A digital twin prediction and simulation system and method for power grids that combines historical and real-time data

Abstract

This invention discloses a digital twin prediction and simulation system and method for power grids that combines historical and real-time data, belonging to the field of power grid simulation technology. The invention acquires historical attribute data and historical operation data of the power grid; constructs a temperature feature fitting model and a current prediction model based on the historical equipment data packets of each device; analyzes the risk index of the devices by combining the temperature feature fitting model and the current prediction model; redistributes power flow on the corrected power grid diagram; analyzes the propagation probability between overloaded routes and adjacent devices; and simulates power flow redistribution on the power grid diagram based on a candidate scheduling strategy set. All candidate scheduling strategies that simultaneously satisfy safety margin verification and temperature rise change verification are combined into an effective strategy set, and the effective strategy set and corresponding risk mitigation rates are transmitted to the system administrator. This improves the accuracy of risk assessment and enhances decision-making efficiency and quality.

Description

Technical Field

[0001] This invention relates to the field of power grid simulation technology, specifically to a power grid digital twin prediction simulation system and method that combines historical and real-time data. Background Technology

[0002] With the increasing proportion of renewable energy connected to the grid, the power system is undergoing a profound transformation towards a new type of power system, facing many new challenges, especially the uncertainty of fault risks, which poses a significant challenge to power system maintenance. Digital twin technology, developed based on next-generation information technology, has unique advantages in addressing this transformation and is an important means of solving and developing new power systems.

[0003] Traditional power grid risk prediction methods heavily rely on static models and isolated data sources, making it difficult to meet the needs of dynamic security assessments of complex power grids. Risk prediction models are often based on extrapolation from a single historical trend, failing to deeply analyze the periodic load fluctuation characteristics in current changes, and failing to establish a precise coupling relationship between equipment heating and electrical load, as well as the cumulative effects over time. Strategy verification relies on manual offline calculations or simplified simulation tools, which are time-consuming and have narrow coverage, making it difficult to respond promptly to real-time risk evolution.

[0004] Therefore, this invention discloses a power grid digital twin prediction simulation system and method that combines historical and real-time data to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a digital twin prediction and simulation system and method for power grids that combines historical and real-time data, in order to solve the problems raised in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a power grid digital twin prediction and simulation method combining historical and real-time data, the method comprising the following steps:

[0007] S1: Obtain historical attribute data and historical operation data of the power grid; align the historical operation data with timestamps and calibrate the equipment temperature; combine the historical attribute data, historical operation data and equipment calibration temperature to generate equipment data packages for each device;

[0008] S2: Based on the historical equipment data packages of each device, construct a temperature feature fitting model and a current prediction model; combine the temperature feature fitting model and the current prediction model to analyze the risk index of the equipment;

[0009] S3: Based on the predicted risk index of each device, the basic power grid diagram is corrected, and the power flow is redistributed in the corrected power grid diagram; the propagation probability between overloaded routes and adjacent devices is analyzed.

[0010] S4: Based on the candidate scheduling strategy set, perform power flow redistribution simulation on the power grid diagram, and form an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification. Then, transmit the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

[0011] According to the above scheme, S1 includes the following:

[0012] S101: The historical attribute data includes topology data and load data; the topology data includes the topology and connection relationships between buses, lines, and power grid equipment; the load data includes active and reactive load data of each device; the historical operating data includes SCADA data, PMU data, equipment infrared temperature, and meteorological data; the SCADA data includes the active power, reactive power, voltage, and current of each device; the PMU data includes phasors and voltage phase angles; the meteorological data includes ambient temperature, humidity, and wind speed.

[0013] For PMU data acquired at non-SCADA data times, a linear interpolation algorithm is used to align the PMU data to the SCADA data times with the smallest time interval; the non-SCADA data times refer to the PMU data acquisition times that are different from the SCADA data acquisition times.

[0014] S102: Analyze the device calibration temperature based on the infrared surface temperature and ambient temperature of the device. The device calibration temperature is equal to the infrared surface temperature plus the correction temperature. The correction temperature is equal to the heat dissipation coefficient multiplied by the difference between the ambient temperature and the standard temperature. The heat dissipation coefficient and the standard temperature are preset constants.

[0015] The normalized topology data, load data, SCADA data, calibration temperature and meteorological data are packaged according to their respective equipment to generate equipment data packages.

[0016] This invention uses equipment calibration temperature to effectively eliminate the influence of ambient temperature fluctuations on infrared temperature measurement of the equipment surface, obtaining temperature values ​​that better reflect the true heating state inside the equipment, and significantly improving the accuracy of subsequent temperature models and risk assessments.

[0017] According to the above scheme, S2 includes the following:

[0018] S201: Extract historical device data packets within a preset historical time window; construct a temperature feature fitting model: Tem=T amb +α×I 2 ×[1-exp(-βt)]÷β+(Tem 0 -T amb) × exp(-βt); where α and β represent the fitting coefficients, t represents the time variable, and T amb I represents the independent variable of ambient temperature, and Tem represents the independent variable of current; Tem represents the dependent variable of equipment calibration temperature. 0 The device calibration temperature represents the initial moment within the historical time window; the least squares method is used to calculate and solve for α and β in the temperature feature fitting model.

[0019] S202: Constructing a current fitting model: I`=I 0 +kt; where t represents the independent variable of time; I` represents the dependent variable of current, and k represents the fitting coefficient; I 0 This represents the current value at the initial moment within the historical time window; the fitting coefficient k in the current fitting model is solved using the least squares method.

[0020] Extract the current curve function from a preset historical time window; the current curve function is a function of current changing over time; after eliminating the DC component from the current curve function, perform a Fast Fourier Transform; select the amplitude, frequency, and phase of the first three dominant frequency components to reconstruct the current curve function.

[0021] ;

[0022] Where I' represents the reduction current, A n f represents the amplitude of the nth dominant frequency component. n φ represents the frequency of the nth dominant frequency component. n This represents the phase of the nth dominant frequency component;

[0023] A current prediction model is constructed by combining the current fitting model and the current curve function:

[0024] ;

[0025] Where I* represents the predicted current;

[0026] S203: Analyze the equipment's risk index R by combining a temperature characteristic fitting model and a current prediction model: R = [(Tem - Tem)] safe )×I*] / [(Tem max -Tem safe )×I rated ]; among them, Tem safe Indicates the safe temperature threshold for equipment calibration, Tem max Indicates the equipment calibration temperature limit threshold; I rated Indicates the rated current.

[0027] The current prediction model constructed in this invention combines linear trends and periodic fluctuations; it can capture the patterns of current changes more comprehensively and accurately, providing high-quality current input for temperature models and risk assessments; the risk index integrates the key risk factors of the equipment, quantifies the degree to which the equipment deviates from a safe operating state, and the higher the value, the higher the risk, providing an intuitive and comparable quantitative basis for the subsequent identification of high-risk equipment.

[0028] According to the above scheme, S3 includes the following:

[0029] S301: Construct a basic power grid graph G=(V,E) from the twin platform; where V represents the node set and E represents the edge set; the nodes include transformer equipment, transmission line equipment and auxiliary equipment; the weight between any two nodes is equal to the apparent power transmitted between the two nodes divided by the maximum transmission power allowed by the system;

[0030] S302: Predict the risk index of each device; if the risk index of a device is greater than the risk index threshold, it is marked as a high-risk device and a set of high-risk devices is generated; remove each high-risk device from the set of high-risk devices from the basic power grid diagram to form a modified power grid diagram; perform power flow redistribution based on the modified power grid diagram and the DC power flow model, and analyze the current of each line.

[0031] S303: If the ratio of the line current to the corresponding rated line current is greater than the overload factor, it is marked as an overloaded line, where the overload factor is a preset constant; analyze the propagation probability between the overloaded line and each adjacent device:

[0032] ;

[0033] Among them, V i This represents the real-time voltage of overloaded line i, V. j I represents the real-time voltage of adjacent device j. i I represents the real-time current of overloaded line i. j V represents the real-time current of adjacent device j. rated The value represents the rated voltage; μ represents the propagation coefficient, which is a preset constant; if the propagation probability between the overloaded route and the adjacent equipment is greater than the propagation probability threshold, the adjacent equipment is marked as a high-risk equipment; and the propagation probability matrix is ​​updated.

[0034] Example 1: In this example, if the risk index of transformer k is greater than the risk index threshold, then transformer k is removed from the basic power grid diagram; after power flow redistribution, the I of line l... l / I rated =1.08; In this embodiment, the overload factor is 1.05; Therefore, line 1 is marked as an overloaded line;

[0035] Calculate the propagation probability of line l to adjacent device m:

[0036] P （l，m） =(|115.2-112.4| / 110)×exp(-0.2|0.95-0.98|)≈0.025×0.994=2.485%;

[0037] In this embodiment, the propagation probability threshold is 2%; therefore, adjacent devices marked m are denoted as high-risk devices.

[0038] S304: Repeat S302 and S303 until no more high-risk devices appear; output the final propagation probability matrix and the historical maximum value of the propagation probability, and mark the high-risk path k→l→m in the propagation probability matrix.

[0039] This invention proactively eliminates high-risk devices based on predicted risk indices, forming a corrected power grid diagram. It simulates scenarios of device failure or forced shutdown, ensuring that subsequent power flow redistribution results accurately reflect the state after network structure changes. It realistically captures how initial overloads or faults gradually spread through power grid topology and power flow relationships. It comprehensively identifies potential risk points, avoiding the possibility of overlooking indirect risk devices in a one-time analysis, and ensuring that all potentially affected high-risk devices are identified.

[0040] According to the above scheme, S4 includes the following:

[0041] S401: Extract a preset set of candidate scheduling strategies from the twin platform. Each candidate scheduling strategy includes generation adjustment amount and grid topology switching method. The grid topology switching method includes branch disconnection, branch closing and bus reconnection. Perform power flow redistribution simulation on the grid diagram after adopting each candidate scheduling strategy.

[0042] S402: Analyze the safety margin of each line and the calibration temperature change of each device after power flow redistribution; the safety margin is equal to the ratio of the maximum current of the line minus the line current to the maximum current.

[0043] All candidate scheduling strategies that simultaneously satisfy both safety margin verification and temperature rise change verification are combined into an effective strategy set, and the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy are transmitted to the system administrator; the safety margin verification is that the minimum safety margin corresponding to the candidate scheduling strategy is greater than a preset safety margin threshold; the temperature rise change verification is that the maximum value of the device's calibration temperature change is less than a preset temperature change threshold; the risk mitigation rate is equal to the ratio of one to the maximum element in the final propagation probability matrix before and after the implementation of the candidate scheduling strategy.

[0044] This invention performs automated power flow simulation on a digital twin, efficiently evaluating the actual effects of a large number of strategies. It provides system administrators with a set of effective strategies that have undergone rigorous verification and their corresponding quantitative performance indicators, enabling administrators to quickly focus on feasible solutions, select the best option based on its effectiveness, and improve decision-making efficiency and quality.

[0045] Another aspect of this application provides a power grid digital twin prediction simulation system that combines historical and real-time data. The system is applied to the above-mentioned power grid digital twin prediction simulation method that combines historical and real-time data. The system includes a power grid data integration module, an equipment risk analysis module, a propagation probability simulation module, and a scheduling strategy analysis module.

[0046] The power grid data integration module is used to acquire historical attribute data and historical operation data of the power grid; to align the historical operation data with timestamps and to calibrate the equipment temperature; and to generate equipment data packages for each device by combining the historical attribute data, historical operation data and equipment calibration temperature.

[0047] The equipment risk analysis module is used to construct a temperature feature fitting model and a current prediction model based on the historical equipment data packages of each device; and to analyze the risk index of the equipment by combining the temperature feature fitting model and the current prediction model.

[0048] The propagation probability simulation module is used to correct the basic power grid diagram based on the predicted risk index of each device, redistribute the power flow in the corrected power grid diagram, and analyze the propagation probability between overloaded routes and adjacent devices.

[0049] The scheduling strategy analysis module is used to perform power flow redistribution simulation on the power grid diagram based on the candidate scheduling strategy set. It forms an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification, and transmits the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

[0050] According to the above scheme, the power grid data integration module includes a time synchronization unit and a data fusion unit;

[0051] The time synchronization unit is used to align the PMU data acquired at non-SCADA data times using a linear interpolation algorithm to the SCADA data times with the smallest time interval.

[0052] The data fusion unit is used to package the normalized topology data, load data, SCADA data, calibration temperature and meteorological data according to their respective equipment to generate equipment data packages.

[0053] According to the above scheme, the equipment risk analysis module includes a model building unit and a risk index analysis unit;

[0054] The model building unit is used to extract historical device data packets within a preset historical time window, build a temperature feature fitting model; build a current fitting model; extract the current curve function within the preset historical time window, perform a fast Fourier transform on the current curve function after eliminating the DC component; select the amplitude, frequency and phase of the first three main frequency components to restore the current curve function; and combine the current fitting model and the current curve function to build a current prediction model.

[0055] The risk index analysis unit is used to analyze the risk index of the equipment by combining a temperature feature fitting model and a current prediction model.

[0056] According to the above scheme, the propagation probability simulation module includes a power flow redistribution unit and a propagation probability analysis unit;

[0057] The power flow redistribution unit is used to construct a basic power grid map from the twin platform; predict the risk index of each device; if the risk index of a device is greater than the risk index threshold, it is marked as a high-risk device and a high-risk device set is generated; each high-risk device in the high-risk device set is removed from the basic power grid map to form a modified power grid map; power flow redistribution is performed based on the modified power grid map and combined with the DC power flow model, and the current of each line is analyzed.

[0058] If the ratio of the line current to the corresponding rated line current is greater than the overload factor, the propagation probability analysis unit marks it as an overloaded line; it analyzes the propagation probability between the overloaded line and each adjacent device; if the propagation probability between the overloaded line and the adjacent device is greater than the propagation probability threshold, the adjacent device is marked as a high-risk device; and the propagation probability matrix is ​​updated; the power flow redistribution, overload analysis, and propagation probability analysis are repeated until no more high-risk devices appear; the final propagation probability matrix and the historical maximum value of the propagation probability are output.

[0059] According to the above scheme, the scheduling strategy analysis module includes a strategy fitting unit and a strategy pre-selection unit;

[0060] The strategy fitting unit is used to extract a preset set of candidate scheduling strategies from the twin platform. Each candidate scheduling strategy includes generation adjustment amount and grid topology switching mode; and to perform power flow redistribution simulation on the grid diagram after adopting each candidate scheduling strategy.

[0061] The strategy pre-selection unit is used to analyze the safety margin of each line and the calibration temperature change value of each device after power flow redistribution; it forms an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification, and transmits the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

[0062] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention effectively eliminates the influence of ambient temperature fluctuations on infrared thermography of the equipment surface by using equipment calibration temperature, obtaining temperature values ​​that better reflect the true internal heating state of the equipment, significantly improving the accuracy of subsequent temperature models and risk assessments; the current prediction model constructed by this invention combines linear trends and periodic fluctuations, enabling it to more comprehensively and accurately capture the patterns of current changes, providing high-quality current input for temperature models and risk assessments; the risk index integrates key risk factors of the equipment, quantifying the degree to which the equipment deviates from a safe operating state, with higher values ​​indicating higher risk, providing an intuitive and comparable quantitative basis for subsequent identification of high-risk equipment; this invention proactively uses the predicted risk index... By eliminating high-risk equipment and generating a revised power grid diagram, this invention simulates scenarios of equipment failure or forced shutdown, ensuring that subsequent power flow redistribution results accurately reflect the state after network structure changes. It realistically captures how initial overloads or faults gradually spread through power grid topology and power flow relationships; comprehensively identifies potential risk points, avoiding the possibility of overlooking indirect risk equipment in a single analysis, and ensuring the identification of all potentially affected high-risk equipment. This invention performs automated power flow simulation on a digital twin, efficiently evaluating the actual effects of numerous strategies. It provides system administrators with a rigorously validated set of effective strategies and their corresponding quantitative performance indicators, enabling administrators to quickly focus on feasible solutions, select the best option based on its effectiveness, and improve decision-making efficiency and quality. Attached Figure Description

[0063] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0064] Figure 1 This is a flowchart illustrating the power grid digital twin prediction and simulation method that combines historical and real-time data according to the present invention.

[0065] Figure 2 This is a schematic diagram of the power grid digital twin prediction simulation system that combines historical and real-time data according to the present invention. Detailed Implementation

[0066] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] Please see Figure 1 This invention provides a technical solution: a power grid digital twin prediction and simulation method combining historical and real-time data, which includes the following steps:

[0068] S1: Obtain historical attribute data and historical operation data of the power grid; align the historical operation data with timestamps and calibrate the equipment temperature; combine the historical attribute data, historical operation data and equipment calibration temperature to generate equipment data packages for each device;

[0069] According to the above scheme, S1 includes the following:

[0070] S101: The historical attribute data includes topology data and load data; the topology data includes the topology and connection relationships between buses, lines, and power grid equipment; the load data includes active and reactive load data of each device; the historical operating data includes SCADA data, PMU data, equipment infrared temperature, and meteorological data; the SCADA data includes the active power, reactive power, voltage, and current of each device; the PMU data includes phasors and voltage phase angles; the meteorological data includes ambient temperature, humidity, and wind speed.

[0071] For PMU data acquired at non-SCADA data times, a linear interpolation algorithm is used to align the PMU data to the SCADA data time with the smallest time interval.

[0072] S102: Analyze the device calibration temperature based on the infrared surface temperature and ambient temperature of the device. The device calibration temperature is equal to the infrared surface temperature plus the correction temperature. The correction temperature is equal to the heat dissipation coefficient multiplied by the difference between the ambient temperature and the standard temperature. The heat dissipation coefficient and the standard temperature are preset constants.

[0073] The normalized topology data, load data, SCADA data, calibration temperature and meteorological data are packaged according to their respective equipment to generate equipment data packages.

[0074] S2: Based on the historical equipment data packages of each device, construct a temperature feature fitting model and a current prediction model; combine the temperature feature fitting model and the current prediction model to analyze the risk index of the equipment;

[0075] According to the above scheme, S2 includes the following:

[0076] S201: Extract historical device data packets within a preset historical time window; construct a temperature feature fitting model: Tem=T amb +α×I 2 ×[1-exp(-βt)]÷β+(Tem 0 -T amb ) × exp(-βt); where α and β represent the fitting coefficients, t represents the time variable, and T ambI represents the independent variable of ambient temperature, and Tem represents the independent variable of current; Tem represents the dependent variable of equipment calibration temperature. 0 The device calibration temperature represents the initial moment within the historical time window; the least squares method is used to calculate and solve for α and β in the temperature feature fitting model.

[0077] S202: Constructing a current fitting model: I`=I 0 +kt; where t represents the independent variable of time; I` represents the dependent variable of current, and k represents the fitting coefficient; I 0 This represents the current value at the initial moment within the historical time window; the fitting coefficient k in the current fitting model is solved using the least squares method.

[0078] Extract the current curve function from a preset historical time window; the current curve function is a function of current changing over time; after eliminating the DC component from the current curve function, perform a Fast Fourier Transform; select the amplitude, frequency, and phase of the first three dominant frequency components to reconstruct the current curve function.

[0079] ;

[0080] Where I' represents the reduction current, A n f represents the amplitude of the nth dominant frequency component. n φ represents the frequency of the nth dominant frequency component. n This represents the phase of the nth dominant frequency component;

[0081] A current prediction model is constructed by combining the current fitting model and the current curve function:

[0082] ;

[0083] Where I* represents the predicted current;

[0084] S203: Analyze the equipment's risk index R by combining a temperature characteristic fitting model and a current prediction model: R = [(Tem - Tem)] safe )×I*] / [(Tem max -Tem safe )×I rated ]; among them, Tem safe Indicates the safe temperature threshold for equipment calibration, Tem max Indicates the equipment calibration temperature limit threshold; I rated Indicates the rated current.

[0085] S3: Based on the predicted risk index of each device, the basic power grid diagram is corrected, and the power flow is redistributed in the corrected power grid diagram; the propagation probability between overloaded routes and adjacent devices is analyzed.

[0086] According to the above scheme, S3 includes the following:

[0087] S301: Construct a basic power grid graph G=(V,E) from the twin platform; where V represents the node set and E represents the edge set; the nodes include transformer equipment, transmission line equipment and auxiliary equipment; the weight between any two nodes is equal to the apparent power transmitted between the two nodes divided by the maximum transmission power allowed by the system;

[0088] S302: Predict the risk index of each device; if the risk index of a device is greater than the risk index threshold, it is marked as a high-risk device and a set of high-risk devices is generated; remove each high-risk device from the set of high-risk devices from the basic power grid diagram to form a modified power grid diagram; perform power flow redistribution based on the modified power grid diagram and the DC power flow model, and analyze the current of each line.

[0089] S303: If the ratio of the line current to the corresponding rated line current is greater than the overload factor, it is marked as an overloaded line, where the overload factor is a preset constant; analyze the propagation probability between the overloaded line and each adjacent device:

[0090] ;

[0091] Among them, V i This represents the real-time voltage of overloaded line i, V. j I represents the real-time voltage of adjacent device j. i I represents the real-time current of overloaded line i. j V represents the real-time current of adjacent device j. rated The value represents the rated voltage; μ represents the propagation coefficient, which is a preset constant; if the propagation probability between the overloaded route and the adjacent equipment is greater than the propagation probability threshold, the adjacent equipment is marked as a high-risk equipment; and the propagation probability matrix is ​​updated.

[0092] S304: Repeat S302 and S303 until no more high-risk devices appear; output the final propagation probability matrix and the historical maximum value of the propagation probability.

[0093] S4: Based on the candidate scheduling strategy set, perform power flow redistribution simulation on the power grid diagram, and form an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification. Then, transmit the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

[0094] According to the above scheme, S4 includes the following:

[0095] S401: Extract a preset set of candidate scheduling strategies from the twin platform. Each candidate scheduling strategy includes generation adjustment amount and grid topology switching method. The grid topology switching method includes branch disconnection, branch closing and bus reconnection. Perform power flow redistribution simulation on the grid diagram after adopting each candidate scheduling strategy.

[0096] S402: Analyze the safety margin of each line and the calibration temperature change of each device after power flow redistribution; the safety margin is equal to the ratio of the maximum current of the line minus the line current to the maximum current.

[0097] All candidate scheduling strategies that simultaneously satisfy both safety margin verification and temperature rise change verification are combined into an effective strategy set, and the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy are transmitted to the system administrator; the safety margin verification is that the minimum safety margin corresponding to the candidate scheduling strategy is greater than a preset safety margin threshold; the temperature rise change verification is that the maximum value of the device's calibration temperature change is less than a preset temperature change threshold; the risk mitigation rate is equal to the ratio of one to the maximum element in the final propagation probability matrix before and after the implementation of the candidate scheduling strategy.

[0098] Please see Figure 2 The present invention provides a technical solution: a power grid digital twin prediction and simulation system that combines historical and real-time data. The system includes a power grid data integration module, an equipment risk analysis module, a propagation probability simulation module, and a scheduling strategy analysis module.

[0099] The power grid data integration module is used to acquire historical attribute data and historical operation data of the power grid; to align the historical operation data with timestamps and to calibrate the equipment temperature; and to generate equipment data packages for each device by combining the historical attribute data, historical operation data and equipment calibration temperature.

[0100] The equipment risk analysis module is used to construct a temperature feature fitting model and a current prediction model based on the historical equipment data packages of each device; and to analyze the risk index of the equipment by combining the temperature feature fitting model and the current prediction model.

[0101] The propagation probability simulation module is used to correct the basic power grid diagram based on the predicted risk index of each device, redistribute the power flow in the corrected power grid diagram, and analyze the propagation probability between overloaded routes and adjacent devices.

[0102] The scheduling strategy analysis module is used to perform power flow redistribution simulation on the power grid diagram based on the candidate scheduling strategy set. It forms an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification, and transmits the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

[0103] According to the above scheme, the power grid data integration module includes a time synchronization unit and a data fusion unit;

[0104] The time synchronization unit is used to align the PMU data acquired at non-SCADA data times using a linear interpolation algorithm to the SCADA data times with the smallest time interval.

[0105] The data fusion unit is used to package the normalized topology data, load data, SCADA data, calibration temperature and meteorological data according to their respective equipment to generate equipment data packages.

[0106] According to the above scheme, the equipment risk analysis module includes a model building unit and a risk index analysis unit;

[0107] The model building unit is used to extract historical device data packets within a preset historical time window, build a temperature feature fitting model; build a current fitting model; extract the current curve function within the preset historical time window, perform a fast Fourier transform on the current curve function after eliminating the DC component; select the amplitude, frequency and phase of the first three main frequency components to restore the current curve function; and combine the current fitting model and the current curve function to build a current prediction model.

[0108] The risk index analysis unit is used to analyze the risk index of the equipment by combining a temperature feature fitting model and a current prediction model.

[0109] According to the above scheme, the propagation probability simulation module includes a power flow redistribution unit and a propagation probability analysis unit;

[0110] The power flow redistribution unit is used to construct a basic power grid map from the twin platform; predict the risk index of each device; if the risk index of a device is greater than the risk index threshold, it is marked as a high-risk device and a high-risk device set is generated; each high-risk device in the high-risk device set is removed from the basic power grid map to form a modified power grid map; power flow redistribution is performed based on the modified power grid map and combined with the DC power flow model, and the current of each line is analyzed.

[0111] If the ratio of the line current to the corresponding rated line current is greater than the overload factor, the propagation probability analysis unit marks it as an overloaded line; it analyzes the propagation probability between the overloaded line and each adjacent device; if the propagation probability between the overloaded line and the adjacent device is greater than the propagation probability threshold, the adjacent device is marked as a high-risk device; and the propagation probability matrix is ​​updated; the power flow redistribution, overload analysis, and propagation probability analysis are repeated until no more high-risk devices appear; the final propagation probability matrix and the historical maximum value of the propagation probability are output.

[0112] According to the above scheme, the scheduling strategy analysis module includes a strategy fitting unit and a strategy pre-selection unit;

[0113] The strategy fitting unit is used to extract a preset set of candidate scheduling strategies from the twin platform. Each candidate scheduling strategy includes generation adjustment amount and grid topology switching mode; and to perform power flow redistribution simulation on the grid diagram after adopting each candidate scheduling strategy.

[0114] The strategy pre-selection unit is used to analyze the safety margin of each line and the calibration temperature change value of each device after power flow redistribution; it forms an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification, and transmits the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

[0115] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0116] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A power grid digital twin prediction and simulation method combining historical and real-time data, characterized in that: The method includes the following steps: S1: Obtain historical attribute data and historical operation data of the power grid; align the historical operation data with timestamps and calibrate the equipment temperature; combine the historical attribute data, historical operation data and equipment calibration temperature to generate equipment data packages for each device; S2: Based on the historical equipment data packages of each device, construct a temperature feature fitting model and a current prediction model; combine the temperature feature fitting model and the current prediction model to analyze the risk index of the equipment; S3: Based on the predicted risk index of each device, the basic power grid diagram is corrected, and the power flow is redistributed in the corrected power grid diagram; the propagation probability between overloaded routes and adjacent devices is analyzed. S4: Based on the candidate scheduling strategy set, perform power flow redistribution simulation on the power grid diagram, and form an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification. Then, transmit the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

2. The power grid digital twin prediction simulation method combining historical and real-time data according to claim 1, characterized in that: S1 contains the following: S101: The historical attribute data includes topology data and load data; the topology data includes the topology and connection relationships between buses, lines, and power grid equipment; the load data includes active and reactive load data of each device; the historical operating data includes SCADA data, PMU data, equipment infrared temperature, and meteorological data; the SCADA data includes the active power, reactive power, voltage, and current of each device; the PMU data includes phasors and voltage phase angles; the meteorological data includes ambient temperature, humidity, and wind speed. For PMU data acquired at non-SCADA data times, a linear interpolation algorithm is used to align the PMU data to the SCADA data time with the smallest time interval. S102: Analyze the device calibration temperature based on the infrared surface temperature and ambient temperature of the device. The device calibration temperature is equal to the infrared surface temperature plus the correction temperature. The correction temperature is equal to the heat dissipation coefficient multiplied by the difference between the ambient temperature and the standard temperature. The heat dissipation coefficient and the standard temperature are preset constants. The normalized topology data, load data, SCADA data, calibration temperature and meteorological data are packaged according to their respective equipment to generate equipment data packages.

3. The power grid digital twin prediction simulation method combining historical and real-time data according to claim 2, characterized in that: S2 includes the following: S201: Extract historical device data packets within a preset historical time window; construct a temperature feature fitting model: Tem=T amb +α×I 2 ×[1-exp(-βt)]÷β+(Tem 0 -T amb ) × exp(-βt); where α and β represent the fitting coefficients, t represents the time variable, and T amb I represents the independent variable of ambient temperature, and Tem represents the independent variable of current; Tem represents the dependent variable of equipment calibration temperature. 0 The device calibration temperature represents the initial moment within the historical time window; the least squares method is used to calculate and solve for α and β in the temperature feature fitting model. S202: Constructing a current fitting model: I`=I 0 +kt; where t represents the independent variable of time; I` represents the dependent variable of current, and k represents the fitting coefficient; I 0 This represents the current value at the initial moment within the historical time window; the fitting coefficient k in the current fitting model is solved using the least squares method. Extract the current curve function from the preset historical time window. The current curve function is a function of current changing with time. After eliminating the DC component of the current curve function, perform a fast Fourier transform. Select the amplitude, frequency and phase of the first three main frequency components to restore the current curve function. A current prediction model is constructed by combining a current fitting model and a current curve function. S203: Analyze the equipment's risk index R by combining a temperature characteristic fitting model and a current prediction model: R = [(Tem - Tem)] safe )×I*] / [(Tem max -Tem safe )×I rated ]; among them, Tem safe Indicates the safe temperature threshold for equipment calibration, Tem max Indicates the equipment calibration temperature limit threshold; I rated Indicates the rated current.

4. The power grid digital twin prediction simulation method combining historical and real-time data according to claim 3, characterized in that: S3 includes the following: S301: Construct a basic power grid diagram G=(V,E) from a twin platform; Where V represents the node set and E represents the edge set; the nodes include transformer-type equipment, transmission line-type equipment and auxiliary equipment; the weight between any two nodes is equal to the apparent power transmitted between the two nodes divided by the maximum transmission power allowed by the system; S302: Predict the risk index of each device; If the risk index of a device is greater than the risk index threshold, it is marked as a high-risk device, and a set of high-risk devices is generated. The revised power grid diagram is formed by removing each high-risk device from the set of high-risk devices in the basic power grid diagram. Power flow redistribution is performed based on the modified power grid diagram and the DC power flow model, and the current of each line is analyzed. S303: If the ratio of the line current to the corresponding rated line current is greater than the overload factor, it is marked as an overloaded line, where the overload factor is a preset constant; analyze the propagation probability between the overloaded line and each adjacent device. If the propagation probability between an overloaded route and adjacent devices is greater than the propagation probability threshold, then the adjacent devices are marked as high-risk devices. And update the propagation probability matrix; S304: Repeat S302 and S303 until no more high-risk devices appear; Output the final propagation probability matrix and the historical maximum propagation probability.

5. The power grid digital twin prediction simulation method combining historical and real-time data according to claim 4, characterized in that: S4 includes the following: S401: Extract a preset set of candidate scheduling strategies from the twin platform. Each candidate scheduling strategy includes a power generation adjustment amount and a grid topology switching method. The grid topology switching method includes branch disconnection, branch closing, and bus reconnection. Perform power flow redistribution simulation on the power grid diagram after adopting each candidate scheduling strategy; S402: Analyze the safety margin of each line and the calibration temperature change of each device after power flow redistribution; the safety margin is equal to the ratio of the maximum current of the line minus the line current to the maximum current. All candidate scheduling strategies that simultaneously satisfy the safety margin check and temperature rise change check are combined into an effective strategy set, and the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy are transmitted to the system administrator; the safety margin check is that the minimum safety margin corresponding to the candidate scheduling strategy is greater than the preset safety margin threshold. The temperature rise change verification indicates that the maximum value of the device's calibrated temperature change is less than the preset temperature change threshold; the risk mitigation rate is equal to the ratio of the largest element in the final propagation probability matrix before and after the implementation of the candidate scheduling strategy.

6. A power grid digital twin prediction and simulation system combining historical and real-time data, wherein the system is implemented using the power grid digital twin prediction and simulation method combining historical and real-time data as described in any one of claims 1-5, characterized in that... The system includes a power grid data integration module, an equipment risk analysis module, a propagation probability simulation module, and a scheduling strategy analysis module; The power grid data integration module is used to acquire historical attribute data and historical operation data of the power grid; to align the historical operation data with timestamps and to calibrate the equipment temperature; and to generate equipment data packages for each device by combining the historical attribute data, historical operation data and equipment calibration temperature. The equipment risk analysis module is used to construct a temperature feature fitting model and a current prediction model based on the historical equipment data packages of each device; and to analyze the risk index of the equipment by combining the temperature feature fitting model and the current prediction model. The propagation probability simulation module is used to correct the basic power grid diagram based on the predicted risk index of each device, redistribute the power flow in the corrected power grid diagram, and analyze the propagation probability between overloaded routes and adjacent devices. The scheduling strategy analysis module is used to perform power flow redistribution simulation on the power grid diagram based on the candidate scheduling strategy set. It forms an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification, and transmits the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

7. The power grid digital twin prediction simulation system combining historical and real-time data according to claim 6, characterized in that: The power grid data integration module includes a time synchronization unit and a data fusion unit; The time synchronization unit is used to align the PMU data acquired at non-SCADA data times using a linear interpolation algorithm to the SCADA data times with the smallest time interval. The data fusion unit is used to package the normalized topology data, load data, SCADA data, calibration temperature and meteorological data according to their respective equipment to generate equipment data packages.

8. The power grid digital twin prediction simulation system combining historical and real-time data according to claim 6, characterized in that: The equipment risk analysis module includes a model building unit and a risk index analysis unit; The model building unit is used to extract historical device data packets within a preset historical time window, build a temperature feature fitting model; build a current fitting model; extract the current curve function within the preset historical time window, perform a fast Fourier transform on the current curve function after eliminating the DC component; select the amplitude, frequency and phase of the first three main frequency components to restore the current curve function; and combine the current fitting model and the current curve function to build a current prediction model. The risk index analysis unit is used to analyze the risk index of the equipment by combining a temperature feature fitting model and a current prediction model.

9. The power grid digital twin prediction simulation system combining historical and real-time data according to claim 6, characterized in that: The propagation probability simulation module includes a power flow redistribution unit and a propagation probability analysis unit; The power flow redistribution unit is used to construct a basic power grid map from the twin platform and predict the risk index of each device. If the risk index of a device is greater than the risk index threshold, it is marked as a high-risk device, and a set of high-risk devices is generated. The high-risk devices in the set of high-risk devices are removed from the basic power grid diagram to form a revised power grid diagram; power flow redistribution is performed based on the revised power grid diagram and the DC power flow model, and the current of each line is analyzed. If the ratio of the line current to the corresponding rated line current is greater than the overload factor, the propagation probability analysis unit marks it as an overloaded line; it analyzes the propagation probability between the overloaded line and each adjacent device; if the propagation probability between the overloaded line and the adjacent device is greater than the propagation probability threshold, the adjacent device is marked as a high-risk device. And update the propagation probability matrix; repeat the power flow redistribution, overload analysis and propagation probability analysis until no more high-risk devices appear; Output the final propagation probability matrix and the historical maximum propagation probability.

10. The power grid digital twin prediction simulation system combining historical and real-time data according to claim 6, characterized in that: The scheduling strategy analysis module includes a strategy fitting unit and a strategy pre-selection unit; The strategy fitting unit is used to extract a preset set of candidate scheduling strategies from the twin platform. Each candidate scheduling strategy includes generation adjustment amount and grid topology switching mode; and to perform power flow redistribution simulation on the grid diagram after adopting each candidate scheduling strategy. The strategy pre-selection unit is used to analyze the safety margin of each line and the calibration temperature change value of each device after power flow redistribution; it forms an effective strategy set by combining all candidate scheduling strategies that simultaneously meet the safety margin verification and temperature rise change verification, and transmits the effective strategy set and the risk mitigation rate corresponding to each candidate scheduling strategy to the system administrator.

Images