Power grid equipment fault simulation method based on digital twinning

Abstract

The present application relates to the technical field of digital twinning, and particularly relates to a power grid equipment fault simulation method based on digital twinning, comprising: determining whether to adjust the disassembly mode from structure analysis to dynamic response analysis based on the load stability of the target power grid model; determining whether the target power grid model is in a fault state based on the node complex correlation degree; screening the target power grid model in the fault state to obtain a first priority sequence; determining whether to adjust the fault analysis benchmark from cycle analysis to fluctuation analysis based on the time sequence fluctuation degree; in the cycle analysis, screening the target power grid model in the healthy state based on the descending order of the cycle impact coefficient to obtain a second priority sequence; in the fluctuation analysis, determining whether to adjust the analysis benchmark from the fixed time sequence cycle to the time sequence segmentation based on the anchor point fluctuation margin to obtain the second priority sequence. The present application improves the adaptability of the power grid equipment fault simulation to the dynamic evolution of the operating state.

Description

Technical Field

[0001] This invention relates to the field of digital twin technology, and in particular to a method for simulating power grid equipment faults based on digital twins. Background Technology

[0002] Power grid equipment fault simulation is a crucial support tool for power system safety analysis and operation and maintenance decision-making. Traditional fault simulations often rely on offline physical models or empirical judgment methods based on fixed rules, lacking adaptive representation capabilities for the dynamic evolution of power grid operating states, complex coupling relationships between nodes, and the non-stationary characteristics of fault propagation. The introduction of digital twin technology provides a new approach to power grid equipment fault simulation through virtual-real mapping and real-time interaction. However, existing digital twin-based fault simulation methods suffer from problems such as lagging fault state identification and unreasonable allocation of simulation resources.

[0003] Chinese Patent Publication No. CN120334680A discloses a power grid digital twin fault simulation device for historical data reconstruction, comprising: a fault feature acquisition unit that collects power grid fault feature information when a target power grid fault occurs; a historical data retrieval unit that indexes historical power grid faults based on the power grid fault feature information to obtain a dataset of historical power grid faults of the same family and historical power grid fault operation data at the time of occurrence of the same family of historical power grid faults; a historical data reconstruction unit that acquires device maintenance data of multiple power grid devices after the same family of historical power grid faults, performs historical data reconstruction to generate a reconstructed power grid fault operation dataset; and a digital twin simulation unit that uses the reconstructed power grid fault operation dataset to construct a power grid fault twin simulation result. It is evident that the above technical solution has the following problems: relying on historical fault data retrieval and reconstruction lacks generalization ability for unknown fault states or topology changes, resulting in fault state identification lagging behind the actual evolution process; and under limited computing resources, the order of fault simulation analysis cannot be effectively controlled, thus reducing the efficiency and real-time response of fault simulation under complex operating conditions. Summary of the Invention

[0004] To address this, the present invention provides a power grid equipment fault simulation method based on digital twins, which overcomes the problems in the prior art caused by the reliance on historical fault data retrieval and reconstruction, such as insufficient generalization ability for unknown fault states or topology changes, lagging fault state identification, and unreasonable allocation of simulation resources.

[0005] To achieve the above objectives, the present invention provides a method for simulating power grid equipment faults based on digital twins, comprising: Based on the load stability corresponding to the target power grid model, determine whether to change the deconstruction method from structural analysis to dynamic response analysis. Based on the deconstruction method, deconstruct the target power grid model to obtain the power grid sub-model, and obtain the node complexity correlation based on the power grid sub-model. Determine whether the target power grid model is in a fault state based on the node complexity correlation; Based on the descending order of topological fault coupling degree, target power grid models in fault state are screened to obtain a first preset number of target power grid models, which is denoted as the first priority sequence. Determine whether to adjust the fault analysis benchmark from periodic analysis to volatility analysis based on the time series volatility. In the periodic analysis, the target power grid models in a healthy state are screened based on the descending order of the periodic impact coefficient to obtain a second preset number of target power grid models, which is denoted as the second priority sequence. In fluctuation analysis, the analysis benchmark is determined based on the anchor point fluctuation margin to determine whether it is adjusted from a fixed time period to a time segment. The target power grid models in a healthy state are screened based on the descending order of the time series dispersion to obtain a second preset number of target power grid models, which is denoted as the second priority sequence. Power grid equipment fault simulation is performed sequentially based on the first priority sequence and the second priority sequence corresponding to the target power grid model.

[0006] Furthermore, for the condition that the load stability is greater than the preset load stability, the deconstruction method is determined to be structural analysis; For conditions where the load stability is less than or equal to the preset load stability, the solution method is determined to be dynamic response analysis; The load stability is determined based on the node static voltage.

[0007] Furthermore, the condition that the node complexity correlation is greater than the preset node complexity correlation determines that the target power grid model is in a fault state; The topology fault coupling degree of the target power grid model is determined based on the electrical distance between nodes and the frequency of historical fault associations.

[0008] Furthermore, the complexity correlation is determined based on the node degree and the port equivalent impedance; The complexity correlation is positively correlated with node degree and port equivalent impedance.

[0009] Furthermore, the condition that the node complexity correlation degree is less than or equal to the preset node complexity correlation degree determines that the target power grid model is in a healthy state; For the condition that the time series fluctuation is less than the preset time series fluctuation, the fault analysis benchmark is determined to be periodic analysis; For the condition that the time series volatility is greater than or equal to the preset time series volatility, the fault analysis benchmark is determined to be volatility analysis; The timing fluctuation is determined based on the state switching frequency.

[0010] Furthermore, the periodic impact coefficient is determined based on the maximum current impact amplitude and rated current of the node within the period; The periodic impact coefficient is positively correlated with the maximum current impact amplitude.

[0011] Furthermore, in the fluctuation analysis, for the condition that the anchor point fluctuation margin is less than the preset anchor point fluctuation margin, the analysis is determined to be based on a fixed time period. For the condition that the anchor point volatility margin is greater than or equal to the preset anchor point volatility margin, it is determined that the analysis is based on time series segmentation. The anchor point fluctuation margin is determined based on the variance and average voltage of the voltage time series data at the preset anchor point.

[0012] Furthermore, the time-series dispersion is determined based on the voltage fluctuation amplitude, power fluctuation amplitude, and state switching frequency; The time-series dispersion is positively correlated with voltage fluctuation amplitude, power fluctuation amplitude, and state switching frequency.

[0013] Furthermore, the topology network corresponding to the target power grid model has substations, buses, generating equipment, load equipment, and switching equipment as nodes, and transmission lines and transformer connection strength as edges.

[0014] Furthermore, the first preset quantity and the second preset quantity are determined based on the fault simulation coverage rate; The first preset quantity and the second preset quantity are positively correlated with the fault simulation coverage rate; the first preset quantity is greater than the second preset quantity.

[0015] Compared with the prior art, the beneficial effect of the present invention is that the present invention reflects the static stability level of the grid node voltage by the load stability corresponding to the target grid model, and compares it with the preset load stability to determine whether to adjust the deconstruction method from structural analysis to dynamic response analysis. This improves the adaptability of the analysis method to the grid operation state, thereby avoiding the problem of simulation accuracy reduction caused by the fixed deconstruction method under the scenario of severe load fluctuation.

[0016] This invention reflects the connection scale and electrical coupling tightness of nodes in the power grid by fusing node degree and port equivalent impedance through node complexity correlation, thereby improving the multi-dimensional assessment capability and accuracy of fault state identification. This avoids fault omission or misjudgment caused by matching a single historical feature. Furthermore, it selects models in fault state as the first priority sequence based on the descending order of topological fault coupling degree, improving the priority response efficiency of high-risk models and ensuring that critical fault states are allocated simulation resources more quickly.

[0017] This invention reflects the frequency of power grid state switching through time-series fluctuation. Based on the time-series fluctuation, it determines whether to adjust the fault analysis benchmark from periodic analysis to fluctuation analysis. In the fluctuation analysis, it dynamically adjusts and supplements the benchmark based on the anchor point fluctuation margin. Different priority sequences are generated by combining the periodic impact coefficient and the anchor point fluctuation margin. Fault simulation is performed sequentially based on the priority sequences, which improves the adaptive switching capability of the analysis benchmark under dynamic operating conditions. This avoids the problems of missing key faults and low simulation efficiency caused by fixed analysis cycles or single sequences. Attached Figure Description

[0018] Figure 1 This is a flowchart of the power grid equipment fault simulation method based on digital twins according to the present invention; Figure 2 This is a flowchart illustrating the present invention's method of determining whether to change the deconstruction approach from structural analysis to dynamic response analysis based on load stability. Figure 3 This is a flowchart illustrating the process of determining whether a target power grid model is in a fault state based on node complexity correlation in this invention. Figure 4 This is a flowchart illustrating the process of determining whether to adjust the fault analysis benchmark from periodic analysis to fluctuation analysis based on the timing volatility of this invention. Detailed Implementation

[0019] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0020] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0021] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0022] Please see Figure 1 The diagram shown is a module connection diagram of a power grid equipment fault simulation method based on digital twins according to an embodiment of the present invention. The present invention provides a power grid equipment fault simulation method based on digital twins, comprising: Based on the load stability corresponding to the target power grid model, determine whether to change the deconstruction method from structural analysis to dynamic response analysis. Based on the deconstruction method, deconstruct the target power grid model to obtain the power grid sub-model, and obtain the node complexity correlation based on the power grid sub-model. Determine whether the target power grid model is in a fault state based on the node complexity correlation; Based on the descending order of topological fault coupling degree, target power grid models in fault state are screened to obtain a first preset number of target power grid models, which is denoted as the first priority sequence. Determine whether to adjust the fault analysis benchmark from periodic analysis to volatility analysis based on the time series volatility. In the periodic analysis, the target power grid models in a healthy state are screened based on the descending order of the periodic impact coefficient to obtain a second preset number of target power grid models, which is denoted as the second priority sequence. In fluctuation analysis, the analysis benchmark is determined based on the anchor point fluctuation margin to determine whether it is adjusted from a fixed time period to a time segment. The target power grid models in a healthy state are screened based on the descending order of the time series dispersion to obtain a second preset number of target power grid models, which is denoted as the second priority sequence. Power grid equipment fault simulation is performed sequentially based on the first priority sequence and the second priority sequence corresponding to the target power grid model.

[0023] In this embodiment of the invention, the selection of an adaptive deconstruction method based on load stability improves the model deconstruction effect. Based on the deconstructed power grid sub-model, the node complexity correlation is analyzed, and the fault state of the power grid model is effectively identified through the node complexity correlation. Furthermore, a fault state model sequence is constructed based on topological fault coupling degree, and a healthy state model sequence is constructed based on different benchmarks selected based on time-series fluctuation degree. Fault simulation is performed based on the model sequence, avoiding the problem that existing technologies cannot effectively control the fault simulation analysis sequence under limited computing resources, thereby reducing the efficiency and real-time response of fault simulation under complex operating conditions.

[0024] In this embodiment of the invention, the method of acquiring the target power grid model includes collecting the operating data of substations, transmission lines, switching equipment and load nodes in the target power grid, constructing a digital twin model corresponding to the actual power grid based on the collected operating data, and continuously updating the operating status of the digital twin model through real-time data stream to ensure that the target power grid model and the actual power grid remain synchronized. The operating data includes acquiring voltage, current, active power, reactive power and switch status.

[0025] In this embodiment of the invention, the topology network corresponding to the target power grid model uses substations, buses, generating equipment, load equipment, and switching equipment as nodes, and transmission lines and transformer connection strength as edges. Weights are assigned to each edge based on the line impedance parameters to reflect the electrical connection strength between nodes. The line impedance is determined based on reactance, and the edge weight is positively correlated with the reactance value. The smaller the reactance value, the closer the electrical distance between the two nodes and the stronger the power transmission capability, and the smaller the corresponding edge weight.

[0026] Please continue reading. Figures 1 to 4 As shown, for the condition that the load stability is greater than the preset load stability, the deconstruction method is determined to be structural analysis; For load stability conditions that are less than or equal to the preset load stability, dynamic response analysis is determined as the decomposition method.

[0027] In this embodiment of the invention, the load stability is used to determine the stability of the current operating state of the target power grid model. It can be understood that a higher load stability indicates that the nodes are farther from the voltage instability boundary, and the overall system operating state is more stable. At this time, the power grid topology changes less, and structural analysis can reflect the fault propagation relationship between devices. Conversely, a lower load stability indicates that the nodes are close to the instability boundary, and the system is more prone to voltage fluctuations and power oscillations after being disturbed. Therefore, dynamic response analysis is needed to improve the accuracy of fault simulation. The load stability is determined based on the static voltage of the nodes. Preferably, the voltage amplitude, active power, and reactive power data of each node under its operating state are obtained, the margin between the nodes and the voltage instability critical point is calculated, and the margin corresponding to each node is normalized to obtain the load stability.

[0028] In this embodiment of the invention, the preset value can be obtained through empirical setting based on historical data. For example, for the preset load stability, the load stability in qualified historical operating data is collected, outliers are removed, and the average value is calculated. The result is 0.15, which is recorded as the preset load stability. The historical operating data refers to the parameters used in this embodiment of the invention during a single historical fault simulation analysis. For any historical fault simulation analysis, whether the historical operating data is qualified historical operating data can be determined by user-defined standards. For example, the fault false alarm rate is set as a user-defined standard. When the fault false alarm rate meets the user's requirements, the historical operating data is qualified historical operating data. The above content is easily understood by those skilled in the art and will not be elaborated here.

[0029] In this embodiment of the invention, when the load stability is greater than a preset load stability, the deconstruction method is determined to be structural analysis. The steps of structural analysis include: based on the steady-state power flow equations and topological constraints of the power grid, obtaining the node injection power and topological connectivity of the target power grid model from a digital twin platform, and establishing a node admittance matrix; solving the power flow equations using the Newton-Raphson method to obtain the steady-state voltage amplitude and phase angle of each node; simulating equipment faults by changing branch states, and observing voltage exceedances and power redistribution at other nodes to determine the scope of fault impact. Through structural analysis, a power grid sub-model is obtained, which includes node topological connectivity, node admittance matrix, and steady-state electrical quantities of each node; based on the power grid sub-model, node degrees and port equivalent impedances are extracted for subsequent acquisition of node complexity correlations.

[0030] In this embodiment of the invention, when the load stability is less than or equal to a preset load stability, the deconstruction method is adjusted to dynamic response analysis. The steps of dynamic response analysis include considering generator rotor motion, excitation regulation, and load dynamic characteristics based on electromechanical transient differential-algebraic equations and electromagnetic transient differential-algebraic equations; establishing a transient stability model combining differential and algebraic equations based on the steady state obtained from structural analysis; and calculating the changes in power angle, frequency, and voltage over time after a fault occurs using time-domain simulation with stepwise integration. Through dynamic response analysis, a power grid sub-model is obtained, which includes time-domain response trajectory, state switching characteristics, and transient electrical quantities. Based on the power grid sub-model, node degrees and port equivalent impedances are extracted for subsequent acquisition of node complexity correlations. This analysis strategy avoids the insufficient accuracy caused by using a unified analysis framework in all scenarios, effectively improving the efficiency and accuracy of fault simulation.

[0031] Specifically, the condition that the node complexity correlation is greater than the preset node complexity correlation determines that the target power grid model is in a fault state; The topology fault coupling degree of the target power grid model is determined based on the electrical distance between nodes and the frequency of historical fault associations.

[0032] In this embodiment of the invention, topology fault coupling is used to determine the strength of fault propagation correlation between nodes under the dual dimensions of electrical connection and historical fault statistics. It can be understood that the smaller the electrical distance, the more direct the impact between nodes; the higher the frequency of historical fault correlations, the stronger the fault propagation correlation between nodes; therefore, the larger the topology fault coupling, the more likely the corresponding node is to trigger cascading failures. The topology fault coupling is determined based on the electrical distance between nodes and the frequency of historical fault correlations. The electrical distance is the shortest electrical path in the target power grid topology network, and the frequency of historical fault correlations is determined by the number of times fault linkages between nodes occurred during historical operation. The topology fault coupling is negatively correlated with the electrical distance and positively correlated with the frequency of historical fault correlations. Preferably, for each pair of nodes in the target power grid model, the reciprocal of the electrical distance is calculated as the product of the historical fault correlation frequency of that pair of nodes. The sum of the products of all node pairs is then used to obtain the topology fault coupling of the model. A preset topology fault coupling is used to determine the sensitivity to the risk of fault propagation correlation between power grid nodes, i.e., a benchmark threshold for determining whether the target power grid model has a high-risk cascading failure tendency. In this embodiment of the invention, the topology fault coupling degree is collected from qualified historical operating data. The topology fault coupling degree is set empirically based on historical data. After removing outliers, the average value of the topology fault coupling degree is calculated to be 0.35, which is recorded as the preset topology fault coupling degree.

[0033] Specifically, the complexity correlation is determined based on the node degree and the port equivalent impedance; The complexity correlation is positively correlated with node degree and port equivalent impedance.

[0034] In this embodiment of the invention, the topology network corresponding to the target power grid model is obtained, and the node degree and port equivalent impedance of each node are calculated. The node complexity correlation is used to determine the connection scale and electrical coupling tightness of the node in the power grid. It can be understood that the node complexity correlation is obtained by scaling the node degree and port equivalent impedance using a transformation parameter. Preferably, the influence of the node degree and port equivalent impedance on the node complexity correlation can be controlled by the transformation parameter. Therefore, the greater the restriction requirement on the adjustment capability of the node degree and port equivalent impedance, the larger the value of the transformation parameter. The preferred value of the transformation parameter is 0.5. The larger the node complexity correlation, the more significant the impact of the node on the operation of the power grid. Therefore, in scenarios with higher requirements for the safe operation of the power grid, the preset node complexity correlation should be smaller. Preferably, the complexity correlation of qualified historical operation data is collected, and the average value is calculated after removing outliers, which is 0.6, and is recorded as the preset node complexity correlation.

[0035] Specifically, the condition that the node complexity correlation degree is less than or equal to the preset node complexity correlation degree determines that the target power grid model is in a healthy state; For the condition that the time series fluctuation is less than the preset time series fluctuation, the fault analysis benchmark is determined to be periodic analysis; For the condition that the time series volatility is greater than or equal to the preset time series volatility, the fault analysis benchmark is determined to be volatility analysis; The timing fluctuation is determined based on the state switching frequency.

[0036] In this embodiment of the invention, the time-series volatility is used to determine the frequency of change in the operating state of the target power grid model. The time-series volatility is determined based on the node state switching frequency, and is the number of times the node switch state or protection action state switches per unit time. It can be understood that when the time-series volatility is small, the system operation pattern has strong periodicity, so periodic analysis is used; when the time-series volatility is large, the system operating state changes frequently, so volatility analysis is used to improve the ability to identify abnormal states. Preferably, time-series volatility data is collected from qualified historical operating data, outliers are removed, and the average value is calculated. The result is 0.05 times per second, which is recorded as the preset time-series volatility.

[0037] Specifically, the periodic impact coefficient is determined based on the maximum current impact amplitude and rated current of the node within the period; The periodic impact coefficient is positively correlated with the maximum current impact amplitude.

[0038] In this embodiment of the invention, during periodic analysis, a second priority sequence is obtained by screening the target power grid model based on the descending order of the periodic impact coefficient. The periodic impact coefficient is used to determine the electrical impact intensity that a node experiences within a period. It can be understood that the periodic impact coefficient is the ratio of the maximum current impact amplitude of a node within a period to its rated current. The larger the periodic impact coefficient, the more pronounced the periodic electrical impact on the node, and the higher the probability of equipment failure. Therefore, fault simulation is prioritized. Preferably, the average value of the periodic impact coefficient obtained by collecting qualified historical operating data, after removing outliers, is 2.5, which is denoted as the preset periodic impact coefficient.

[0039] Specifically, in the fluctuation analysis, for the condition that the anchor point fluctuation margin is less than the preset anchor point fluctuation margin, the analysis is determined to be based on a fixed time period. For the condition that the anchor point volatility margin is greater than or equal to the preset anchor point volatility margin, it is determined that the analysis is based on time series segmentation. The anchor point fluctuation margin is determined based on the variance and average voltage of the voltage time series data at the preset anchor point.

[0040] In this embodiment of the invention, during fluctuation analysis, the supplementary benchmark is adjusted from a fixed time period to time segmentation based on the anchor point fluctuation margin. The anchor point fluctuation margin is used to determine the relative severity of voltage fluctuations at a preset anchor point, which is a node in the power grid with stable voltage and complete historical data. The anchor point fluctuation margin is determined based on the variance and average voltage of the voltage time series data at the preset anchor point. Voltage time series sample values ​​within the anchor point period are obtained, and the statistical variance and the arithmetic mean of the sample values ​​are calculated. The anchor point fluctuation margin is obtained by dividing the variance by the average value. A larger anchor point fluctuation margin indicates that a fixed period is insufficient to accurately reflect actual operating characteristics. Therefore, a time segmentation approach is used for supplementary analysis to improve the accuracy and timeliness of fault simulation results. Preferably, the anchor point fluctuation margin obtained by collecting qualified historical operating data, removing outliers, and calculating the average value is 0.05, which is denoted as the preset anchor point fluctuation margin.

[0041] In this embodiment of the invention, during fluctuation analysis, a supplementary benchmark is determined based on the comparison between the anchor point fluctuation margin and the preset anchor point fluctuation margin. When the anchor point fluctuation margin is less than the preset anchor point fluctuation margin, a fixed time-series periodic analysis is adopted, that is, the time-series sampling data of the target power grid model is extracted at fixed time intervals. When the anchor point fluctuation margin is greater than or equal to the preset anchor point fluctuation margin, a time-series segmented analysis is adopted, that is, the entire time axis is divided into several continuous and non-overlapping time periods according to the change points or state switching density of the voltage fluctuation intensity. Within each segment, the sampling step size or analysis parameters are adaptively adjusted according to the fluctuation intensity of that segment to accurately capture the sudden change process and avoid information loss during periods of intense fluctuation in fixed-period sampling.

[0042] Specifically, the target power grid models in a healthy state are screened based on the descending order of time-series discreteness to obtain a second preset number of target power grid models, denoted as the second priority sequence; The time-series dispersion is determined based on the voltage fluctuation amplitude, power fluctuation amplitude, and state switching frequency.

[0043] In this embodiment of the invention, the time-series dispersion is determined based on voltage fluctuation amplitude, power fluctuation amplitude, and state switching frequency. The time-series dispersion is obtained by weighting and summing the difference between the maximum and minimum voltage values, the peak-valley difference of active or reactive power, and the number of state switching times of circuit breakers and protection devices per unit time within a preset time window. It can be understood that the time-series dispersion is used to determine the severity and complexity of the fluctuations in the operating state of the target power grid model on the time axis. A larger time-series dispersion indicates a more unstable and frequent fluctuation in the model's operating state, requiring priority for fault simulation. Therefore, the higher the user's requirements for power grid operation stability, the smaller the preset time-series dispersion should be. Preferably, the time-series dispersion of qualified historical operating data is collected, outliers are removed, and the average value is calculated, resulting in a value of 0.35, which is denoted as the preset time-series dispersion.

[0044] Specifically, the topology network corresponding to the target power grid model has substations, buses, generating equipment, load equipment and switching equipment as nodes, and transmission lines and transformer connection strength as edges.

[0045] Specifically, the first preset quantity and the second preset quantity are determined based on the fault simulation coverage rate; The first preset quantity and the second preset quantity are positively correlated with the fault simulation coverage rate; the first preset quantity is greater than the second preset quantity.

[0046] In this embodiment of the invention, the first preset quantity is used to determine the maximum simulation capacity of the system for high-risk models that are in a fault state and have a high degree of topological fault coupling, so as to ensure that critical fault states are covered first. The second preset quantity is used to determine the supplementary simulation capacity of the system for periodic impact risk models, so as to cover potential faults under periodic electrical impacts, and the supplementary simulation capacity of the system for fluctuating operating condition models, so as to cover abnormal scenarios under frequent fluctuations in operating conditions. It can be understood that the first preset quantity and the second preset quantity are determined based on the fault simulation coverage rate, which is the proportion of historical fault events and potential fault states covered by the selected target power grid model. The first preset quantity and the second preset quantity are positively correlated with the fault simulation coverage rate.

[0047] In this embodiment of the invention, the preset values ​​can be obtained through simulation. For example, for the first preset quantity and the second preset quantity, fault scenario data of the target power grid model under different operating conditions are collected, and each model is sorted according to the topological fault coupling degree. The fault simulation coverage rate under different combinations of preset quantities is calculated, and fault simulation tests are conducted with different first and second preset quantities. Using fault simulation coverage rate and simulation resource consumption as evaluation indicators, the value corresponding to the coverage rate reaching the saturation point and the inflection point of resource consumption increase is selected as the preset value. Simulation results show that when the first preset quantity is 30% of the total number of target power grid models and the second preset quantity is 10% of the total number of target power grid models, the overall fault simulation coverage rate can reach 92%, which reduces resource consumption by about 45% compared with full coverage simulation, and the coverage rate of high-risk fault states remains at 100%, resulting in the best overall performance.

[0048] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0049] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for simulating power grid equipment faults based on digital twins, characterized in that, include: Based on the load stability corresponding to the target power grid model, determine whether to change the deconstruction method from structural analysis to dynamic response analysis. Based on the deconstruction method, deconstruct the target power grid model to obtain the power grid sub-model, and obtain the node complexity correlation based on the power grid sub-model. Determine whether the target power grid model is in a fault state based on the node complexity correlation; Based on the descending order of topological fault coupling degree, target power grid models in fault state are screened to obtain a first preset number of target power grid models, which is denoted as the first priority sequence. Determine whether to adjust the fault analysis benchmark from periodic analysis to volatility analysis based on the time series volatility. In the periodic analysis, the target power grid models in a healthy state are screened based on the descending order of the periodic impact coefficient to obtain a second preset number of target power grid models, which is denoted as the second priority sequence. In fluctuation analysis, the analysis benchmark is determined based on the anchor point fluctuation margin to determine whether it is adjusted from a fixed time period to a time segment. The target power grid models in a healthy state are screened based on the descending order of the time series dispersion to obtain a second preset number of target power grid models, which is denoted as the second priority sequence. Power grid equipment fault simulation is performed sequentially based on the first priority sequence and the second priority sequence corresponding to the target power grid model.

2. The method for simulating power grid equipment faults based on digital twins according to claim 1, characterized in that, For the condition that the load stability is greater than the preset load stability, the deconstruction method is determined to be structural analysis; For conditions where the load stability is less than or equal to the preset load stability, the solution method is determined to be dynamic response analysis; The load stability is determined based on the node static voltage.

3. The method for simulating power grid equipment faults based on digital twins according to claim 2, characterized in that, If the node complexity correlation is greater than the preset node complexity correlation, the target power grid model is determined to be in a fault state. The topology fault coupling degree of the target power grid model is determined based on the electrical distance between nodes and the frequency of historical fault associations.

4. The method for simulating power grid equipment faults based on digital twins according to claim 3, characterized in that, The node complexity correlation is determined based on the node degree and the port equivalent impedance. The node complexity correlation is positively correlated with the node degree and the port equivalent impedance.

5. The method for simulating power grid equipment faults based on digital twins according to claim 4, characterized in that, If the node complexity correlation is less than or equal to the preset node complexity correlation, the target power grid model is determined to be in a healthy state. For the condition that the time series fluctuation is less than the preset time series fluctuation, the fault analysis benchmark is determined to be periodic analysis; For the condition that the time series volatility is greater than or equal to the preset time series volatility, the fault analysis benchmark is determined to be volatility analysis; The timing fluctuation is determined based on the state switching frequency.

6. The method for simulating power grid equipment faults based on digital twins according to claim 5, characterized in that, The periodic impact coefficient is determined based on the maximum current impact amplitude and rated current of the node within the period; The periodic impact coefficient is positively correlated with the maximum current impact amplitude.

7. The method for simulating power grid equipment faults based on digital twins according to claim 6, characterized in that, In the fluctuation analysis, for the condition that the anchor point fluctuation margin is less than the preset anchor point fluctuation margin, the analysis is determined to be based on a fixed time period. For the condition that the anchor point volatility margin is greater than or equal to the preset anchor point volatility margin, it is determined that the analysis is based on time series segmentation. The anchor point fluctuation margin is determined based on the variance and average voltage of the voltage time series data at the preset anchor point.

8. The method for simulating power grid equipment faults based on digital twins according to claim 7, characterized in that, The time-series dispersion is determined based on the voltage fluctuation amplitude, power fluctuation amplitude, and state switching frequency; The time-series dispersion is positively correlated with voltage fluctuation amplitude, power fluctuation amplitude, and state switching frequency.

9. The method for simulating power grid equipment faults based on digital twins according to claim 8, characterized in that, The topology network corresponding to the target power grid model has substations, buses, generating equipment, load equipment and switching equipment as nodes, and transmission lines and transformer connection strength as edges.

10. The method for simulating power grid equipment faults based on digital twins according to claim 9, characterized in that, The first preset quantity and the second preset quantity are determined based on the fault simulation coverage rate; The first preset quantity, the second preset quantity, and the fault simulation coverage rate are positively correlated. The first preset quantity is greater than the second preset quantity.

Images