Power grid data fusion analysis method, medium and system for live-action twins

Abstract

The invention discloses a live-action twinborn power grid data fusion analysis method, which comprises the following steps: S1, acquiring power grid data, and constructing a physical mechanism model of a power grid; s2, constructing a simulation model of the power grid based on the physical mechanism model; s3, introducing an AI algorithm model into the simulation model; s4, performing simulation analysis on the simulation model, performing real-time analysis on real-time simulation data in the simulation model based on an AI algorithm model, and adjusting simulation parameters of the simulation model; and S5, outputting a simulation analysis result by the simulation model. According to the power grid data fusion analysis method designed by the invention, high-approximation simulation is carried out by constructing a hybrid driving mode of a physical mechanism model, a simulation model and an AI algorithm model, so that the defects caused by single driving of data and the model are overcome, a power grid data twin entity can be constructed, high-quality digital-analog fusion is carried out, and the data fusion efficiency is improved. The generalization ability of the power grid system operation state evaluation method is improved, and the method has good popularization prospects and application values.

Description

Technical Field

[0001] The present invention relates to the technical field of digital analysis of power grids, and in particular to a real-scene twin power grid data fusion analysis method, medium and system. Background Art

[0002] As we all know, in the current existing technology, the methods for fusion analysis of multi-source spatiotemporal data mainly include: comprehensive evaluation theory, machine learning theory and uncertainty theory, etc. Among them, comprehensive evaluation theory specifically includes hierarchical analysis method, expert system and other methods; machine learning theory specifically includes deep belief network, support vector machine, random forest and other methods, which trains the model by inputting a large amount of data, and finally obtains a complex mapping relationship between multi-source data and output, but often the model performance is not ideal due to the heterogeneity of the data; and uncertainty theory specifically includes Bayesian network and DS evidence theory.

[0003] At present, in the field of electric power, existing researchers are considering using Bayesian networks in the process of electric power data fusion to conduct online learning at any time, thereby optimizing the structure and parameters of the network and improving the accuracy of probabilistic reasoning. In addition, some researchers are considering integrating multiple fault information and using improved DS evidence theory to achieve fault location. In addition, some researchers have also tried to use improved Bayesian classifiers to classify and assign credibility to transformer equipment, thereby effectively improving precision and diagnostic accuracy; in addition, some researchers have tried to solve the problem of information uncertainty to a certain extent based on DS evidence theory, but when the number of evidence bodies is large, it is difficult to determine its uncertainty and basic probability distribution function.

[0004] Therefore, for power data analysis in the power sector, the existing data analysis methods still have the following problems: (1) Traditional data processing technology cannot meet the analysis needs of quickly obtaining information and knowledge from massive power grid data; (2) The use of comprehensive evaluation, machine learning, uncertainty theory and other theories for digital-analog fusion is driven by data and models alone and cannot accurately reflect the complex power grid structure. Summary of the invention

[0005] The technical problem to be solved by the present invention is: the present invention discloses a real-scene twin power grid data fusion analysis method, medium and system to solve the problem that the existing data processing technology cannot meet the power grid analysis needs and cannot accurately reflect the complex power grid structure.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: a real-scene twin power grid data fusion analysis method, which comprises the steps of: S1: Obtain power grid data and build a physical mechanism model of the power grid; S2: constructing a simulation model of the power grid based on the physical mechanism model; S3: Introducing an AI algorithm model into the simulation model; S4: the simulation model performs simulation analysis, and based on the AI ​​algorithm model, performs real-time analysis on the real-time simulation data in the simulation model, and adjusts the simulation parameters of the simulation model; S5: The simulation model outputs simulation analysis results.

[0007] In the above-mentioned technical scheme of the present invention, the present invention designs a real-scene twin power grid data fusion analysis method, which can achieve high-approximation simulation and information virtual-real space alignment by constructing a hybrid driving mode of mechanism model + simulation model + AI algorithm model, so as to effectively overcome the shortcomings brought by the single driving of data and model. It can effectively improve the generalization ability of the power grid system operation status evaluation method by constructing a power grid data twin entity and performing high-quality digital-analog fusion, and realize the risk identification of component-level and system-level performance of the main equipment under complex working conditions in the digital space of the power equipment.

[0008] It should be pointed out in particular that the hybrid drive mode used in the power grid data fusion analysis method of the present invention not only involves the construction of the simulation model and data transmission, but also introduces an AI algorithm model into the above-mentioned simulation model to enhance the comprehensive, integrated and realistic modeling and simulation analysis of complex power grid structures and dynamic characteristics, which has good promotion prospects and application value.

[0009] Furthermore, in a real-scene twin power grid data fusion analysis method described in the present invention, in step S1, the power grid data includes: power grid topology, power grid equipment parameters and power grid equipment operating status data.

[0010] Furthermore, in the real-scene twin power grid data fusion analysis method described in the present invention, in step S2, specifically: On the basis of the physical mechanism model, the multi-physical field and multi-scale characteristics of the power grid are covered, and a structure tree of the simulation model is generated to construct a simulation model of the power grid and to simulate the dynamic operation process of the power grid.

[0011] Furthermore, in the real-scene twin power grid data fusion analysis method described in the present invention, in step S2, the following steps are specifically included: S21: Based on the physical mechanism model, obtaining the grid topology structure, grid equipment parameters and grid equipment operation status data including all grid equipment points; S22: Generate a shortest path tree based on the grid topology structure, grid equipment parameters and grid equipment operation status data of all grid equipment points; S23: Generate a structure tree of the simulation model according to the dependency relationship between each power grid equipment point, and determine the transmission path and transmission sequence of each power grid equipment point to construct the simulation model.

[0012] Furthermore, in the real-scene twin power grid data fusion analysis method described in the present invention, in step S3, the AI ​​algorithm model includes: a machine learning algorithm and a deep learning algorithm.

[0013] Furthermore, in a real-scene twin power grid data fusion analysis method described in the present invention, in step S3, specifically: introducing an AI algorithm model into the simulation model, and collecting real-time operation data, historical operation data and simulation data of the power grid, so as to clean and optimize the data in the simulation model through machine learning algorithms and deep learning algorithms, thereby improving the accuracy of the simulation model.

[0014] Furthermore, in the power grid data fusion analysis method of a real-scene twin described in the present invention, in step S4, specifically: The simulation model performs simulation analysis to obtain real-time simulation data; Analyzing the real-time simulation data using the AI ​​algorithm model to obtain analysis results, and generating adjustment strategies or control instructions according to the analysis results; The adjustment strategy or control instruction is fed back to the simulation model to adjust the simulation parameters of the simulation model.

[0015] Furthermore, in the real-scene twin power grid data fusion analysis method described in the present invention, step S6 is also included: Evaluate the simulation analysis results output by the simulation model, and iteratively optimize the simulation model and the AI ​​algorithm model based on the evaluation results of the simulation analysis results.

[0016] Accordingly, another object of the present invention is to disclose a computer medium, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor; Wherein, the processor is used to implement a real-scene twin power grid data fusion analysis method as described above in the present invention when executing the computer program.

[0017] In addition, another object of the present invention is to disclose a real-scene twin power grid data fusion analysis system, which is used to implement the real-scene twin power grid data fusion analysis method of the present invention, which specifically includes: Data acquisition module, to obtain power grid data; A physical model building module receives the power grid data and builds a physical mechanism model of the power grid; The simulation module constructs a simulation model of the power grid based on the physical mechanism model, and introduces an AI algorithm model into the simulation model to perform simulation analysis using the simulation model, and performs real-time analysis on real-time simulation data in the simulation model based on the AI ​​algorithm model to adjust simulation parameters of the simulation model, and the simulation model outputs simulation analysis results.

[0018] The beneficial effects of the present invention are: the present invention discloses a real-scene twin power grid data fusion analysis method, medium and system to solve the problem that the existing data processing technology cannot meet the power grid analysis needs and cannot accurately reflect the complex power grid structure; the power grid data fusion analysis method, medium and system can achieve high-approximation simulation and information virtual and real space alignment by constructing a hybrid drive mode of mechanism model + simulation model + AI algorithm model to effectively overcome the shortcomings brought by the single drive of data and model. It can effectively improve the generalization ability of the power grid system operation status evaluation method by constructing a power grid data twin entity and performing high-quality digital-analog fusion, and realize the risk identification of component-level and system-level performance of the main equipment under complex working conditions in the digital space of power equipment, and enhance the comprehensive, comprehensive and real modeling and simulation analysis of complex power grid structures and dynamic characteristics. It has good promotion prospects and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 The present invention is a flowchart of the steps of a real-scene twin power grid data fusion analysis method in one implementation.

[0020] Figure 2 It is a structural block diagram of a real-scene twin power grid data fusion analysis system according to one implementation mode of the present invention. DETAILED DESCRIPTION

[0021] In order to explain the technical content, achieved objectives and effects of the present invention in detail, the following is an explanation in combination with the implementation modes and the accompanying drawings.

[0022] The study found that in the current existing power grid field, the power grid data collected by the power grid digital perception layer is characterized by large scale, fast update speed, multi-source heterogeneity and low value density; and traditional data processing technology has encountered bottlenecks and cannot meet the analysis needs of quickly acquiring information and knowledge from massive data. Therefore, multi-source data fusion is a key technology for future power digital space research.

[0023] To do this, please refer to Figure 1 As shown, the present invention discloses a real-scene twin power grid data fusion analysis method, which specifically includes the steps of: S1: Obtain power grid data and build a physical mechanism model of the power grid; S2: constructing a simulation model of the power grid based on the physical mechanism model; S3: Introducing an AI algorithm model into the simulation model; S4: the simulation model performs simulation analysis, and based on the AI ​​algorithm model, performs real-time analysis on the real-time simulation data in the simulation model, and adjusts the simulation parameters of the simulation model; S5: The simulation model outputs simulation analysis results.

[0024] Based on the above settings, in response to the prominent problems of complex structure, uncertainty and complexity of power equipment in the power grid, the present invention can conduct comprehensive, integrated and realistic modeling and simulation analysis from the perspective of multi-physical fields and multi-scales. It constructs a hybrid driving method of mechanism model + simulation model + AI algorithm model to perform high-approximation simulation and information virtual and real space alignment to overcome the shortcomings brought by the single drive of data and model, thereby constructing a power grid data twin entity, and performing high-quality digital and analog fusion, improving the generalization ability of the system operation status evaluation method, and realizing the risk analysis and identification of component-level and system-level performance under complex working conditions of the main equipment in the digital space of the equipment.

[0025] Accordingly, in order to improve the accuracy and enhance the processing capability of the simulation model, the hybrid drive mode used in the power grid data fusion analysis method not only involves the construction of the simulation model and data transmission, but also specially introduces an AI algorithm model in the above-mentioned simulation model to enhance the comprehensive, integrated and realistic modeling and simulation analysis of complex power grid structures and dynamic characteristics, which has a good use effect.

[0026] In addition, it should be noted that in the present invention, when actually applying the power grid data fusion analysis method, a data mapping method that is actually applicable to the power grid data fusion analysis method can also be selected to realize the spatial transformation of multi-dimensional data, facilitate the digital presentation of the operating status of the power grid, and facilitate users to conduct intuitive analysis.

[0027] See also Figure 1 As shown, in order to ensure that the power grid data fusion analysis method of the present invention can perform comprehensive, integrated and realistic modeling and simulation analysis from the perspective of multi-physical fields and multi-scales, in the above step S1, the power grid topology structure, power grid equipment parameters and power grid equipment operating status data can be specifically obtained to construct a physical mechanism model, so as to facilitate in the subsequent step S2, on the basis of the above physical mechanism model, covering the multi-physical fields and multi-scale characteristics of the power grid, generating a structure tree of the simulation model, so as to construct a simulation model of the power grid, and use it to simulate the dynamic operation process of the power grid.

[0028] It should be noted that when the step S2 of the power grid data fusion analysis method designed by the present invention is actually executed, the following steps S21-S23 may be specifically included: S21: Based on the physical mechanism model, obtaining the grid topology structure, grid equipment parameters and grid equipment operation status data including all grid equipment points; S22: Generate a shortest path tree based on the grid topology structure, grid equipment parameters and grid equipment operation status data of all grid equipment points; S23: Generate a structure tree of the simulation model according to the dependency relationship between each power grid equipment point, and determine the transmission path and transmission sequence of each power grid equipment point to construct the simulation model.

[0029] Accordingly, in this embodiment, after constructing the above-mentioned simulation model, it is also necessary to introduce an AI algorithm model in the above-mentioned step S3 to process and optimize the data of the simulation model. The above-mentioned AI algorithm model may specifically include a machine learning algorithm and a deep learning algorithm.

[0030] Among them, in this embodiment, step S3 in the power grid data analysis method can be specifically as follows: introducing an AI algorithm model into the simulation model, and collecting real-time operation data, historical operation data and simulation data of the power grid, so as to clean and optimize the data in the simulation model through machine learning algorithms and deep learning algorithms, thereby improving the accuracy of the simulation model.

[0031] Therefore, if Figure 1 As shown, after the AI ​​algorithm model is introduced into the above-mentioned simulation model, the simulation model can be used to perform simulation analysis in step S4 of the power grid data analysis method of the present invention, and the real-time simulation data in the simulation model can be analyzed in real time based on the above-mentioned AI algorithm model, and the simulation parameters of the simulation model can be adjusted, which can specifically include the following steps: Use simulation models to conduct simulation analysis to obtain real-time simulation data; Analyzing the real-time simulation data using the AI ​​algorithm model to obtain analysis results, and generating adjustment strategies or control instructions according to the analysis results; The adjustment strategy or control instruction is fed back to the simulation model to adjust the simulation parameters of the simulation model.

[0032] Based on this, the above-mentioned AI algorithm model can be used to continuously optimize the above-mentioned simulation model to improve the simulation accuracy of the simulation model, thereby obtaining a better simulation effect, and in the subsequent step S5, the simulation model is used to output the simulation analysis results.

[0033] Of course, in actual application, in some preferred embodiments, based on the simulation analysis results obtained by the output of step S5, the staff can also evaluate them to evaluate the simulation analysis results output by the simulation model, and continuously iterate and optimize the above-mentioned simulation model and AI algorithm model according to the actual evaluation results, so as to improve the accuracy of the simulation analysis and the effectiveness of the adjustment strategy.

[0034] Correspondingly, in the present invention, the present invention also discloses a computer medium, which includes: a memory, a processor, and a computer program stored in the memory and executable on the processor; wherein the processor is used to implement a real-scene twin power grid data fusion analysis method as described above in the present invention when executing the computer program.

[0035] In addition, if Figure 2 As shown, in order to facilitate the implementation of the above-mentioned power grid data fusion analysis method of the present invention, the present invention also discloses a real-scene twin power grid data fusion analysis system, which can be specifically used to implement the above-mentioned power grid data fusion analysis method of the present invention, and can specifically include: a data acquisition module, a physical model building module and a simulation module; wherein the above-mentioned data acquisition module is used to acquire power grid data; the above-mentioned physical model building module is used to receive the above-mentioned power grid data, and construct a physical mechanism model of the power grid based on the power grid data.

[0036] The above-mentioned simulation module can construct a simulation model of the power grid based on the physical mechanism model established by the physical model building module, and introduce an AI algorithm model into the simulation model to perform simulation analysis using the simulation model, and perform real-time analysis on real-time simulation data in the simulation model based on the AI ​​algorithm model to adjust the simulation parameters of the simulation model, and the simulation model outputs the simulation analysis results.

[0037] Of course, in actual application, the power grid data fusion analysis system may also include a data storage module to store various historical data based on the data storage module, so as to facilitate the subsequent AI algorithm model and staff's call and analysis.

[0038] From the above, it can be seen that the power grid data fusion analysis method, medium and system of the present invention can achieve high-approximation simulation and information virtual and real space alignment by constructing a hybrid drive mode of mechanism model + simulation model + AI algorithm model, so as to effectively overcome the shortcomings brought by the single drive of data and model. It can effectively improve the generalization ability of the power grid system operation status evaluation method by constructing a power grid data twin entity and performing high-quality digital-analog fusion, and realize the risk identification of component-level and system-level performance of the main equipment under complex working conditions in the digital space of power equipment, and realize comprehensive, integrated and real modeling and simulation analysis of complex power grid structure and dynamic characteristics, and its promotion prospect and application value are relatively good.

[0039] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent transformations made using the contents of the present invention's specification and drawings, or directly or indirectly applied in related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A real-scene twin power grid data fusion analysis method, characterized in that: Includes steps: S1: Obtain power grid data and build a physical mechanism model of the power grid; S2: constructing a simulation model of the power grid based on the physical mechanism model; S3: Introducing an AI algorithm model into the simulation model; S4: the simulation model performs simulation analysis, and based on the AI ​​algorithm model, performs real-time analysis on the real-time simulation data in the simulation model, and adjusts the simulation parameters of the simulation model; S5: The simulation model outputs simulation analysis results.

2. The power grid data fusion analysis method of a real-scene twin according to claim 1 is characterized in that: In step S1, the power grid data includes: power grid topology, power grid equipment parameters and power grid equipment operating status data.

3. The power grid data fusion analysis method of a real-scene twin according to claim 1 is characterized in that: In step S2, specifically: On the basis of the physical mechanism model, the multi-physical field and multi-scale characteristics of the power grid are covered, and a structure tree of the simulation model is generated to construct a simulation model of the power grid and to simulate the dynamic operation process of the power grid.

4. The power grid data fusion analysis method of a real-scene twin according to claim 2 is characterized in that: In step S2, the steps are specifically included: S21: Based on the physical mechanism model, obtaining the grid topology structure, grid equipment parameters and grid equipment operation status data including all grid equipment points; S22: Generate a shortest path tree based on the grid topology structure, grid equipment parameters and grid equipment operation status data of all grid equipment points; S23: Generate a structure tree of the simulation model according to the dependency relationship between each power grid equipment point, and determine the transmission path and transmission sequence of each power grid equipment point to construct the simulation model.

5. The power grid data fusion analysis method of a real-scene twin according to claim 1 is characterized in that: In step S3, the AI ​​algorithm model includes: a machine learning algorithm and a deep learning algorithm.

6. The power grid data fusion analysis method of a real-scene twin according to claim 5 is characterized in that: In step S3, specifically: an AI algorithm model is introduced into the simulation model, and real-time operation data, historical operation data and simulation data of the power grid are collected, so as to clean and optimize the data in the simulation model through machine learning algorithms and deep learning algorithms, so as to improve the accuracy of the simulation model.

7. The power grid data fusion analysis method of a real-scene twin according to claim 1 is characterized in that: In step S4, specifically: The simulation model performs simulation analysis to obtain real-time simulation data; Analyzing the real-time simulation data using the AI ​​algorithm model to obtain analysis results, and generating adjustment strategies or control instructions according to the analysis results; The adjustment strategy or control instruction is fed back to the simulation model to adjust the simulation parameters of the simulation model.

8. The power grid data fusion analysis method of a real-scene twin according to claim 1 is characterized in that: The step S6 is also included: Evaluate the simulation analysis results output by the simulation model, and iteratively optimize the simulation model and the AI ​​algorithm model based on the evaluation results of the simulation analysis results.

9. A computer medium, characterized in that include: A memory, a processor, and a computer program stored in the memory and executable on the processor; Wherein, the processor is used to implement a real-scene twin power grid data fusion analysis method as described in any one of claims 1-8 when executing the computer program.

10. A real-scene twin power grid data fusion analysis system, which is used to implement a real-scene twin power grid data fusion analysis method as described in any one of claims 1 to 8, characterized in that: include: Data acquisition module, to obtain power grid data; A physical model building module receives the power grid data and builds a physical mechanism model of the power grid; The simulation module constructs a simulation model of the power grid based on the physical mechanism model, and introduces an AI algorithm model into the simulation model to perform simulation analysis using the simulation model, and performs real-time analysis on real-time simulation data in the simulation model based on the AI ​​algorithm model to adjust simulation parameters of the simulation model, and the simulation model outputs simulation analysis results.

Images