A method, system, and medium for power grid natural disaster prediction and defense

Abstract

This application provides a method, system, and medium for predicting and defending against natural disasters in power grids. The method includes: constructing a knowledge-enhanced digital twin of power grid disasters by acquiring multi-source power grid operation record data; extracting and analyzing disaster event assessment datasets to obtain a set of virtual disaster event sequence scenarios; acquiring real-time disaster assessment data and analyzing it using a preset disaster intensity time-series prediction model to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators; analyzing and processing real-time status data in conjunction with the power grid operation risk assessment report and the set of power grid operation resilience indicators to obtain a list of high-frequency faults in power grid equipment; and performing matching analysis using the knowledge-enhanced digital twin of power grid disasters to determine the optimal solution for power grid disaster handling. This application achieves the prediction and defense against natural disasters in power grids by constructing a digital twin, generating a disaster scenario set, conducting dual-channel collaborative early warning simulations, and implementing human-computer interactive resilience assessments.

Description

Technical Field

[0001] This application relates to the field of power grid security and disaster prevention technology, and more specifically, to a method, system, and medium for predicting and defending against natural disasters affecting power grids. Background Technology

[0002] The environment in which power grid equipment is located is susceptible to erosion from various natural disasters such as typhoons, freezing, wildfires, and flash floods. Traditional disaster prediction, early warning, and defense technologies rely on purely data-driven early warning models, but lack the constraints of physical mechanisms, resulting in insufficient reliability of model outputs. Existing early warning technologies focus on short-term warnings and lack quantitative assessments of the dynamic resilience of power grids under continuous disaster impacts, which is insufficient to provide defensive measures based on early warnings. Disaster scenario analysis mainly relies on historical disaster data, which cannot cover extreme disaster scenarios. The ability to handle low-probability extreme disaster events is insufficient, and data, knowledge, and models have not been deeply integrated, resulting in insufficient reasoning, analysis, and decision-making capabilities.

[0003] Effective technical solutions are urgently needed to address the above problems. Summary of the Invention

[0004] The purpose of this application is to provide a method, system, and medium for predicting and defending against natural disasters in power grids. This can be achieved by constructing digital twins, generating disaster scenario sets, conducting dual-channel collaborative early warning simulations, and performing human-computer interaction resilience assessments.

[0005] Firstly, this application provides methods for predicting and mitigating natural disasters affecting power grids, including the following steps:

[0006] Acquire multi-source power grid operation record data, and construct and process digital twins based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin;

[0007] Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, and the disaster event assessment dataset is analyzed and processed to obtain a set of virtual disaster event sequence scenarios.

[0008] Real-time disaster assessment data is acquired and analyzed using a pre-set disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators.

[0009] The real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin is obtained, and analyzed and processed in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment.

[0010] Based on the high-frequency fault list of power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin, a matching analysis is performed to obtain a preset number of power grid disaster handling schemes. These schemes are then analyzed and processed to obtain optimized values ​​for power grid operational resilience indicators and to determine the optimal power grid disaster handling scheme.

[0011] Optionally, in the method for predicting and preventing natural disasters in power grids described in this application, the step of acquiring multi-source power grid operation record data and performing digital twin construction processing based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin includes:

[0012] Acquire multi-source power grid operation record data, including static power grid parameter data, dynamic power grid operation record data, real-time disaster and environmental impact data, historical power grid disaster record data, and power grid equipment operation and maintenance record data;

[0013] A power grid physical and safety mechanism model library is constructed based on the static power grid parameter data, dynamic power grid operation record data, and historical power grid disaster record data.

[0014] A natural disaster dynamics model library is constructed based on the real-time disaster environmental impact data and historical power grid disaster record data.

[0015] Based on the historical power grid disaster record data and power grid equipment operation and maintenance record data, text parsing, entity recognition, relationship extraction and data fusion processing are performed to obtain a power grid operation knowledge graph;

[0016] The multi-source power grid operation record data, power grid physical and safety mechanism model library, natural disaster dynamics model library, and power grid operation knowledge graph are integrated through data models and service interfaces to obtain a knowledge-enhanced power grid disaster-coupled digital twin.

[0017] Optionally, in the method for predicting and preventing natural disasters in power grids described in this application, the step of extracting a disaster event assessment dataset from the multi-source power grid operation record data, and analyzing and processing the disaster event assessment dataset to obtain a set of virtual disaster event sequence scenarios includes:

[0018] Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, including disaster type characteristic data, disaster intensity, disaster location data, disaster duration and corresponding historical geographical and climatic data;

[0019] Obtain geographic climate prediction data;

[0020] Based on the disaster type characteristic data, disaster intensity, disaster location data, and disaster duration data, as well as the corresponding historical geographic and climate data, and combined with the geographic and climate prediction data, the model is trained to obtain a disaster statistical model and a disaster generation adversarial network model.

[0021] Based on the disaster statistical model, disaster parameters are extracted and combined to generate a disaster event sequence;

[0022] Based on the disaster generation adversarial network model and preset extreme disaster generation conditions, an extreme disaster event sequence is generated;

[0023] The disaster event sequence and the extreme disaster event sequence are fused together to obtain a set of virtual disaster event sequence scenarios.

[0024] Optionally, in the method for predicting and preventing natural disasters in power grids described in this application, the step of acquiring real-time disaster assessment data and analyzing and processing it through a preset disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a power grid operation resilience index set includes:

[0025] Based on the aforementioned knowledge, the power grid disaster-coupled digital twin is enhanced to obtain real-time disaster assessment data, including meteorological remote sensing monitoring data stream, ground meteorological monitoring data stream, power grid operation status data stream, and equipment operation status data stream;

[0026] The meteorological remote sensing monitoring data stream, the ground meteorological monitoring data stream, the power grid operation status data stream, and the equipment operation status data stream are input into a preset disaster intensity time series prediction model for analysis and processing to obtain disaster field prediction data, a list of risky equipment, and physical consistency scores.

[0027] Based on the disaster field prediction data, risk equipment list, and physical consistency score, a physical simulation analysis is performed using a natural disaster dynamics model to obtain the disaster intensity field.

[0028] Based on the disaster intensity field, a coupled extrapolation and analysis is performed using a power grid physics and safety mechanism model to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators.

[0029] Optionally, in the method for predicting and preventing natural disasters in power grids described in this application, the step of obtaining the real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin, and analyzing and processing it in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults in power grid equipment includes:

[0030] Acquire the real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin;

[0031] By acquiring dynamic adjustment parameters for disaster prevention, and combining the real-time status data, power grid operation risk assessment report, and power grid operation resilience index set, the disaster intensity field is re-analyzed and processed through the power grid physical and safety mechanism model, and the power grid weak link analysis is performed to obtain a list of high-frequency faults of power grid equipment.

[0032] Optionally, in the method for predicting and preventing power grid natural disasters described in this application, the step of performing matching analysis based on the high-frequency fault list of power grid equipment combined with the knowledge-enhanced power grid disaster coupled digital twin to obtain a preset number of power grid disaster handling schemes, and performing analysis and processing to obtain optimized values ​​of power grid operation resilience indicators, and determining the optimal power grid disaster handling scheme, includes:

[0033] Based on the high-frequency fault list of the power grid equipment and the power grid operation knowledge graph of the knowledge-enhanced power grid disaster coupled digital twin, a predetermined number of power grid disaster handling solutions are obtained through matching analysis.

[0034] The parameters of the power grid disaster handling scheme are obtained, and the typical disaster event scenario subset is extracted from the virtual disaster event sequence scenario set through cluster analysis. The disaster intensity field is then re-analyzed and processed through the power grid physical and safety mechanism model to obtain the optimized value of the power grid operation resilience index corresponding to the power grid disaster handling scheme.

[0035] The optimized values ​​of the power grid operation resilience index are sorted in descending order, and the power grid disaster handling scheme with the largest optimized value is determined as the optimal power grid disaster handling scheme.

[0036] Secondly, this application provides a system for predicting and preventing natural disasters in power grids. The system includes a memory and a processor. The memory includes a program for predicting and preventing natural disasters in power grids. When the program for predicting and preventing natural disasters in power grids is executed by the processor, it implements the following steps:

[0037] Acquire multi-source power grid operation record data, and construct and process digital twins based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin;

[0038] Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, and the disaster event assessment dataset is analyzed and processed to obtain a set of virtual disaster event sequence scenarios.

[0039] Real-time disaster assessment data is acquired and analyzed using a pre-set disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators.

[0040] The real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin is obtained, and analyzed and processed in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment.

[0041] Based on the high-frequency fault list of power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin, a matching analysis is performed to obtain a preset number of power grid disaster handling schemes. These schemes are then analyzed and processed to obtain optimized values ​​for power grid operational resilience indicators and to determine the optimal power grid disaster handling scheme.

[0042] Optionally, in the power grid natural disaster prediction and prevention system described in this application, the step of acquiring multi-source power grid operation record data and performing digital twin construction processing based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin includes:

[0043] Acquire multi-source power grid operation record data, including static power grid parameter data, dynamic power grid operation record data, real-time disaster and environmental impact data, historical power grid disaster record data, and power grid equipment operation and maintenance record data;

[0044] Based on the static power grid parameter data, dynamic power grid operation record data, and historical power grid disaster record data, a power grid physical and safety mechanism model library is constructed, including power flow calculation models, transient stability models, equipment heat capacity models, and tower mechanical strength models;

[0045] A natural disaster dynamics model library is constructed based on the real-time disaster environmental impact data and historical power grid disaster record data, including typhoon wind field model, icing growth model, wildfire spread model and flood evolution model;

[0046] Based on the historical power grid disaster record data and power grid equipment operation and maintenance record data, text parsing, entity recognition, relationship extraction and data fusion processing are performed to obtain a power grid operation knowledge graph;

[0047] The multi-source power grid operation record data, power grid physical and safety mechanism model library, natural disaster dynamics model library, and power grid operation knowledge graph are integrated through data models and service interfaces to obtain a knowledge-enhanced power grid disaster-coupled digital twin.

[0048] Optionally, in the power grid natural disaster prediction and prevention system described in this application, the step of extracting a disaster event assessment dataset from the multi-source power grid operation record data, and analyzing and processing the disaster event assessment dataset to obtain a virtual disaster event sequence scenario set includes:

[0049] Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, including disaster type characteristic data, disaster intensity, disaster location data, disaster duration and corresponding historical geographical and climatic data;

[0050] Obtain geographic climate prediction data;

[0051] Based on the disaster type characteristic data, disaster intensity, disaster location data, and disaster duration data, as well as the corresponding historical geographic and climate data, and combined with the geographic and climate prediction data, the model is trained to obtain a disaster statistical model and a disaster generation adversarial network model.

[0052] Based on the disaster statistical model, disaster parameters are extracted and combined to generate a disaster event sequence;

[0053] Based on the disaster generation adversarial network model and preset extreme disaster generation conditions, an extreme disaster event sequence is generated;

[0054] The disaster event sequence and the extreme disaster event sequence are fused together to obtain a set of virtual disaster event sequence scenarios.

[0055] Thirdly, this application also provides a computer-readable storage medium storing a method program for predicting and preventing natural disasters in a power grid. When executed by a processor, the method program implements the steps of the method for predicting and preventing natural disasters in a power grid as described in any of the preceding claims.

[0056] As can be seen from the above, the method, system and medium for predicting and defending against natural disasters in the power grid provided in this application realize the prediction and defense against natural disasters in the power grid by constructing a digital twin, generating a disaster scenario set, performing dual-channel collaborative early warning simulation and human-computer interactive resilience assessment.

[0057] Other features and advantages of this application will be set forth in the following description and will be apparent in part from the description or may be learned by practicing embodiments of this application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings. Attached Figure Description

[0058] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0059] Figure 1 A flowchart illustrating a method for predicting and preventing natural disasters in a power grid, provided as an embodiment of this application;

[0060] Figure 2 A flowchart illustrating the acquisition of a knowledge-enhanced power grid disaster-coupled digital twin, provided for an embodiment of this application, in a method for predicting and mitigating natural disasters in a power grid;

[0061] Figure 3 A high-level flowchart of a method for predicting and preventing natural disasters in power grids, provided in an embodiment of this application. Detailed Implementation

[0062] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0063] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0064] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating a method for predicting and mitigating natural disasters in a power grid, as described in some embodiments of this application. This method is used in terminal devices, such as computers and mobile terminals. The method includes the following steps:

[0065] S11. Obtain multi-source power grid operation record data, and perform digital twin construction processing based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin;

[0066] S12. Extract disaster event assessment dataset from the multi-source power grid operation record data, analyze and process the disaster event assessment dataset, and obtain a set of virtual disaster event sequence scenarios.

[0067] S13. Obtain real-time disaster assessment data, and analyze and process it through a preset disaster intensity time series prediction model and a knowledge-enhanced power grid disaster coupling digital twin to obtain a power grid operation risk assessment report and a power grid operation resilience index set.

[0068] S14. Obtain the real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin, and analyze and process it in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment.

[0069] S15. Based on the high-frequency fault list of the power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin, a matching analysis is performed to obtain a preset number of power grid disaster handling schemes, and the analysis and processing are performed to obtain the optimized value of the power grid operation resilience index, and the optimal power grid disaster handling scheme is determined.

[0070] Further explanation is needed regarding the implementation architecture for accurately predicting and effectively defending against natural disasters. This architecture incorporates a hybrid physical information neural network, adversarial scenario generation, knowledge graph enhancement, and human-computer interactive simulation. First, by fusing static data, dynamic data, physical mechanism models, and knowledge graphs, a knowledge-enhanced digital twin of the power grid disaster is constructed. Then, a set of virtual disaster event sequence scenarios, including historical and extreme disasters, is generated and subjected to dual-channel collaborative early warning analysis. Next, human-computer interactive dynamic resilience simulation and weak node assessment are achieved through manual dynamic adjustment of scheme parameters. Finally, defense measures are quantitatively evaluated based on typical disaster scenarios to achieve accurate selection of defense methods.

[0071] Please refer to Figure 2 , Figure 2 This is a flowchart illustrating the acquisition of a knowledge-enhanced power grid disaster-coupled digital twin, part of a method for predicting and mitigating natural disasters in a power grid, as described in some embodiments of this application. According to an embodiment of the present invention, the step of acquiring multi-source power grid operation record data and performing digital twin construction processing based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin includes:

[0072] S21. Obtain multi-source power grid operation record data, including static power grid parameter data, dynamic power grid operation record data, real-time disaster and environmental impact data, historical power grid disaster record data, and power grid equipment operation and maintenance record data;

[0073] S221. Construct a power grid physical and safety mechanism model library based on the static power grid parameter data, dynamic power grid operation record data, and historical power grid disaster record data;

[0074] S222. Construct a natural disaster dynamics model library based on the real-time disaster environmental impact data and historical power grid disaster record data;

[0075] S223. Based on the historical power grid disaster record data and power grid equipment operation and maintenance record data, perform text parsing, entity recognition, relationship extraction and data fusion processing to obtain a power grid operation knowledge graph;

[0076] S23. Integrate the multi-source power grid operation record data, power grid physical and safety mechanism model library, natural disaster dynamics model library, and power grid operation knowledge graph through data models and service interfaces to obtain a knowledge-enhanced power grid disaster-coupled digital twin.

[0077] Further explanation is needed regarding the construction of a knowledge-enhanced power grid disaster-coupled digital twin that integrates multi-source information and possesses knowledge reasoning capabilities, serving as a dynamic virtual power grid mirror. First, static power grid parameter data is acquired from the enterprise asset management system, including power grid geographic information system data and equipment building information model data. The power grid geographic information system data includes line routes, tower coordinates, and substation locations, while the equipment building information model data includes equipment 3D models, materials, and design parameters. Dynamic power grid operation record data is then collected, such as node voltage, line power flow, switch status, and equipment temperature. Real-time disaster data is extracted from meteorological, water resources, and natural resources departments. Environmental impact data, including wind speed, rainfall, temperature, and geological and hydrological data, is extracted from unstructured documents such as archives, equipment manuals, operation and maintenance procedures, and technical standards. This includes historical disaster event records, equipment fault logs, maintenance reports, and text information. Data cleaning and alignment preprocessing are then performed. Based on the constraints of physical laws and prior experience, a power grid physical and safety mechanism model library and a natural disaster dynamics model library are constructed. The power grid physical and safety mechanism model library includes power flow calculation models, transient stability models, equipment heat capacity models, and tower mechanical strength models. The construction methods for the power grid physical and safety mechanism models are shown in Table 1.

[0078] Table 1

[0079] The natural disaster dynamics model library includes typhoon wind field models, icing growth models, wildfire spread models, and flood evolution models. The construction methods for natural disaster dynamics models are shown in Table 2.

[0080] Table 2

[0081] Natural language processing technology is used to automatically extract entities, such as insulators and typhoons, from historical power grid disaster records and power grid equipment operation and maintenance records, including operation and maintenance documents and fault reports. Entities and their relationships, such as causes and preventive measures, are also extracted. The extracted equipment-fault-disaster-measure data are then associated with cleaned structured data, such as equipment IDs and fault codes, to construct a power grid operation knowledge graph with semantic relationships. Finally, a knowledge-enhanced power grid disaster-coupled digital twin is integrated through a unified data model and service interface. This digital twin is dynamically updated, interactively queried, and capable of simulation and deduction.

[0082] According to an embodiment of the present invention, the step of extracting a disaster event assessment dataset from the multi-source power grid operation record data, and analyzing and processing the disaster event assessment dataset to obtain a virtual disaster event sequence scenario set includes:

[0083] Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, including disaster type characteristic data, disaster intensity, disaster location data, disaster duration and corresponding historical geographical and climatic data;

[0084] Obtain geographic climate prediction data;

[0085] Based on the disaster type characteristic data, disaster intensity, disaster location data, and disaster duration data, as well as the corresponding historical geographic and climate data, and combined with the geographic and climate prediction data, the model is trained to obtain a disaster statistical model and a disaster generation adversarial network model.

[0086] Based on the disaster statistical model, disaster parameters are extracted and combined to generate a disaster event sequence;

[0087] Based on the disaster generation adversarial network model and preset extreme disaster generation conditions, an extreme disaster event sequence is generated;

[0088] The disaster event sequence and the extreme disaster event sequence are fused together to obtain a set of virtual disaster event sequence scenarios.

[0089] Further explanation is needed regarding the following: using key features of historical disasters, geographical and climatic information, and matching information from predictions as training samples, a statistical model that learns the joint distribution law of disaster parameters is trained, and a generative adversarial network that learns the spatiotemporal evolution pattern of disasters is trained. Physical constraints are introduced into the generator input to ensure that the generated scene conforms to physical laws. A disaster event sequence that conforms to historical statistical laws is generated by randomly selecting parameter combinations from the statistical model. The generator AI input generates a scene that is 30% stronger than the strongest disaster in history. After generation, it is checked against physical laws to generate an extreme disaster event sequence, which is then merged into a set of virtual disaster event sequence scenes.

[0090] According to an embodiment of the present invention, the step of acquiring real-time disaster assessment data and analyzing and processing it through a preset disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a power grid operation resilience index set includes:

[0091] Based on the aforementioned knowledge, the power grid disaster-coupled digital twin is enhanced to obtain real-time disaster assessment data, including meteorological remote sensing monitoring data stream, ground meteorological monitoring data stream, power grid operation status data stream, and equipment operation status data stream;

[0092] The meteorological remote sensing monitoring data stream, the ground meteorological monitoring data stream, the power grid operation status data stream, and the equipment operation status data stream are input into a preset disaster intensity time series prediction model for analysis and processing to obtain disaster field prediction data, a list of risky equipment, and physical consistency scores.

[0093] Based on the disaster field prediction data, risk equipment list, and physical consistency score, a physical simulation analysis is performed using a natural disaster dynamics model to obtain the disaster intensity field.

[0094] Based on the disaster intensity field, a coupled extrapolation and analysis is performed using a power grid physics and safety mechanism model to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators.

[0095] Further explanation is needed regarding the process: The encoder extracts features from multi-source data, the decoder predicts disaster fields, a graph neural network branch processes the power grid topology, identifies associated risks of equipment, and supplements the model architecture with physical constraint technology and physical residuals. A preset disaster intensity time-series prediction model is obtained by training based on real-time disaster assessment data and corresponding disaster field data. The data stream from the knowledge-enhanced power grid disaster-coupled digital twin is processed through the model to obtain disaster field prediction data, a list of risky equipment, and a physical consistency score. The disaster field prediction data is a grid map of disaster intensity within a preset future time period. The list of risky equipment is automatically selected based on disaster intensity and equipment vulnerability thresholds. The physical consistency score is obtained by calculating the consistency between the prediction results and the physical equations, with a value range of 0-100. Then, natural... The disaster dynamics model undergoes physical rationality verification and refined simulation to obtain a gridded disaster intensity field, such as the wind load on the base towers. Finally, the disaster intensity field is used to calculate the mechanical response of equipment through the power grid physical and safety mechanism model. Based on the failure criteria, the set of faulty equipment is mapped, and the power grid topology is updated. Power flow calculation and stability analysis are performed to simulate cascading failures, simulate preset scheduling strategies (such as load shedding), and simulate fault repair processes constrained by resources. The system function changes over time are dynamically extrapolated, thereby generating a power grid operation risk assessment report and a power grid operation resilience index set. The power grid operation risk assessment report includes the risk location and risk intensity, and the power grid operation resilience index set includes the system function retention rate curve, cumulative load loss, and recovery time, thus realizing the collaborative processing of the rapid sensing channel and the high-fidelity physical simulation and resilience extrapolation channel.

[0096] According to an embodiment of the present invention, the step of obtaining real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin, and analyzing and processing it in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment includes:

[0097] Acquire the real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin;

[0098] By acquiring dynamic adjustment parameters for disaster prevention, and combining the real-time status data, power grid operation risk assessment report, and power grid operation resilience index set, the disaster intensity field is re-analyzed and processed through the power grid physical and safety mechanism model, and the power grid weak link analysis is performed to obtain a list of high-frequency faults of power grid equipment.

[0099] Further explanation is needed regarding the human-computer interaction and dynamic assessment process. First, real-time status data and disaster prevention dynamic adjustment parameters are acquired. Real-time status data includes power grid topology data, equipment parameters, real-time operating status data, and disaster environment data. Disaster prevention dynamic adjustment parameters include resource repair parameters (number of repair teams, equipment configuration parameters), control power grid parameters (load shedding, protection settings), and strategy scheduling parameters (repair priority, task allocation algorithm). Then, the relevant parameters in the digital twin are updated, and the power grid physical and safety mechanism model (a simplified model, such as using a coarser time step, can be used for rapid re-deduction to achieve real-time interaction) is invoked to generate a new power grid operation risk assessment report and a set of power grid operation resilience indicators. Finally, weak links are diagnosed through two indicators: high-frequency initial fault equipment and systemic resource bottlenecks, resulting in a list of high-frequency faults in power grid equipment.

[0100] According to an embodiment of the present invention, the step of performing matching analysis based on the high-frequency fault list of power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin to obtain a preset number of power grid disaster handling schemes, and performing analysis and processing to obtain optimized values ​​of power grid operational resilience indicators, and determining the optimal power grid disaster handling scheme, includes:

[0101] Based on the high-frequency fault list of the power grid equipment and the power grid operation knowledge graph of the knowledge-enhanced power grid disaster coupled digital twin, a predetermined number of power grid disaster handling solutions are obtained through matching analysis.

[0102] The parameters of the power grid disaster handling scheme are obtained, and the typical disaster event scenario subset is extracted from the virtual disaster event sequence scenario set through cluster analysis. The disaster intensity field is then re-analyzed and processed through the power grid physical and safety mechanism model to obtain the optimized value of the power grid operation resilience index corresponding to the power grid disaster handling scheme.

[0103] The optimized values ​​of the power grid operation resilience index are sorted in descending order, and the power grid disaster handling scheme with the largest optimized value is determined as the optimal power grid disaster handling scheme.

[0104] Further explanation is needed regarding the process of matching weak links with the power grid operation knowledge graph, retrieving associated measures, and generating multiple candidate measures, i.e., power grid disaster management solutions. The knowledge-enhanced power grid disaster-coupled digital twin updates the corresponding solution parameters. The virtual disaster event sequence scenario set undergoes scenario feature extraction and cluster analysis. One or more representative scenarios are selected to form a subset of typical disaster event scenarios, enabling the evaluation of the effectiveness of defensive measures using a limited set of scenarios. Finally, the power grid physical and safety mechanism model is used to re-execute the simulation, quantitatively calculating the optimized value of the power grid operation resilience index corresponding to each power grid disaster management solution. The power grid disaster management solution with the largest optimized value of the power grid operation resilience index is determined as the optimal power grid disaster management solution.

[0105] Please refer to Figure 3 , Figure 3 This is a high-level flowchart of a method for predicting and preventing natural disasters in power grids, as described in some embodiments of this application.

[0106] This invention also discloses a system for predicting and preventing natural disasters in power grids, including a memory and a processor. The memory includes a program for predicting and preventing natural disasters in power grids. When the processor executes the program for predicting and preventing natural disasters in power grids, it performs the following steps:

[0107] Acquire multi-source power grid operation record data, and construct and process digital twins based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin;

[0108] Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, and the disaster event assessment dataset is analyzed and processed to obtain a set of virtual disaster event sequence scenarios.

[0109] Real-time disaster assessment data is acquired and analyzed using a pre-set disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators.

[0110] The real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin is obtained, and analyzed and processed in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment.

[0111] Based on the high-frequency fault list of power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin, a matching analysis is performed to obtain a preset number of power grid disaster handling schemes. These schemes are then analyzed and processed to obtain optimized values ​​for power grid operational resilience indicators and to determine the optimal power grid disaster handling scheme.

[0112] Further explanation is needed regarding the implementation architecture for accurately predicting and effectively defending against natural disasters. This architecture incorporates a hybrid physical information neural network, adversarial scenario generation, knowledge graph enhancement, and human-computer interactive simulation. First, by fusing static data, dynamic data, physical mechanism models, and knowledge graphs, a knowledge-enhanced digital twin of the power grid disaster is constructed. Then, a set of virtual disaster event sequence scenarios, including historical and extreme disasters, is generated and subjected to dual-channel collaborative early warning analysis. Next, human-computer interactive dynamic resilience simulation and weak node assessment are achieved through manual dynamic adjustment of scheme parameters. Finally, defense measures are quantitatively evaluated based on typical disaster scenarios to achieve accurate selection of defense methods.

[0113] According to an embodiment of the present invention, the step of acquiring multi-source power grid operation record data, and performing digital twin construction processing based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin includes:

[0114] Acquire multi-source power grid operation record data, including static power grid parameter data, dynamic power grid operation record data, real-time disaster and environmental impact data, historical power grid disaster record data, and power grid equipment operation and maintenance record data;

[0115] A power grid physical and safety mechanism model library is constructed based on the static power grid parameter data, dynamic power grid operation record data, and historical power grid disaster record data.

[0116] A natural disaster dynamics model library is constructed based on the real-time disaster environmental impact data and historical power grid disaster record data.

[0117] Based on the historical power grid disaster record data and power grid equipment operation and maintenance record data, text parsing, entity recognition, relationship extraction and data fusion processing are performed to obtain a power grid operation knowledge graph;

[0118] The multi-source power grid operation record data, power grid physical and safety mechanism model library, natural disaster dynamics model library, and power grid operation knowledge graph are integrated through data models and service interfaces to obtain a knowledge-enhanced power grid disaster-coupled digital twin.

[0119] Further explanation is needed regarding the construction of a knowledge-enhanced power grid disaster-coupled digital twin that integrates multi-source information and possesses knowledge reasoning capabilities, serving as a dynamic virtual power grid mirror. First, static power grid parameter data is acquired from the enterprise asset management system, including power grid geographic information system data and equipment building information model data. The power grid geographic information system data includes line routes, tower coordinates, and substation locations; the equipment building information model data includes equipment 3D models, materials, and design parameters. Dynamic power grid operation record data is collected, such as node voltage, line power flow, switch status, and equipment temperature. Real-time disaster environmental impact data, including wind speed, rainfall, temperature, and geological and hydrological data, is extracted from meteorological, water resources, and natural resources departments. Power grid equipment operation and maintenance record data, such as historical disaster event records, equipment fault logs, maintenance reports, and text information, are extracted from archives and unstructured documents such as equipment manuals, operation and maintenance procedures, and technical standards. Data cleaning and alignment preprocessing are then performed. Finally, a power grid is constructed based on the constraints of physical laws and prior experience. The system comprises two libraries: a physical and safety mechanism model library and a natural disaster dynamics model library. The physical and safety mechanism model library includes power flow calculation models, transient stability models, equipment thermal capacity models, and tower mechanical strength models. The construction methods for these models are shown in Table 1. The natural disaster dynamics model library includes typhoon wind field models, icing growth models, wildfire spread models, and flood evolution models. The construction methods for these models are shown in Table 2. Natural language processing technology is used to automatically extract entities, such as insulators and typhoons, from historical power grid disaster records and power grid equipment operation and maintenance records, including operation and maintenance documents and fault reports. Entities and their relationships, such as causes and preventive measures, are extracted. The extracted equipment-fault-disaster-measure data are then associated with cleaned structured data, such as equipment IDs and fault codes, to construct a power grid operation knowledge graph with semantic relationships. Finally, a dynamically updated, interactively queried, and simulated knowledge-enhanced power grid disaster-coupled digital twin is integrated through a unified data model and service interface.

[0120] According to an embodiment of the present invention, the step of extracting a disaster event assessment dataset from the multi-source power grid operation record data, and analyzing and processing the disaster event assessment dataset to obtain a virtual disaster event sequence scenario set includes:

[0121] Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, including disaster type characteristic data, disaster intensity, disaster location data, disaster duration and corresponding historical geographical and climatic data;

[0122] Obtain geographic climate prediction data;

[0123] Based on the disaster type characteristic data, disaster intensity, disaster location data, and disaster duration data, as well as the corresponding historical geographic and climate data, and combined with the geographic and climate prediction data, the model is trained to obtain a disaster statistical model and a disaster generation adversarial network model.

[0124] Based on the disaster statistical model, disaster parameters are extracted and combined to generate a disaster event sequence;

[0125] Based on the disaster generation adversarial network model and preset extreme disaster generation conditions, an extreme disaster event sequence is generated;

[0126] The disaster event sequence and the extreme disaster event sequence are fused together to obtain a set of virtual disaster event sequence scenarios.

[0127] Further explanation is needed regarding the following: using key features of historical disasters, geographical and climatic information, and matching information from predictions as training samples, a statistical model that learns the joint distribution law of disaster parameters is trained, and a generative adversarial network that learns the spatiotemporal evolution pattern of disasters is trained. Physical constraints are introduced into the generator input to ensure that the generated scene conforms to physical laws. A disaster event sequence that conforms to historical statistical laws is generated by randomly selecting parameter combinations from the statistical model. The generator AI input generates a scene that is 30% stronger than the strongest disaster in history. After generation, it is checked against physical laws to generate an extreme disaster event sequence, which is then merged into a set of virtual disaster event sequence scenes.

[0128] According to an embodiment of the present invention, the step of acquiring real-time disaster assessment data and analyzing and processing it through a preset disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a power grid operation resilience index set includes:

[0129] Based on the aforementioned knowledge, the power grid disaster-coupled digital twin is enhanced to obtain real-time disaster assessment data, including meteorological remote sensing monitoring data stream, ground meteorological monitoring data stream, power grid operation status data stream, and equipment operation status data stream;

[0130] The meteorological remote sensing monitoring data stream, the ground meteorological monitoring data stream, the power grid operation status data stream, and the equipment operation status data stream are input into a preset disaster intensity time series prediction model for analysis and processing to obtain disaster field prediction data, a list of risky equipment, and physical consistency scores.

[0131] Based on the disaster field prediction data, risk equipment list, and physical consistency score, a physical simulation analysis is performed using a natural disaster dynamics model to obtain the disaster intensity field.

[0132] Based on the disaster intensity field, a coupled extrapolation and analysis is performed using a power grid physics and safety mechanism model to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators.

[0133] Further explanation is needed regarding the process: The encoder extracts features from multi-source data, the decoder predicts disaster fields, a graph neural network branch processes the power grid topology, identifies associated risks of equipment, and supplements the model architecture with physical constraint technology and physical residuals. A preset disaster intensity time-series prediction model is obtained by training based on real-time disaster assessment data and corresponding disaster field data. The data stream from the knowledge-enhanced power grid disaster-coupled digital twin is processed through the model to obtain disaster field prediction data, a list of risky equipment, and a physical consistency score. The disaster field prediction data is a grid map of disaster intensity within a preset future time period. The list of risky equipment is automatically selected based on disaster intensity and equipment vulnerability thresholds. The physical consistency score is obtained by calculating the consistency between the prediction results and the physical equations, with a value range of 0-100. Then, natural... The disaster dynamics model undergoes physical rationality verification and refined simulation to obtain a gridded disaster intensity field, such as the wind load on the base towers. Finally, the disaster intensity field is used to calculate the mechanical response of equipment through the power grid physical and safety mechanism model. Based on the failure criteria, the set of faulty equipment is mapped, and the power grid topology is updated. Power flow calculation and stability analysis are performed to simulate cascading failures, simulate preset scheduling strategies (such as load shedding), and simulate fault repair processes constrained by resources. The system function changes over time are dynamically extrapolated, thereby generating a power grid operation risk assessment report and a power grid operation resilience index set. The power grid operation risk assessment report includes the risk location and risk intensity, and the power grid operation resilience index set includes the system function retention rate curve, cumulative load loss, and recovery time, thus realizing the collaborative processing of the rapid sensing channel and the high-fidelity physical simulation and resilience extrapolation channel.

[0134] According to an embodiment of the present invention, the step of obtaining real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin, and analyzing and processing it in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment includes:

[0135] Acquire the real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin;

[0136] By acquiring dynamic adjustment parameters for disaster prevention, and combining the real-time status data, power grid operation risk assessment report, and power grid operation resilience index set, the disaster intensity field is re-analyzed and processed through the power grid physical and safety mechanism model, and the power grid weak link analysis is performed to obtain a list of high-frequency faults of power grid equipment.

[0137] Further explanation is needed regarding the human-computer interaction and dynamic assessment process. First, real-time status data and disaster prevention dynamic adjustment parameters are acquired. Real-time status data includes power grid topology data, equipment parameters, real-time operating status data, and disaster environment data. Disaster prevention dynamic adjustment parameters include resource repair parameters (number of repair teams, equipment configuration parameters), control power grid parameters (load shedding, protection settings), and strategy scheduling parameters (repair priority, task allocation algorithm). Then, the relevant parameters in the digital twin are updated, and the power grid physical and safety mechanism model (a simplified model, such as using a coarser time step, can be used for rapid re-deduction to achieve real-time interaction) is invoked to generate a new power grid operation risk assessment report and a set of power grid operation resilience indicators. Finally, weak links are diagnosed through two indicators: high-frequency initial fault equipment and systemic resource bottlenecks, resulting in a list of high-frequency faults in power grid equipment.

[0138] According to an embodiment of the present invention, the step of performing matching analysis based on the high-frequency fault list of power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin to obtain a preset number of power grid disaster handling schemes, and performing analysis and processing to obtain optimized values ​​of power grid operational resilience indicators, and determining the optimal power grid disaster handling scheme, includes:

[0139] Based on the high-frequency fault list of the power grid equipment and the power grid operation knowledge graph of the knowledge-enhanced power grid disaster coupled digital twin, a predetermined number of power grid disaster handling solutions are obtained through matching analysis.

[0140] The parameters of the power grid disaster handling scheme are obtained, and the typical disaster event scenario subset is extracted from the virtual disaster event sequence scenario set through cluster analysis. The disaster intensity field is then re-analyzed and processed through the power grid physical and safety mechanism model to obtain the optimized value of the power grid operation resilience index corresponding to the power grid disaster handling scheme.

[0141] The optimized values ​​of the power grid operation resilience index are sorted in descending order, and the power grid disaster handling scheme with the largest optimized value is determined as the optimal power grid disaster handling scheme.

[0142] Further explanation is needed regarding the process of matching weak links with the power grid operation knowledge graph, retrieving associated measures, and generating multiple candidate measures, i.e., power grid disaster management solutions. The knowledge-enhanced power grid disaster-coupled digital twin updates the corresponding solution parameters. The virtual disaster event sequence scenario set undergoes scenario feature extraction and cluster analysis. One or more representative scenarios are selected to form a subset of typical disaster event scenarios, enabling the evaluation of the effectiveness of defensive measures using a limited set of scenarios. Finally, the power grid physical and safety mechanism model is used to re-execute the simulation, quantitatively calculating the optimized value of the power grid operation resilience index corresponding to each power grid disaster management solution. The power grid disaster management solution with the largest optimized value of the power grid operation resilience index is determined as the optimal power grid disaster management solution.

[0143] A third aspect of the present invention provides a readable storage medium storing a method program for predicting and preventing natural disasters in a power grid. When the method program is executed by a processor, it implements the steps of the method for predicting and preventing natural disasters in a power grid as described in any of the preceding claims.

[0144] This invention discloses a method, system, and medium for predicting and defending against natural disasters in power grids. By constructing a digital twin, generating a disaster scenario set, performing dual-channel collaborative early warning simulations, and conducting human-computer interactive resilience assessments, it enables the prediction and defense against natural disasters in power grids.

[0145] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or components can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, and the indirect coupling or communication connection between devices or units can be electrical, mechanical, or other forms.

[0146] The units described above as separate components may or may not be physically separate. The components shown as units may or may not be physical units. They may be located in one place or distributed across multiple network units. Some or all of the units may be selected to achieve the purpose of this embodiment according to actual needs.

[0147] In addition, in the various embodiments of the present invention, each functional unit can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be implemented in hardware or in the form of hardware plus software functional units.

[0148] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0149] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, can be embodied in the form of a software product. This software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

Claims

1. A method for predicting and defending against natural disasters of power grid, characterized in that, Includes the following steps: Acquire multi-source power grid operation record data, and construct and process digital twins based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin; Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, and the disaster event assessment dataset is analyzed and processed to obtain a set of virtual disaster event sequence scenarios. Real-time disaster assessment data is acquired and analyzed using a preset disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators. The real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin is obtained, and analyzed and processed in combination with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment. Based on the high-frequency fault list of power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin, a matching analysis is performed to obtain a preset number of power grid disaster handling schemes. These schemes are then analyzed and processed to obtain optimized values ​​for power grid operational resilience indicators and to determine the optimal power grid disaster handling scheme.

2. The power grid natural disaster prediction and defense method of claim 1, wherein, The process of acquiring multi-source power grid operation record data, constructing a digital twin based on the multi-source power grid operation record data, and obtaining a knowledge-enhanced power grid disaster-coupled digital twin includes: Acquire multi-source power grid operation record data, including static power grid parameter data, dynamic power grid operation record data, real-time disaster and environmental impact data, historical power grid disaster record data, and power grid equipment operation and maintenance record data; A power grid physical and safety mechanism model library is constructed based on the static power grid parameter data, dynamic power grid operation record data, and historical power grid disaster record data. A natural disaster dynamics model library is constructed based on the real-time disaster environmental impact data and historical power grid disaster record data. Based on the historical power grid disaster record data and power grid equipment operation and maintenance record data, text parsing, entity recognition, relationship extraction and data fusion processing are performed to obtain a power grid operation knowledge graph; The multi-source power grid operation record data, power grid physical and safety mechanism model library, natural disaster dynamics model library, and power grid operation knowledge graph are integrated through data models and service interfaces to obtain a knowledge-enhanced power grid disaster-coupled digital twin.

3. The power grid natural disaster prediction and defense method of claim 2, wherein, The step involves extracting a disaster event assessment dataset from the multi-source power grid operation record data, analyzing and processing the disaster event assessment dataset to obtain a set of virtual disaster event sequence scenarios, including: Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, including disaster type characteristic data, disaster intensity, disaster location data, disaster duration and corresponding historical geographical and climatic data; Obtain geographic climate prediction data; Based on the disaster type characteristic data, disaster intensity, disaster location data, and disaster duration data, as well as the corresponding historical geographic and climate data, and combined with the geographic and climate prediction data, the model is trained to obtain a disaster statistical model and a disaster generation adversarial network model. Based on the disaster statistical model, disaster parameters are extracted and combined to generate a disaster event sequence; Based on the disaster generation adversarial network model and preset extreme disaster generation conditions, an extreme disaster event sequence is generated; The disaster event sequence and the extreme disaster event sequence are fused together to obtain a set of virtual disaster event sequence scenarios.

4. The method for predicting and preventing natural disasters in power grids according to claim 3, characterized in that, The process involves acquiring real-time disaster assessment data and analyzing it using a preset disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a power grid operation resilience index set, including: Based on the aforementioned knowledge, the power grid disaster-coupled digital twin is enhanced to obtain real-time disaster assessment data, including meteorological remote sensing monitoring data stream, ground meteorological monitoring data stream, power grid operation status data stream, and equipment operation status data stream; The meteorological remote sensing monitoring data stream, the ground meteorological monitoring data stream, the power grid operation status data stream, and the equipment operation status data stream are input into a preset disaster intensity time series prediction model for analysis and processing to obtain disaster field prediction data, a list of risky equipment, and physical consistency scores. Based on the disaster field prediction data, risk equipment list, and physical consistency score, a physical simulation analysis is performed using a natural disaster dynamics model to obtain the disaster intensity field. Based on the disaster intensity field, a coupled extrapolation and analysis is performed using a power grid physics and safety mechanism model to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators.

5. The method for predicting and preventing natural disasters in power grids according to claim 4, characterized in that, The process involves acquiring real-time status data from the knowledge-enhanced power grid disaster-coupled digital twin, analyzing and processing this data in conjunction with the power grid operation risk assessment report and the power grid operation resilience index set, to obtain a list of high-frequency faults in power grid equipment, including: Acquire the real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin; By acquiring dynamic adjustment parameters for disaster prevention, and combining the real-time status data, power grid operation risk assessment report, and power grid operation resilience index set, the disaster intensity field is re-analyzed and processed through the power grid physical and safety mechanism model, and the power grid weak link analysis is performed to obtain a list of high-frequency faults of power grid equipment.

6. The method for predicting and preventing natural disasters in power grids according to claim 5, characterized in that, The process involves matching and analyzing the high-frequency fault list of power grid equipment with the knowledge-enhanced power grid disaster-coupled digital twin to obtain a preset number of power grid disaster handling solutions, analyzing and processing these solutions to obtain optimized values ​​for power grid operational resilience indicators, and determining the optimal power grid disaster handling solution. This includes: Based on the high-frequency fault list of the power grid equipment and the power grid operation knowledge graph of the knowledge-enhanced power grid disaster coupled digital twin, a predetermined number of power grid disaster handling solutions are obtained through matching analysis. The parameters of the power grid disaster handling scheme are obtained, and the typical disaster event scenario subset is extracted from the virtual disaster event sequence scenario set through cluster analysis. The disaster intensity field is then re-analyzed and processed through the power grid physical and safety mechanism model to obtain the optimized value of the power grid operation resilience index corresponding to the power grid disaster handling scheme. The optimized values ​​of the power grid operation resilience index are sorted in descending order, and the power grid disaster handling scheme with the largest optimized value is determined as the optimal power grid disaster handling scheme.

7. A power grid natural disaster prediction and prevention system, characterized in that, The system includes a memory and a processor. The memory contains a program for predicting and mitigating natural disasters affecting the power grid. When executed by the processor, the program for predicting and mitigating natural disasters affecting the power grid implements the following steps: Acquire multi-source power grid operation record data, and construct and process digital twins based on the multi-source power grid operation record data to obtain a knowledge-enhanced power grid disaster-coupled digital twin; Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, and the disaster event assessment dataset is analyzed and processed to obtain a set of virtual disaster event sequence scenarios. Real-time disaster assessment data is acquired and analyzed using a preset disaster intensity time-series prediction model and a knowledge-enhanced power grid disaster-coupled digital twin to obtain a power grid operation risk assessment report and a set of power grid operation resilience indicators. The real-time status data of the knowledge-enhanced power grid disaster-coupled digital twin is obtained, and analyzed and processed in combination with the power grid operation risk assessment report and the power grid operation resilience index set to obtain a list of high-frequency faults of power grid equipment. Based on the high-frequency fault list of power grid equipment and the knowledge-enhanced power grid disaster coupled digital twin, a matching analysis is performed to obtain a preset number of power grid disaster handling schemes. These schemes are then analyzed and processed to obtain optimized values ​​for power grid operational resilience indicators and to determine the optimal power grid disaster handling scheme.

8. The power grid natural disaster prediction and prevention system according to claim 7, characterized in that, The process of acquiring multi-source power grid operation record data, constructing a digital twin based on the multi-source power grid operation record data, and obtaining a knowledge-enhanced power grid disaster-coupled digital twin includes: Acquire multi-source power grid operation record data, including static power grid parameter data, dynamic power grid operation record data, real-time disaster and environmental impact data, historical power grid disaster record data, and power grid equipment operation and maintenance record data; Based on the static power grid parameter data, dynamic power grid operation record data, and historical power grid disaster record data, a power grid physical and safety mechanism model library is constructed, including power flow calculation models, transient stability models, equipment heat capacity models, and tower mechanical strength models; A natural disaster dynamics model library is constructed based on the real-time disaster environmental impact data and historical power grid disaster record data, including typhoon wind field model, icing growth model, wildfire spread model and flood evolution model; Based on the historical power grid disaster record data and power grid equipment operation and maintenance record data, text parsing, entity recognition, relationship extraction and data fusion processing are performed to obtain a power grid operation knowledge graph; The multi-source power grid operation record data, power grid physical and safety mechanism model library, natural disaster dynamics model library, and power grid operation knowledge graph are integrated through data models and service interfaces to obtain a knowledge-enhanced power grid disaster-coupled digital twin.

9. The power grid natural disaster prediction and prevention system according to claim 8, characterized in that, The step involves extracting a disaster event assessment dataset from the multi-source power grid operation record data, analyzing and processing the disaster event assessment dataset to obtain a set of virtual disaster event sequence scenarios, including: Based on the multi-source power grid operation record data, a disaster event assessment dataset is extracted, including disaster type characteristic data, disaster intensity, disaster location data, disaster duration and corresponding historical geographical and climatic data; Obtain geographic climate prediction data; Based on the disaster type characteristic data, disaster intensity, disaster location data, and disaster duration data, as well as the corresponding historical geographic and climate data, and combined with the geographic and climate prediction data, the model is trained to obtain a disaster statistical model and a disaster generation adversarial network model. Based on the disaster statistical model, disaster parameters are extracted and combined to generate a disaster event sequence; Based on the disaster generation adversarial network model and preset extreme disaster generation conditions, an extreme disaster event sequence is generated; The disaster event sequence and the extreme disaster event sequence are fused together to obtain a set of virtual disaster event sequence scenarios.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a method program for predicting and preventing natural disasters in the power grid. When the method program is executed by a processor, it implements the steps of a method for predicting and preventing natural disasters in the power grid as described in any one of claims 1 to 6.

Images