Dj n , n represents the total number of power samples.

5. The power digital modeling sharing service platform based on digital twin technology according to claim 4 is characterized by: The dynamic model error statistics subunit calculates the dynamic model error value Ij according to the dynamic model update data, and its calculation formula is: In the formula, Ij represents the dynamic model error value, D1, D2, D3, ...D n represents the actual observed value of the power sample, D i represents the i-th actual observation value, Dj1, Dj2, Dj3, …Dj n Denotes the predicted value of the power sample, Dj i represents the i-th predicted value, and n represents the number of samples.

6. The power digital modeling sharing service platform based on digital twin technology according to claim 1 is characterized by: The simulation enhancement subunit numbers the platform power, equipment and temperature data according to the simulation enhancement data characteristics, and the platform power, equipment and temperature data are numbered as L1, L2, L3, ... L n , n represents the total number of types of simulation enhancement data registration.

7. The power digital modeling sharing service platform based on digital twin technology according to claim 6 is characterized by: The simulation model subunit calculates the simulation model interval credibility Fl according to the simulation enhancement data, and the calculation formula is: In the formula, Fl represents the interval credibility of the simulation model, L1, L2, L3, ... L n represents platform power, equipment and temperature data, n represents the total number of types of simulation enhancement data registration, L i represents the i-th simulation enhancement data registration type, Represents the sample mean of platform power, equipment and temperature data. Represents the square of the difference between the data point and the sample mean of the platform power, equipment and temperature data. represents the critical value of the t distribution, Indicates the sample standard deviation of platform power, equipment, and temperature data.

8. The power digital modeling sharing service platform based on digital twin technology according to claim 1 is characterized by: The intelligent decision support subunit numbers the correctly predicted number of power grid failures, the incorrectly predicted number of power grid failures, the correctly predicted number of normal system operation states, and the number of actual failures that could not be predicted according to the characteristics of the intelligent decision support data. The correctly predicted number of power grid failures, the incorrectly predicted number of power grid failures, the correctly predicted number of normal system operation states, and the number of actual failures that could not be predicted are numbered Z1, Z2, Z3, and Z4.

9. The power digital modeling sharing service platform based on digital twin technology according to claim 8, characterized in that: The intelligent decision evaluation subunit calculates the intelligent decision accuracy Kz based on the intelligent decision support data, and the calculation formula is: In the formula, Kz represents the accuracy of intelligent decision-making, Z1, Z2, Z3, and Z4 represent the number of correctly predicted power grid failures, the number of incorrectly predicted power grid failures, the number of correctly predicted normal system operating states, and the number of actual failures that were not predicted, Z1+Z3 represents the number of correctly predicted events, and Z1+Z2+Z3+Z4 represents the total number of predicted events.

10. The electric power digital modeling sharing service platform based on digital twin technology according to claim 9, characterized in that: When the interval credibility of the simulation model Fl<0.1, the data prediction system will increase the real-time monitoring of the power simulation model in advance, and adjust the basic parameters of the safety performance of the power simulation model. When the interval credibility of the simulation model Fl≥0.1, the data prediction system will support the long-term trend analysis of the existing model, and establish a connection between the actual system and the virtual model; at the same time, the data transmission frequency and data update frequency of the synchronous data subunit are optimized according to the real-time index Hr; The data optimization system optimizes the model parameters and model structure of the dynamic model update subunit according to the dynamic model error value Ij, and optimizes the learning rate range, regularization coefficient and batch size of the intelligent decision support subunit according to the intelligent decision accuracy rate Kz.

Images