Power grid risk assessment method, device, equipment and medium

Abstract

Embodiments of the present application provide a power grid risk assessment method, device, equipment and medium. The method comprises: acquiring multi-source operation data of a power grid, the multi-source operation data at least comprising historical load data, real-time load data, meteorological environment data and regional economic data; then performing standardization processing on the multi-source operation data to obtain a standardized load data set; then, based on the standardized load data set, generating a multi-level fault scenario by using a fault propagation model; finally, according to the multi-level fault scenario, performing quantitative assessment on the operation risk of the power grid under each fault scenario according to a preset multi-dimensional risk assessment system to determine a corresponding risk level. Through the above method, accurate assessment and judgment of the operation risk of the power grid are realized.

Description

Technical Field

[0001] This application relates to the field of power grid risk assessment technology, and in particular to a power grid risk assessment method, apparatus, equipment and medium. Background Technology

[0002] With the global energy structure transformation and the continuous growth of electricity demand, modern power grids are facing unprecedented complexity and dynamic challenges. In practical applications, power grid operation needs to simultaneously cope with dynamic load fluctuations (such as sudden load surges caused by high summer temperatures), complex fault coupling effects (such as cascading faults caused by the complete shutdown of lines on the same tower), multi-dimensional risk indicators (such as voltage stability and equipment lifespan loss), and the need for rapid optimization decisions (such as restoring power supply within 10 minutes after a fault). Existing technologies lack the ability to dynamically integrate meteorological data (temperature, humidity) and economic data (GDP growth), resulting in load forecasting errors as high as 28%. At the same time, single-dimensional risk assessment (such as focusing only on load loss) cannot fully reflect the vulnerability of the power grid. Therefore, there is an urgent need for a multi-dimensional risk assessment method to achieve accurate assessment and efficient response to power grid risks, supporting the safe and stable operation of new power systems.

[0003] In existing technologies, the commonly used risk assessment method is to establish a steady-state model of the power grid based on electromagnetic transient simulation tools; then disconnect equipment at voltage levels of 110kV and above in sequence to simulate faults; after each fault, overload judgment is made based on whether the main transformer load rate exceeds 130% or whether the line current exceeds the fault current limit value (such as 1.2 times the rated current of a 220kV line); finally, indicators such as the proportion of risky loads (for example, the risky load after N-1 of the main transformer in a certain substation is 14.72MW) are statistically analyzed to complete the risk assessment.

[0004] However, existing solutions suffer from incomplete coverage of complex fault scenarios and a single dimension of risk assessment. Summary of the Invention

[0005] This application provides a power grid risk assessment method, apparatus, equipment, and medium to address the problems of incomplete coverage of complex fault scenarios and single risk assessment dimensions in the prior art.

[0006] In a first aspect, embodiments of this application provide a power grid risk assessment method, including:

[0007] Acquire multi-source operation data of the power grid, including at least historical load data, real-time load data, meteorological and environmental data, and regional economic data;

[0008] The multi-source operational data is standardized to obtain a standardized load dataset;

[0009] Based on the standardized load dataset, a multi-level fault scenario is generated using a fault propagation model.

[0010] Based on the aforementioned multi-level fault scenarios, and in accordance with a pre-set multi-dimensional risk assessment system, the operational risks of the power grid under each fault scenario are quantitatively assessed, and the corresponding risk levels are determined.

[0011] By employing the aforementioned technical means, it is possible to comprehensively utilize heterogeneous data from multiple sources, such as historical load, real-time load, meteorological environment, and regional economy, to fully reflect the power grid's operating status and its external influencing factors. Based on this, a more accurate basic data system for fault scenario analysis and risk assessment can be constructed, thereby improving the comprehensiveness and reliability of power grid risk identification, avoiding the one-sided or distorted risk assessment results caused by relying solely on single operating data, and enhancing the accuracy of power grid operation risk early warning and decision support capabilities.

[0012] In one possible implementation, the standardization process of the multi-source operational data to obtain a standardized load dataset includes:

[0013] The multi-source operational data is normalized to obtain the first dataset;

[0014] The missing data in the first dataset is imputed to obtain the second dataset;

[0015] Abnormal load data in the second dataset are identified and removed based on statistical criteria to obtain the standardized load dataset.

[0016] By employing the aforementioned technical means, the impact of differences in units and value ranges between data from different sources on the analysis results can be eliminated. At the same time, missing data can be repaired and abnormal data can be removed, thereby improving the consistency, completeness, and effectiveness of input data. This provides high-quality data support for subsequent load forecasting, fault scenario generation, and risk assessment, reduces the interference of noisy data on the assessment results, and improves the stability and credibility of risk analysis results.

[0017] In one possible implementation, after obtaining the standardized load dataset, the method further includes:

[0018] The standardized load dataset is input into the trained load prediction model to dynamically predict the power grid load. The load prediction model is an LSTM neural network model trained based on historical multi-source operating data and corresponding historical loads.

[0019] The predicted load data is used as the load boundary condition for generating the multi-level fault scenarios and conducting risk assessments.

[0020] By employing the aforementioned technical means, we can fully explore the temporal variation patterns in historical operational data, achieve dynamic prediction of future power grid load change trends, and use the prediction results as boundary conditions for fault analysis and risk assessment. This allows risk assessment to not only reflect the current operating status but also the potential risks brought about by future load fluctuations, thereby improving the foresight and predictability of risk assessment and providing a more accurate basis for power grid planning and operation decisions.

[0021] In one possible implementation, generating multi-level fault scenarios based on the standardized load dataset using a fault propagation model includes:

[0022] Based on the relationship between power grid topology and equipment, a Petri net fault propagation model is constructed;

[0023] A stratified sampling method is used to generate multi-level fault scenarios, including single-device faults, dual-device faults, and complex coupling faults, through the Petri net fault propagation model.

[0024] Through the above-mentioned technical means, it is possible to simulate the propagation process of faults in the power grid based on the power grid topology and the interrelationship between equipment, and generate multi-level fault scenarios covering single equipment faults, multi-equipment faults and complex coupled faults. This allows the risk analysis scope to cover more actual operating conditions and extreme fault situations, thereby improving the comprehensiveness and authenticity of risk assessment and facilitating the early identification of potential weak links and key risk nodes.

[0025] In one possible implementation, the step of quantitatively assessing the operational risk of the power grid under each fault scenario according to a preset multi-dimensional risk assessment system, and determining the corresponding risk level, includes:

[0026] For each fault scenario, transient simulation data of the power grid is obtained, including node voltage, line current and transformer load rate data.

[0027] Based on the simulation data and the predicted load data, calculate the load risk parameter, voltage risk parameter, and equipment life risk parameter based on the equipment life prediction model.

[0028] According to the preset weighting coefficients, the load risk parameters, voltage risk parameters and equipment life risk parameters are weighted and summed to obtain the comprehensive risk value of power grid operation risk;

[0029] The overall risk value is compared with a preset threshold range to determine the risk level.

[0030] By employing the aforementioned technical means, it is possible to comprehensively consider the influencing factors of multiple dimensions, such as load risk, voltage risk, and equipment lifespan risk, and to quantitatively evaluate the power grid operation status through a unified comprehensive risk index. This enables a coordinated assessment of power grid safety, stability, and equipment health status, avoiding assessment biases caused by single risk indicators, improving the scientificity and accuracy of risk level determination, and providing a reliable basis for risk warning and operation and maintenance decisions.

[0031] In one possible implementation, after determining the risk level, the method further includes:

[0032] Based on the risk level results, power grid topology data, and predicted load data, a target optimization model is constructed with the objective of minimizing the sum of power grid construction costs and power outage losses, and constrained by the N-1 safety criterion and equipment operating limits.

[0033] The objective optimization model is solved using an objective optimization algorithm to obtain at least one power grid operation optimization scheme, which includes a power grid structure adjustment scheme and / or an operation mode adjustment scheme.

[0034] Through the above-mentioned technical means, an optimized decision-making model that takes into account both construction costs and power outage losses can be automatically constructed based on the risk assessment results. Under the premise of meeting the constraints of power grid safe operation, the optimal operation plan can be generated, thereby optimizing the allocation of power grid resources and operation mode, reducing power grid operation risks and economic losses, and improving the reliability of power supply, operation economy and overall operation efficiency of the power grid.

[0035] In one possible implementation, the method further includes:

[0036] The effectiveness of the power grid operation optimization scheme was verified through digital twin simulation.

[0037] If the verification passes, the optimization scheme will be output and executed.

[0038] If the verification fails, adjust the optimization objective or constraints and solve again until a successful optimization solution is obtained.

[0039] Through the aforementioned technical means, the implementation effect of the optimized scheme can be simulated and verified in a digital twin environment before the actual implementation of the scheme. Potential problems and unreasonable configurations can be identified in advance, avoiding the operational risks caused by applying unverified schemes directly to the actual power grid. At the same time, the optimization objectives and constraints can be continuously corrected through iterative optimization mechanisms, improving the feasibility, accuracy and success rate of the final scheme, and enhancing the safety and reliability of power grid operation decisions.

[0040] Secondly, embodiments of this application provide a power grid risk assessment device, comprising:

[0041] The acquisition module is used to acquire multi-source operation data of the power grid, which includes at least historical load data, real-time load data, meteorological environment data, and regional economic data.

[0042] The processing module is used to standardize the multi-source operating data to obtain a standardized load dataset;

[0043] The generation module is used to generate multi-level fault scenarios based on the standardized load dataset and using a fault propagation model.

[0044] The determination module is used to quantitatively assess the operational risks of the power grid under each fault scenario according to the multi-level fault scenarios and a preset multi-dimensional risk assessment system, and to determine the corresponding risk level.

[0045] In one possible implementation, the processing module is specifically used for:

[0046] The multi-source operational data is normalized to obtain the first dataset;

[0047] The missing data in the first dataset is imputed to obtain the second dataset;

[0048] Abnormal load data in the second dataset are identified and removed based on statistical criteria to obtain the standardized load dataset.

[0049] In one possible implementation, after obtaining the standardized load dataset, the processing module is further configured to:

[0050] The standardized load dataset is input into the trained load prediction model to dynamically predict the power grid load. The load prediction model is an LSTM neural network model trained based on historical multi-source operating data and corresponding historical loads.

[0051] The predicted load data is used as the load boundary condition for generating the multi-level fault scenarios and conducting risk assessments.

[0052] In one possible implementation, the generation module is specifically used for:

[0053] Based on the relationship between power grid topology and equipment, a Petri net fault propagation model is constructed;

[0054] A stratified sampling method is used to generate multi-level fault scenarios, including single-device faults, dual-device faults, and complex coupling faults, through the Petri net fault propagation model.

[0055] In one possible implementation, the determining module is specifically used for:

[0056] For each fault scenario, transient simulation data of the power grid is obtained, including node voltage, line current and transformer load rate data.

[0057] Based on the simulation data and the predicted load data, calculate the load risk parameter, voltage risk parameter, and equipment life risk parameter based on the equipment life prediction model.

[0058] According to the preset weighting coefficients, the load risk parameters, voltage risk parameters and equipment life risk parameters are weighted and summed to obtain the comprehensive risk value of power grid operation risk;

[0059] The overall risk value is compared with a preset threshold range to determine the risk level.

[0060] In one possible implementation, the power grid risk assessment device further includes an optimization module, which, after determining the risk level, is configured to:

[0061] Based on the risk level results, power grid topology data, and predicted load data, a target optimization model is constructed with the objective of minimizing the sum of power grid construction costs and power outage losses, and constrained by the N-1 safety criterion and equipment operating limits.

[0062] The objective optimization model is solved using an objective optimization algorithm to obtain at least one power grid operation optimization scheme, which includes a power grid structure adjustment scheme and / or an operation mode adjustment scheme.

[0063] In one possible implementation, the power grid risk assessment device further includes an optimization module for:

[0064] The effectiveness of the power grid operation optimization scheme was verified through digital twin simulation.

[0065] If the verification passes, the optimization scheme will be output and executed.

[0066] If the verification fails, adjust the optimization objective or constraints and solve again until a successful optimization solution is obtained.

[0067] Thirdly, embodiments of this application provide an electronic device, including: a memory and a processor;

[0068] The memory stores computer-executed instructions;

[0069] The processor executes computer execution instructions stored in the memory, causing the processor to perform the first aspect and / or various possible implementations of the first aspect as described above.

[0070] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the first aspect and / or various possible implementations of the first aspect.

[0071] Fifthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the first aspect and / or various possible implementations of the first aspect.

[0072] The power grid risk assessment method, apparatus, equipment, and medium provided in this application first construct a comprehensive and realistic representation of the power grid's operational status by collecting multi-source operational data, thereby achieving comprehensive coverage of internal and external influencing factors at the data level. Subsequently, the multi-source operational data undergoes unified standardization processing. By eliminating dimensional differences, scaling the data distribution range, and suppressing the interference of outliers on the calculation results, a standardized load dataset that can be directly used for modeling and analysis is formed, improving the stability and comparability of subsequent model calculations. Based on this, a fault propagation model is constructed and introduced based on the standardized load dataset, characterizing the power grid topology, electrical coupling relationships, and fault propagation. The causal transmission mechanism between faults is simulated and extrapolated to simulate the chain reaction of faults that may be triggered by a single point of failure under different operating conditions. This generates multi-level fault scenarios covering different propagation paths and impact ranges, enabling a systematic characterization of complex fault evolution processes. Finally, for the generated multi-level fault scenarios, based on a pre-constructed multi-dimensional risk assessment system, the power grid operation risk is quantitatively calculated from multiple dimensions such as power supply reliability, load loss, impact on key equipment, and fault propagation range. The risk level corresponding to each fault scenario is determined through comprehensive assessment results, thereby achieving a refined and quantifiable assessment of the power grid's operation risk under different potential fault conditions, improving the power grid's risk prediction capability and the scientific nature of operation decisions. Attached Figure Description

[0073] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0074] Figure 1 A flowchart illustrating the power grid risk assessment method provided in this application embodiment. Figure 1 ;

[0075] Figure 2A flowchart illustrating the power grid risk assessment method provided in this application embodiment. Figure 2 ;

[0076] Figure 3 This is a schematic diagram of the structure of the power grid risk assessment device provided in the embodiments of this application;

[0077] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0078] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0079] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0080] With the global energy structure transformation and the continuous growth of electricity demand, modern power grids are facing unprecedented complexity and dynamic challenges. In practical applications, power grid operation needs to simultaneously cope with dynamic load fluctuations (such as sudden load surges caused by high summer temperatures), complex fault coupling effects (such as cascading faults caused by the complete shutdown of lines on the same tower), multi-dimensional risk indicators (such as voltage stability and equipment lifespan loss), and the need for rapid optimization decisions (such as restoring power supply within 10 minutes after a fault). Existing technologies lack the ability to dynamically integrate meteorological data (temperature, humidity) and economic data (GDP growth), resulting in load forecasting errors as high as 28%. At the same time, single-dimensional risk assessment (such as focusing only on load loss) cannot fully reflect the vulnerability of the power grid. Therefore, there is an urgent need for a multi-dimensional risk assessment method to achieve accurate assessment and efficient response to power grid risks, supporting the safe and stable operation of new power systems.

[0081] In existing technologies, commonly used risk assessment methods first establish a steady-state model of the Dongguan power grid, including a three-level grid structure of 500kV / 220kV / 110kV, based on electromagnetic transient simulation tools; then, they simulate 246 single faults in equipment at voltage levels of 110kV and above; after each fault, overload judgment is made based on whether the main transformer load rate exceeds 130% or whether the line current exceeds the fault current limit value (such as 1.2 times the rated current of a 220kV line); finally, indicators such as the risk load ratio (for example, the risk load after N-1 of the main transformer in a certain substation is 14.72MW) are statistically analyzed to complete the risk assessment.

[0082] However, existing solutions only simulate single faults in 110kV and above equipment (a total of 246 times), failing to cover complex fault scenarios such as multiple faults, cascading faults, or severe weather. This results in insufficient assessment of higher-order risks that may occur in actual power grid operation. At the same time, the risk assessment is based solely on overload judgment when the main transformer load rate exceeds 130% or the line current exceeds the fault current limit, and calculates the risk load ratio as a single-dimensional indicator. This lacks comprehensive consideration of safety dimensions such as voltage stability, transient stability, power angle stability, and cascading tripping after a fault, making it difficult to fully reflect the true risk level of the power grid under complex operating conditions.

[0083] To address the shortcomings of existing technologies that rely primarily on single data sources or static operating condition analysis for power grid risk assessment, which fails to comprehensively reflect the risk evolution process of the power grid under complex operating environments and lacks sufficient detail in characterizing fault cascading propagation and risk levels, the inventors propose a power grid risk assessment method based on the understanding that power grid operational risks are essentially driven by multiple factors and amplified step-by-step through topological coupling. In its implementation, this method first collects multi-source operational data, including historical load, real-time load, meteorological environment, and regional economic data, to construct an operational status description covering internal and external influencing factors of the power grid. Furthermore, standardization processing eliminates dimensional and scale differences between different data sources, facilitating unified modeling. The system provides a foundation; subsequently, a fault propagation model is constructed based on a standardized load dataset to simulate the propagation path and evolution process of a single fault under the constraints of power grid topology and operating conditions, generating multi-level fault scenarios; on this basis, according to a preset multi-dimensional risk assessment index system, the operational risks of the power grid under different fault scenarios are quantitatively calculated and graded; through the above technical solution, the transformation of potential power grid risks from static assessment to dynamic evolution assessment is realized, which can more accurately characterize the comprehensive impact of fault propagation on power grid security, improve the comprehensiveness, precision and foresight of risk assessment results, and provide reliable technical support for power grid operation scheduling and risk prevention and control.

[0084] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0085] It should be noted that this application adopts a layered distributed architecture, covering a data awareness layer, an edge computing layer, a core decision-making layer, an execution control layer, and a visualization layer. The hardware composition and functions of each layer are as follows:

[0086] Data perception layer (data acquisition unit): It consists of intelligent sensors, load monitoring terminals, meteorological data acquisition modules and economic data interfaces deployed on 110kV and above lines, main transformers and busbars; it is responsible for collecting power grid operation parameters, real-time load, meteorological environment and economic indicator data, and uploading them to edge computing nodes through 5G / fiber optic communication with a transmission latency of ≤100ms and data integrity of ≥99.9%.

[0087] Edge computing layer (edge ​​computing nodes): The processor is configured with 4GB cache and 1TB local storage; deploys data preprocessing programs, short-term load forecasting models (LSTM), and fault scenario screening modules; implements data cleaning and outlier removal. Guidelines), short-term load forecasting and basic fault scenario generation reduce the computing power pressure on the central server, and the local data retention period is ≥72 hours.

[0088] Core decision-making layer (central server cluster): consists of 8 GPU servers with a peak computing power of ≥100TFLOPS; deploys Petri net fault propagation model, multi-dimensional risk assessment engine, NSGA-III multi-objective optimization algorithm library and digital twin simulation module; responsible for complex fault scenario simulation, risk index calculation, optimization solution solution and effect verification, and supports parallel computing of more than 1,000 fault scenarios.

[0089] The execution control layer (execution control unit) includes substation automation terminals (RTUs), line sectionalizing switch controllers, distribution network tie switch actuators, and main transformer voltage regulating controllers; it receives optimization instructions from the central server to perform operations such as grid structure adjustment (e.g., line splitting, tie switch operation) and main transformer load rate adjustment, with a response time of ≤500ms.

[0090] Visualization layer (visualization terminal): adopts an industrial-grade touch screen (1920×1080 resolution) and deploys a decision support interface; displays risk assessment results, fault scenario simulation process, optimization solutions and equipment operating status in real time, and supports manual intervention (one-click solution issuance / rejection) and historical data backtracking query.

[0091] Figure 1 A flowchart illustrating the power grid risk assessment method provided in this application embodiment. Figure 1 ;like Figure 1 As shown, the method includes:

[0092] S101. Obtain multi-source operation data of the power grid.

[0093] The multi-source operational data includes at least historical load data, real-time load data, meteorological and environmental data, and regional economic data.

[0094] It should be understood that traditional risk assessment methods often rely on static or single load data, failing to reflect the dynamic changes in the power grid's operating status and the impact of the external environment, leading to discrepancies between risk assessment results and actual conditions. To address this issue, this application's embodiments employ a multi-source data fusion approach. Specifically, this involves deploying intelligent sensors (current / voltage accuracy) at key nodes in the power grid (such as 110kV and above substations, main transformers, and lines). High-frequency (≥1 time / minute) load monitoring terminals, high-precision (temperature) monitoring terminals. ,humidity The system utilizes a meteorological data acquisition module and a data interface that connects to the regional macroeconomic statistics platform to simultaneously collect four types of core data: 1) historical load data (e.g., time-series data spanning 10 years from 2012 to 2022, with over 100,000 sample points), used for analyzing long-term trends and training prediction models; 2) real-time load data, used to capture current operating status; 3) meteorological environmental data (temperature, humidity, precipitation, etc.), as they directly affect electricity load (e.g., air conditioning, heating load); and 4) regional economic data (e.g., GDP growth rate, industrial structure), used to correlate with medium- and long-term load growth trends. All data is aggregated in real time via 5G or fiber optic communication networks (transmission latency ≤100ms), forming a panoramic data pool covering time, space, physical, and socioeconomic dimensions, laying the data foundation for subsequent precise analysis.

[0095] Historical load data reflects the load variation patterns and periodic characteristics of the power grid at different time scales, while real-time load data characterizes the dynamic changes in load levels under current operating conditions. Meteorological environmental data (such as temperature, humidity, and rainfall) describes the impact of the natural environment on fluctuations in electricity demand and the probability of equipment failure. Regional economic data depicts the correlation between regional industrial structure, economic activity, and electricity demand growth trends. In practice, the above-mentioned multi-source operational data can be collected synchronously or quasi-synchronously through various data interfaces such as dispatch automation systems, power acquisition systems, meteorological information platforms, and statistical databases, and aligned according to a unified time scale to form a raw multi-source operational data set for subsequent analysis. For example, under high-temperature conditions in summer, historical load data for the same period, current real-time load data, corresponding high-temperature meteorological data, and regional industrial electricity consumption growth data can be acquired simultaneously to comprehensively reflect the operating environment of the power grid.

[0096] Understandably, by constructing a multi-dimensional and highly timely data collection system, the limitations of traditional methods with their single data source have been fundamentally changed. This allows risk assessment to fully consider the time-varying nature of loads, climate sensitivity, and correlation with economic development, significantly improving the quality of input data and the relevance to reality of the assessment model.

[0097] S102. Standardize the multi-source operating data to obtain a standardized load dataset.

[0098] In one possible implementation, the multi-source operational data is first normalized to obtain a first dataset; then, missing data in the first dataset is imputed to obtain a second dataset; finally, abnormal load data in the second dataset is identified and removed based on statistical criteria to obtain a standardized load dataset.

[0099] It should be understood that, due to significant differences in dimensions, value ranges, and data quality among multi-source operational data, directly using them for modeling and analysis can easily lead to decreased model convergence and stability. Therefore, this step unifies the multi-source data through standardization. This embodiment aims to transform raw data into high-precision dynamic load forecasting results through a standardized process, executed by edge computing nodes. Specifically, firstly, a normalization method is used to map data from different sources to a unified numerical range, eliminating the influence of dimensions; then, for missing data generated during the collection process, time series interpolation or weighted interpolation of adjacent time points is used to complete the data to ensure data continuity; finally, abnormal load data is identified and removed based on statistical criteria (such as the mean-standard deviation criterion or box plot criterion), thereby obtaining a reliable standardized load dataset.

[0100] It should be noted that after obtaining the standardized load dataset, the method also includes:

[0101] First, the standardized load dataset is input into the trained load prediction model to dynamically predict the power grid load. Then, the predicted load data is used as the load boundary conditions for generating multi-level fault scenarios and conducting risk assessment.

[0102] The load forecasting model is an LSTM neural network model trained based on historical multi-source operating data and corresponding historical loads.

[0103] It should be understood that historical multi-source operational data and corresponding historical loads are used as input features to construct a prediction model containing a 3-layer LSTM network (64-128 hidden units) and using the Adam optimizer (learning rate 0.001-0.005). This model is trained and updated online at the edge computing layer, and can simultaneously output: short-term predictions (hourly loads for the next 24 hours, updated hourly, with prediction accuracy...). ) and medium- to long-term forecasts (monthly load breakdown for the next year, revised monthly, forecast error) ).

[0104] Building upon this foundation, a standardized load dataset is input into a pre-trained load forecasting model to dynamically predict grid load for future periods. This model employs a long short-term memory (LSTM) neural network structure and is trained using historical multi-source operating data and their corresponding historical loads, enabling the model to fully learn the temporal characteristics of load changes. The predicted load data is further used as load boundary conditions for generating multi-level fault scenarios and conducting risk assessments. For example, based on the predicted load curve for the next 24 hours, fault propagation and risk levels under high load conditions can be analyzed.

[0105] Understandably, combining standardized processing with load forecasting not only improves data quality and the stability of model calculations, but also transforms the risk assessment process from one based on static load assumptions to one based on dynamic load boundary conditions, thereby enhancing the adaptability and foresight of the risk assessment results to actual operating conditions.

[0106] S103. Based on the standardized load dataset, generate multi-level fault scenarios using the fault propagation model.

[0107] In one possible implementation, a Petri net fault propagation model is first constructed based on the relationship between the power grid topology and the equipment; then, a hierarchical sampling method is used to generate multi-level fault scenarios, including single-equipment faults, dual-equipment faults, and complex coupling faults, through the Petri net fault propagation model.

[0108] It should be understood that the traditional N-1 criterion is difficult to simulate complex cascading faults. This application's embodiments construct a fault propagation model by introducing Petri net theory. This model can formally describe the dynamic logical relationship between equipment states (storage points) and fault events (transitions) in a power grid. In the specific implementation process, a digital Petri net model is first constructed based on the actual power grid topology, equipment connection relationships, and protection configuration logic. Then, a hierarchical sampling strategy is used to extrapolate from this model: the first layer samples to generate all possible single initial faults (such as a single line or a single main transformer exiting operation); the second layer, based on the consequences of the initial fault (such as power flow shift leading to equipment overload) and the propagation rules defined by the Petri net, samples to generate dual fault scenarios (such as a bus fault leading to a main transformer fault); the third layer further extrapolates complex coupled fault scenarios (such as the simultaneous complete shutdown of four lines due to being erected on the same tower). Through this method, a multi-level fault scenario set containing up to 2000 typical faults can be systematically and automatically generated, comprehensively covering various risk events from simple to extremely complex.

[0109] Understandably, this method breaks through the limitations of traditional single fault analysis and realizes the active simulation of complex cascading faults.

[0110] S104. Based on multi-level fault scenarios, and in accordance with the preset multi-dimensional risk assessment index system, quantitatively assess the operational risks of the power grid under each fault scenario, and determine the corresponding risk level.

[0111] In one possible implementation, firstly, for each fault scenario, simulation data of the power grid is obtained through transient simulation; then, based on the simulation data and the predicted load data, load risk parameters, voltage risk parameters, and equipment life risk parameters based on the equipment life prediction model are calculated respectively; then, the load risk parameters, voltage risk parameters, and equipment life risk parameters are weighted and summed according to preset weight coefficients to obtain the comprehensive risk value of the power grid operation risk; finally, the comprehensive risk value is compared with a preset threshold range to determine the risk level.

[0112] The simulation data includes node voltage, line current, and transformer load rate data.

[0113] It should be understood that, to address the issue of relying solely on traditional risk indicators (such as power shortage), this application embodiment constructs a three-dimensional risk assessment indicator system encompassing load, voltage, and equipment lifespan, comprehensively measuring the consequences of failures from different dimensions. The specific implementation process is as follows: For each failure scenario generated in S103, electromagnetic transient simulation tools, such as power system computer-aided design software, are invoked in the central server cluster (e.g., 8 GPU servers). Combined with the dynamic load forecast data obtained in S102, high-precision transient simulation is performed (the step size can be set to...). The simulation model simulates the dynamic process of the power grid for several seconds to several minutes after a fault, outputting simulation data (including voltage deviation at each node, line current, main transformer load rate, etc.). Based on this, three parameters are calculated in parallel: 1) Load risk parameter: the percentage of risky load to total load; 2) Voltage risk parameter: the weighted sum of voltage deviation at each node relative to the rated voltage; 3) Equipment life risk parameter: based on the Weibull distribution model, considering the number of overloads and years of operation, predicting the remaining life loss of key equipment. Subsequently, the three parameters are weighted and summed according to pre-set weighting coefficients reflecting the importance of each dimension (e.g., 0.5:0.3:0.2) to obtain the comprehensive risk value under this fault scenario. Finally, based on the threshold range of the comprehensive risk value (e.g., high risk ≥0.6, medium risk [0.3, 0.6), low risk <0.3), the risk level is automatically classified, and a risk assessment containing detailed indicators and levels is generated.

[0114] It should be noted that, to verify the effectiveness of the solution in this embodiment, this application selected a typical fault in the Dongguan power grid in 2012: a complete outage of the Shuiwan A and B lines on the same tower (the risk load calculated using traditional methods was 611MW), and verified it based on the power grid risk assessment method provided in this application. The results are shown in the table below:

[0115] Table 1. Validation Table of Risk Assessment Methods

[0116]

[0117] Understandably, multi-dimensional assessment provides a more comprehensive view of risk. For example, a fault might cause minimal load loss (low load risk) but trigger a severe voltage dip (high voltage risk) or cause a critical transformer to age faster due to short-term overload (high equipment life risk). Traditional methods might assess such events as low-risk and ignore them, while this invention, through comprehensive calculation, can accurately identify them as medium or high risk. In a verification case in the Dongguan power grid, for the complex fault of "complete shutdown of the Shuiwan A and B lines on the same tower," the risk load value assessed by this invention was 589MW, which is more accurate than the 611MW calculated by the traditional static method. The assessment time was reduced from 45 minutes to 8 minutes, improving efficiency by 82%.

[0118] The power grid risk assessment method provided in this application first constructs a comprehensive and realistic representation of the power grid's operational status by collecting multi-source operational data, thereby achieving comprehensive coverage of internal and external influencing factors at the data level. Subsequently, the multi-source operational data undergoes unified standardization processing. By eliminating dimensional differences, scaling the data distribution range, and suppressing the interference of outliers on the calculation results, a standardized load dataset is formed that can be directly used for modeling and analysis, improving the stability and comparability of subsequent model calculations. Based on this, a fault propagation model is constructed and introduced based on the standardized load dataset, characterizing the power grid topology, electrical coupling relationships, and causal transmission between faults. The mechanism simulates and extrapolates the cascading failure process that a single point of failure may trigger under different operating conditions, thereby generating multi-level failure scenarios covering different propagation paths and impact ranges, achieving a systematic characterization of the complex failure evolution process. Finally, for the generated multi-level failure scenarios, based on a pre-constructed multi-dimensional risk assessment index system, the power grid operation risk is quantitatively calculated from multiple dimensions such as power supply reliability, load loss degree, impact on key equipment, and failure propagation range. The risk level corresponding to each failure scenario is determined through comprehensive assessment results, thereby achieving a refined and quantifiable assessment of the power grid's operation risk under different potential failure scenarios, improving the power grid's risk prediction capability and the scientific nature of operation decisions.

[0119] Figure 2 A flowchart illustrating the power grid risk assessment method provided in this application embodiment. Figure 2 ;like Figure 2 As shown, in this embodiment... Figure 1 Based on the examples, the process of obtaining the power grid operation optimization scheme is described in detail. The method includes:

[0120] S201. Based on the risk level results, power grid topology data, and predicted load data, construct an objective optimization model with the goal of minimizing the sum of power grid construction costs and power outage losses, and constrained by the N-1 safety criterion and equipment operating limits.

[0121] Among them, the risk level assessment result refers to the conclusion obtained after quantitatively classifying the degree of risk of each node or line in the power grid potentially exceeding the limit under different preset fault scenarios; the power grid topology relationship data is a data set describing the electrical connection structure and connectivity between primary equipment such as transmission lines, transformers, busbars, and switches; the predicted load data is the active and reactive power demand values ​​of each node at future times, obtained through the new energy output prediction model and the user electricity consumption prediction model.

[0122] It should be understood that, in this embodiment, according to Figure 1The method described in this embodiment obtains risk level results, incorporates power grid topology data and predicted load data, and constructs a multi-objective optimization model for power grid operation. This model abstracts the power grid operation problem into a constrained optimization problem under given topology and load boundary conditions. The objective function is to minimize the sum of power grid construction costs and outage losses, comprehensively reflecting the investment costs and power supply reliability impacts of structural adjustments or operational mode changes. Simultaneously, the N-1 safety criterion is used as the core safety constraint to ensure safe operation of the power grid even under the failure of any single piece of equipment. Furthermore, equipment operating limits are introduced as constraints to restrict key operating parameters such as line current, node voltage, and transformer load rate, thereby ensuring the feasibility of the optimization results in engineering. It should be noted that in practical applications, the weights of the objective function can be adjusted according to different risk levels.

[0123] Understandably, in this embodiment, a target optimization model is first constructed based on the risk level assessment results, power grid topology data, and predicted load data. The objective function of this model is set to minimize the sum of power grid construction costs and power outage losses. Power grid construction costs refer to the engineering construction expenses incurred to improve the power grid's supply capacity, such as the construction of new lines and the expansion of substations. Power outage losses refer to the economic losses caused by power outages leading to production stoppages and equipment damage, typically estimated by multiplying the power outage cost by the estimated power loss. The model's constraints mainly include: the N-1 safety criterion, which is the reliability standard that ensures the remaining equipment can maintain normal power supply without triggering cascading overloads or voltage exceedances after any component (such as a line, transformer, or generator set) fails and goes out of service; and equipment operating limits, which are the upper and lower boundaries of the allowable operation of each device under steady-state conditions, such as the thermal stability limit power of lines, the rated capacity of transformers, and the allowable fluctuation range of node voltages. By constructing the aforementioned optimization model with economic efficiency and reliability as dual objectives and N-1 criterion and equipment limits as hard constraints, we can solve for the planning or operation scheme that meets the power grid safety redundancy requirements and has the lowest overall cost using algorithms such as mixed integer programming.

[0124] S202. Solve the target optimization model using the target optimization algorithm to obtain at least one power grid operation optimization scheme.

[0125] Among them, the power grid operation optimization scheme includes power grid structure adjustment scheme and / or operation mode adjustment scheme.

[0126] It should be noted that after obtaining the corresponding power grid operation optimization scheme, it is also necessary to verify the power grid operation optimization scheme to determine whether the optimization scheme can be operated. Specifically, the implementation effect of the optimization scheme is verified by digital twin simulation; if the verification is successful, the optimization scheme is output and executed; if the verification fails, the optimization objective or constraint conditions are adjusted and the solution is resolved until a verified optimization scheme is obtained.

[0127] It should be understood that, based on the constructed multi-objective optimization model, objective optimization algorithms (such as NSGA-III (Non-dominated sorting genetic algorithm III)) are used to solve the objective optimization model. That is, the Pareto optimal solution set is found in the feasible solution space composed of the objective function and constraints through a search algorithm, thereby obtaining at least one power grid operation optimization scheme that satisfies safety constraints and achieves a balance between economy and reliability. These schemes are mainly divided into two categories: 1) Power grid structure adjustment schemes: such as suggesting the splitting of heavily loaded 110kV lines, adding interconnecting lines at key sections, or connecting terminal 220kV substations to the 500kV ring network to enhance power supply support; 2) Operation mode adjustment schemes: for example, using the PPO (Proximal Policy Optimization) reinforcement learning algorithm, an agent is trained based on more than 5,000 historical operation mode data to dynamically optimize the open loop point position of the power grid to minimize the power outage recovery time under complex faults.

[0128] Furthermore, after obtaining the corresponding power grid operation optimization scheme, to avoid the optimization results being only theoretically feasible but posing risks in actual operation, the implementation effect of the optimization scheme is verified through digital twin simulation. The verification pass criteria are quantitative, such as requiring that after implementation, the risk load reduction of the corresponding fault scenario is ≥40%, the voltage recovery and stabilization are within a safe range, and the remaining lifespan of key equipment meets the requirements. If the verification passes, the scheme is marked as feasible and can be sent to the substation automation terminal, switch controller, and other equipment in the execution control layer for execution with one click via the visualization terminal. If the verification fails, the system automatically adjusts the optimization objectives or constraints and iterates again until a feasible scheme is obtained. After execution, the actual operation data collected by the data perception layer is fed back to the system, forming a complete closed loop of "evaluation-optimization-verification-execution-feedback", enabling continuous optimization of the model and strategy.

[0129] Understandably, the above technical solutions achieve deep integration of risk assessment results and power grid operation optimization, enabling optimization decisions to be adaptively adjusted for different risk levels. At the same time, the digital twin simulation verification mechanism effectively avoids the direct application of unfeasible or high-risk solutions. Under the premise of ensuring the N-1 safety criterion and the safe operation of equipment, the economic efficiency and reliability of power grid operation are taken into account, thereby improving the engineering feasibility and practical application value of power grid operation optimization solutions.

[0130] The power grid risk assessment method provided in this application first constructs a comprehensive and realistic representation of the power grid's operational status by collecting multi-source operational data, thereby achieving comprehensive coverage of internal and external influencing factors at the data level. Subsequently, the multi-source operational data undergoes unified standardization processing. By eliminating dimensional differences, scaling the data distribution range, and suppressing the interference of outliers on the calculation results, a standardized load dataset is formed that can be directly used for modeling and analysis, improving the stability and comparability of subsequent model calculations. Based on this, a fault propagation model is constructed and introduced based on the standardized load dataset, characterizing the power grid topology, electrical coupling relationships, and causal transmission between faults. The mechanism simulates and extrapolates the cascading failure process that a single point of failure may trigger under different operating conditions, thereby generating multi-level failure scenarios covering different propagation paths and impact ranges, achieving a systematic characterization of the complex failure evolution process. Finally, for the generated multi-level failure scenarios, based on a pre-constructed multi-dimensional risk assessment index system, the power grid operation risk is quantitatively calculated from multiple dimensions such as power supply reliability, load loss degree, impact on key equipment, and failure propagation range. The risk level corresponding to each failure scenario is determined through comprehensive assessment results, thereby achieving a refined and quantifiable assessment of the power grid's operation risk under different potential failure scenarios, improving the power grid's risk prediction capability and the scientific nature of operation decisions.

[0131] Figure 3 This is a schematic diagram of the structure of the power grid risk assessment device provided in the embodiments of this application; as shown below. Figure 3 As shown, the device includes:

[0132] The acquisition module 301 is used to acquire multi-source operation data of the power grid, which includes at least historical load data, real-time load data, meteorological environment data, and regional economic data.

[0133] Processing module 302 is used to standardize multi-source operating data to obtain a standardized load dataset;

[0134] The generation module 303 is used to generate multi-level fault scenarios based on a standardized load dataset and a fault propagation model.

[0135] The determination module 304 is used to quantitatively assess the operational risks of the power grid under various fault scenarios according to a preset multi-dimensional risk assessment system, based on multi-level fault scenarios, and to determine the corresponding risk level.

[0136] In one possible implementation, the processing module 302 is specifically used for:

[0137] The multi-source operational data is normalized to obtain the first dataset;

[0138] The missing data in the first dataset is imputed to obtain the second dataset;

[0139] Abnormal load data in the second dataset are identified and removed based on statistical criteria to obtain a standardized load dataset.

[0140] In one possible implementation, after obtaining the standardized load dataset, the processing module 302 is further configured to:

[0141] The load forecasting model dynamically predicts the power grid load by inputting a standardized load dataset into the trained load forecasting model. The load forecasting model is an LSTM neural network model trained based on historical multi-source operating data and corresponding historical loads.

[0142] The predicted load data is used as the load boundary condition for generating multi-level fault scenarios and conducting risk assessments.

[0143] In one possible implementation, the generation module 303 is specifically used for:

[0144] Based on the relationship between power grid topology and equipment, a Petri net fault propagation model is constructed;

[0145] A stratified sampling method is adopted to generate multi-level fault scenarios, including single-device faults, dual-device faults, and complex coupling faults, through a Petri net fault propagation model.

[0146] In one possible implementation, the determining module 304 is specifically used for:

[0147] For each fault scenario, transient simulation data of the power grid is obtained, including node voltage, line current and transformer load rate data.

[0148] Based on the simulation data and the predicted load data, the load risk parameters, voltage risk parameters, and equipment life risk parameters based on the equipment life prediction model are calculated respectively.

[0149] According to the preset weighting coefficients, the load risk parameters, voltage risk parameters and equipment life risk parameters are weighted and summed to obtain the comprehensive risk value of power grid operation risk;

[0150] The overall risk value is compared with a preset threshold range to determine the risk level.

[0151] In one possible implementation, the power grid risk assessment device further includes an optimization module, which, after determining the risk level, is used to:

[0152] Based on the risk level results, power grid topology data, and predicted load data, a target optimization model is constructed with the objective of minimizing the sum of power grid construction costs and power outage losses, and constrained by the N-1 safety criterion and equipment operating limits.

[0153] The objective optimization model is solved by using an objective optimization algorithm to obtain at least one power grid operation optimization scheme, which includes a power grid structure adjustment scheme and / or an operation mode adjustment scheme.

[0154] In one possible implementation, the power grid risk assessment device further includes an optimization module for:

[0155] The effectiveness of the optimization scheme was verified through digital twin simulation.

[0156] If the verification passes, the optimization plan will be output and executed.

[0157] If the verification fails, adjust the optimization objective or constraints and solve again until a successful optimization solution is obtained.

[0158] The power grid risk assessment device provided in this application embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and will not be described in detail here.

[0159] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 40 provided in this embodiment includes at least one processor 401 and a memory 402. Optionally, the device 40 also includes a communication component 403.

[0160] The processor 401, communication component 403, and memory 402 communicate with each other via communication bus 404. Communication component 403 communicates with other network elements, such as clients or other servers. Processor 402 executes programs, specifically the steps described above in the embodiment of the power grid risk assessment method.

[0161] In the above embodiments, it should be understood that the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this invention can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0162] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0163] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0164] Specifically, the program can be invoked by the processor to enable the control equipment for power grid risk assessment to perform the following operations:

[0165] Acquire multi-source operation data of the power grid, which includes at least historical load data, real-time load data, meteorological and environmental data, and regional economic data;

[0166] Standardize the multi-source operational data to obtain a standardized load dataset;

[0167] Based on a standardized load dataset, a multi-level fault scenario is generated using a fault propagation model.

[0168] Based on multi-level fault scenarios and in accordance with a pre-set multi-dimensional risk assessment system, the operational risks of the power grid under each fault scenario are quantitatively assessed, and the corresponding risk levels are determined.

[0169] In one possible implementation, the multi-source operational data is standardized to obtain a standardized load dataset, including:

[0170] The multi-source operational data is normalized to obtain the first dataset;

[0171] The missing data in the first dataset is imputed to obtain the second dataset;

[0172] Abnormal load data in the second dataset are identified and removed based on statistical criteria to obtain a standardized load dataset.

[0173] By employing the aforementioned technical means, the impact of differences in units and value ranges between data from different sources on the analysis results can be eliminated. At the same time, missing data can be repaired and abnormal data can be removed, thereby improving the consistency, completeness, and effectiveness of input data. This provides high-quality data support for subsequent load forecasting, fault scenario generation, and risk assessment, reduces the interference of noisy data on the assessment results, and improves the stability and credibility of risk analysis results.

[0174] In one possible implementation, after obtaining the standardized load dataset, the method further includes:

[0175] The load forecasting model dynamically predicts the power grid load by inputting a standardized load dataset into the trained load forecasting model. The load forecasting model is an LSTM neural network model trained based on historical multi-source operating data and corresponding historical loads.

[0176] The predicted load data is used as the load boundary condition for generating multi-level fault scenarios and conducting risk assessments.

[0177] In one possible implementation, based on a standardized load dataset, a multi-level fault scenario is generated using a fault propagation model, including:

[0178] Based on the relationship between power grid topology and equipment, a Petri net fault propagation model is constructed;

[0179] A stratified sampling method is adopted to generate multi-level fault scenarios, including single-device faults, dual-device faults, and complex coupling faults, through a Petri net fault propagation model.

[0180] In one possible implementation, based on multi-level fault scenarios and according to a pre-set multi-dimensional risk assessment system, the operational risks of the power grid under each fault scenario are quantitatively assessed to determine the corresponding risk levels, including:

[0181] For each fault scenario, transient simulation data of the power grid is obtained, including node voltage, line current and transformer load rate data.

[0182] Based on the simulation data and the predicted load data, the load risk parameters, voltage risk parameters, and equipment life risk parameters based on the equipment life prediction model are calculated respectively.

[0183] According to the preset weighting coefficients, the load risk parameters, voltage risk parameters and equipment life risk parameters are weighted and summed to obtain the comprehensive risk value of power grid operation risk;

[0184] The overall risk value is compared with a preset threshold range to determine the risk level.

[0185] By employing the aforementioned technical means, it is possible to comprehensively consider the influencing factors of multiple dimensions, such as load risk, voltage risk, and equipment lifespan risk, and to quantitatively evaluate the power grid operation status through a unified comprehensive risk index. This enables a coordinated assessment of power grid safety, stability, and equipment health status, avoiding assessment biases caused by single risk indicators, improving the scientificity and accuracy of risk level determination, and providing a reliable basis for risk warning and operation and maintenance decisions.

[0186] In one possible implementation, after determining the risk level, the method further includes:

[0187] Based on the risk level results, power grid topology data, and predicted load data, a target optimization model is constructed with the objective of minimizing the sum of power grid construction costs and power outage losses, and constrained by the N-1 safety criterion and equipment operating limits.

[0188] The objective optimization model is solved by using an objective optimization algorithm to obtain at least one power grid operation optimization scheme, which includes a power grid structure adjustment scheme and / or an operation mode adjustment scheme.

[0189] In one possible implementation, the method further includes:

[0190] The effectiveness of the power grid operation optimization scheme was verified through digital twin simulation.

[0191] If the verification passes, the optimization plan will be output and executed.

[0192] If the verification fails, adjust the optimization objective or constraints and solve again until a successful optimization solution is obtained.

[0193] This application also provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the above-described method.

[0194] The aforementioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The readable storage medium can be any available medium accessible to a general-purpose or special-purpose computer.

[0195] An exemplary readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can reside in an Application Specific Integrated Circuit (ASIC). Alternatively, the processor and the readable storage medium can exist as discrete components in the device.

[0196] The division of units is merely a logical functional division; in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0197] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0198] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0199] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer 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 steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0200] Based on the above disclosure, those skilled in the art will understand that the computer-readable storage medium and its related implementation methods provided in the embodiments of the present invention are not limited to the specific hardware or logical partitioning forms described above. In actual product or system deployments, computer execution instructions can be encapsulated in different software distribution packages or firmware modules and distributed via networks, optical discs, or mobile storage devices. When computer execution instructions are loaded and executed by one or more processors, in addition to implementing the aforementioned method steps, hardware resource configurations can be automatically identified according to specific application scenarios, and thread priorities or memory allocation strategies can be dynamically adjusted to optimize execution efficiency without manual intervention. Furthermore, the computer-readable storage medium can also be integrated into IoT terminals, edge computing nodes, or cloud servers to form a distributed instruction storage and collaborative execution system to meet differentiated needs such as high concurrency, low latency, or data privacy protection.

[0201] Furthermore, it should be emphasized that the logical functions corresponding to the methods, units, modules, and storage media in this application, without departing from the core idea of ​​this invention, allow developers to make equivalent refactoring, renaming, or re-division based on specific programming languages ​​(such as C++, Java, Python, etc.) or framework characteristics. For example, the functions of multiple "units" can be merged into a single service process, or message queues or pipe communication can be used instead of direct function calls. Even if these implementation details are not completely the same as the foregoing embodiments, as long as they ultimately achieve the same technical effect and no creative effort is required, they should all be considered to fall within the protection scope of this invention. Similarly, any intermediate data processing forms, exception handling mechanisms, or logging methods embodied in the instructions in the computer-readable storage medium when executed are also derivative technologies of this invention and should not be construed as substantial limitations on the original solution.

[0202] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0203] Finally, it should be noted that other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein, and is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of the invention is limited only by the appended claims.

Claims

1. A power grid risk assessment method, characterized in that, include: Acquire multi-source operation data of the power grid, including at least historical load data, real-time load data, meteorological environmental data, and regional economic data; The multi-source operational data is standardized to obtain a standardized load dataset; Based on the standardized load dataset, a multi-level fault scenario is generated using a fault propagation model. Based on the aforementioned multi-level fault scenarios, and in accordance with a pre-set multi-dimensional risk assessment system, the operational risks of the power grid under each fault scenario are quantitatively assessed, and the corresponding risk levels are determined.

2. The method according to claim 1, characterized in that, The standardization process of the multi-source operational data to obtain a standardized load dataset includes: The multi-source operational data is normalized to obtain the first dataset; The missing data in the first dataset is imputed to obtain the second dataset; Abnormal load data in the second dataset are identified and removed based on statistical criteria to obtain the standardized load dataset.

3. The method according to claim 2, characterized in that, After obtaining the standardized load dataset, the method further includes: The standardized load dataset is input into the trained load prediction model to dynamically predict the power grid load. The load prediction model is an LSTM neural network model trained based on historical multi-source operating data and corresponding historical loads. The predicted load data is used as the load boundary condition for generating the multi-level fault scenarios and conducting risk assessments.

4. The method according to claim 1, characterized in that, The generation of multi-level fault scenarios based on the standardized load dataset and using a fault propagation model includes: Based on the relationship between power grid topology and equipment, a Petri net fault propagation model is constructed; A stratified sampling method is used to generate multi-level fault scenarios, including single-device faults, dual-device faults, and complex coupling faults, through the Petri net fault propagation model.

5. The method according to claim 1 or 2, characterized in that, The step involves quantifying the operational risks of the power grid under each fault scenario based on the multi-level fault scenarios and a pre-defined multi-dimensional risk assessment system, thereby determining the corresponding risk levels. This includes: For each fault scenario, transient simulation data of the power grid is obtained, including node voltage, line current and transformer load rate data. Based on the simulation data and the predicted load data, calculate the load risk parameter, voltage risk parameter, and equipment life risk parameter based on the equipment life prediction model. According to the preset weighting coefficients, the load risk parameters, voltage risk parameters and equipment life risk parameters are weighted and summed to obtain the comprehensive risk value of power grid operation risk; The overall risk value is compared with a preset threshold range to determine the risk level.

6. The method according to claim 5, characterized in that, After determining the risk level, the method further includes: Based on the risk level results, power grid topology data, and predicted load data, a target optimization model is constructed with the objective of minimizing the sum of power grid construction costs and power outage losses, and constrained by the N-1 safety criterion and equipment operating limits. The objective optimization model is solved using an objective optimization algorithm to obtain at least one power grid operation optimization scheme, which includes a power grid structure adjustment scheme and / or an operation mode adjustment scheme.

7. The method according to claim 6, characterized in that, The method further includes: The effectiveness of the power grid operation optimization scheme was verified through digital twin simulation. If the verification passes, the optimization scheme will be output and executed. If the verification fails, adjust the optimization objective or constraints and solve again until a successful optimization solution is obtained.

8. A power grid risk assessment device, characterized in that, include: The acquisition module is used to acquire multi-source operation data of the power grid, which includes at least historical load data, real-time load data, meteorological environment data, and regional economic data. The processing module is used to standardize the multi-source operating data to obtain a standardized load dataset; The generation module is used to generate multi-level fault scenarios based on the standardized load dataset and using a fault propagation model. The determination module is used to quantitatively assess the operational risks of the power grid under each fault scenario according to the multi-level fault scenarios and a preset multi-dimensional risk assessment system, and to determine the corresponding risk level.

9. An electronic device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1-7.

Images