Closed-loop verification and parameter optimization method for network-type SVG electromagnetic transient model
By combining a hardware-in-the-loop testing platform and intelligent optimization algorithms, the entire process of closed-loop verification and parameter optimization of the network-type SVG electromagnetic transient model was realized. This solved the deviation problem of the model under different operating conditions, improved the accuracy and adaptability of the model, and made it suitable for simulation analysis of new energy power systems.
Patent Information
- Application Number
- CN202611094512.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-07-22
- Publication Date
- 2026-11-10
AI Technical Summary
In existing technologies, the electromagnetic transient model of the grid-type SVG lacks full-process closed-loop verification and parameter optimization, resulting in large deviations in the model under different operating conditions, and failing to effectively cope with transient overvoltage and sub/supersynchronous oscillation problems in new energy power systems.
By building a hardware-in-the-loop testing platform, collecting electrical quantity response data, constructing a corresponding electromagnetic transient model, performing multi-dimensional deviation matrix analysis, and combining sensitivity analysis and intelligent optimization algorithms, key parameters are automatically optimized, achieving closed-loop verification of the entire process from hardware-in-the-loop testing to electromagnetic model simulation.
It significantly improves the accuracy and modeling efficiency of the model, enhances the model's adaptability and stability under different operating conditions, and is suitable for simulation analysis and planning operation of high-proportion new energy power systems.
Smart Images

Figure CN123069633A_ABST
Abstract
Description
Technical Field
[0001] This invention mainly relates to the field of power system simulation and modeling technology, specifically to a closed-loop verification and parameter optimization method for a network-type SVG electromagnetic transient model. Background Technology
[0002] With the rapid development of new energy power generation technologies, a high-proportion new energy power structure has been initially formed. The power electronic power generation equipment in the sending-end systems of new energy bases is numerous, with complex control and multi-timescale coupling, making transient overvoltage and sub- / supersynchronous oscillations increasingly prominent issues. Grid-based reactive power compensation devices (such as grid-based SVG) play a crucial role in improving the system's short-circuit ratio and supporting voltage stability, actively providing voltage support and transient support to the power grid.
[0003] Currently, AC substations in large-scale renewable energy power generation bases are typically equipped with SVG reactive power compensation devices accounting for 25%-40% of the station's capacity. These devices play a crucial role in improving renewable energy transmission power and ensuring and improving power quality. To accurately grasp the electrical and control characteristics of SVG, effectively identify the risks of transient overvoltages and sub / supersynchronous oscillations in the system, and propose effective suppression measures, it is necessary to build an accurate electromagnetic transient model of the SVG.
[0004] However, current modeling methods largely rely on offline simulation, lacking closed-loop verification with actual controller hardware-in-the-loop test data, which can lead to significant deviations in the model under different operating conditions. While existing technologies include methods for electromagnetic transient simulation of grid-connected equipment connected to the power grid, these primarily focus on system-level simulation analysis and do not address the refined verification and automatic parameter optimization of single-machine models. Other parameter identification methods identify control parameters through hardware-in-the-loop testing, but fail to achieve closed-loop verification and iterative optimization under multiple operating conditions. Therefore, there is an urgent need for a method that can achieve closed-loop verification and automatic optimization throughout the entire process from hardware-in-the-loop testing to electromagnetic model simulation, to significantly improve model accuracy and modeling efficiency. Summary of the Invention
[0005] To address the technical problems existing in the prior art, this invention provides a closed-loop verification and parameter optimization method for a network-type SVG electromagnetic transient model that achieves closed-loop verification and automatic optimization throughout the entire process from HIL testing to electromagnetic model simulation, significantly improving model accuracy and modeling efficiency.
[0006] To solve the above-mentioned technical problems, the technical solution proposed by this invention is as follows: A closed-loop verification and parameter optimization method for a network-type SVG electromagnetic transient model includes the following steps: Step S1: Build a hardware-in-the-loop test platform that includes an actual SVG controller and a real-time simulator, set up various operating conditions, collect electrical quantity response data of the SVG grid connection point under each operating condition, and form a benchmark dataset; Step S2: Build an SVG electromagnetic transient model in the electromagnetic transient simulation software that is consistent with the primary topology and control strategy of the hardware-in-the-loop test platform, and run the simulation under the same operating conditions to collect the corresponding simulation response data. Step S3: Divide the simulation response data and the benchmark dataset into multiple intervals in the time domain, calculate the deviation index in each interval, and generate a multidimensional deviation matrix; Step S4: Determine whether each deviation index in the multidimensional deviation matrix is less than a preset threshold. If yes, output the current SVG electromagnetic transient model as the verified model; otherwise, proceed to step S5. Step S5: Identify the key model parameters affecting the deviation through sensitivity analysis, use an intelligent optimization algorithm to automatically optimize the key model parameters, update the SVG electromagnetic transient model, and repeat steps S2 to S4 until all deviation indicators meet the requirements.
[0007] Preferably, the various operating conditions in step S1 include: fault ride-through test condition, continuous fault ride-through test condition, reactive power step test condition, voltage reference value step test condition, grid connection / network mode switching test condition, short-circuit ratio adaptability test condition, frequency adaptability test condition, and phase angle jump test condition.
[0008] Preferably, the fault ride-through test conditions include: three-phase or two-phase symmetrical / asymmetrical faults, with voltage drops or rises of 5%Un, 20%Un, 50%Un, 120%Un, and 130%Un, lasting for 150ms, 625ms, 1214ms, or 10s; the continuous fault ride-through test conditions include: voltage drops to 5%Un for 150ms followed by a rise to 120%Un for 10s without delay, and voltage drops to 5%Un for 150ms followed by a rise to 130%Un for 0.5s without delay; where Un is the rated voltage.
[0009] Preferably, the multi-interval division in step S3 includes: a pre-fault steady-state interval, a fault-interrupted transient interval, a fault-interrupted steady-state interval, a post-fault transient interval, and a post-fault steady-state interval; the deviation index includes: maximum deviation, average deviation, and transient average deviation; the multi-dimensional deviation matrix includes at least voltage amplitude deviation, reactive power deviation, reactive current deviation, and active power deviation.
[0010] Preferably, the method for generating the multidimensional deviation matrix in step S3 is as follows: for each time interval of each operating condition, calculate the maximum and average deviations of voltage amplitude, reactive power, reactive current and active power respectively, and form a four-dimensional deviation matrix with operating condition type, time interval, electrical quantity and deviation type as dimensions.
[0011] Preferably, the key model parameters in step S5 include: virtual impedance, voltage loop proportional-integral coefficient, current loop proportional-integral coefficient, current limiting value, and fault ride-through control coefficient; the intelligent optimization algorithm includes particle swarm optimization, genetic algorithm, or differential evolution algorithm.
[0012] Preferably, the method for identifying key model parameters through sensitivity analysis in step S5 is as follows: apply a small perturbation to each adjustable parameter, observe the rate of change of each index in the multidimensional deviation matrix, and determine the parameters whose rate of change exceeds the set sensitivity threshold as key parameters.
[0013] Preferably, in step S5, when using an intelligent optimization algorithm to automatically optimize the key model parameters, the weighted sum of each deviation index in the multidimensional deviation matrix is used as the objective function, and the physical feasible range of the model parameters is used as the constraint condition.
[0014] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, the computer program performing the steps of the method described above when run by a processor.
[0015] This invention further discloses a closed-loop verification and parameter optimization system for a network-type SVG electromagnetic transient model, comprising an interconnected memory and a processor. The memory stores a computer program, which executes the steps of the method described above when run by the processor.
[0016] Compared with the prior art, the advantages of the present invention are as follows: 1. A closed-loop verification process from hardware-in-the-loop testing to electromagnetic model simulation was achieved, significantly improving model accuracy by using the actual controller response as a benchmark; 2. A multi-dimensional, multi-interval deviation analysis method is proposed, covering steady-state and transient processes, generating a four-dimensional deviation matrix, which can comprehensively evaluate the performance of the model under different working conditions; 3. An automatic parameter optimization mechanism combining sensitivity analysis and intelligent optimization algorithms is introduced to reduce the cost of manual parameter tuning and improve modeling efficiency; 4. The model has been verified under multiple operating conditions and has good short-circuit ratio adaptability, frequency adaptability and phase jump capability, making it suitable for simulation analysis and planning operation of high-proportion new energy power systems; 5. The optimized model can be used for electromagnetic transient simulation at the station level and system level, providing an accurate basic model for the study of the stability of new energy grid connection. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating the closed-loop verification and parameter optimization method for a network-type SVG electromagnetic transient model according to an embodiment of the present invention. Detailed Implementation
[0018] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.
[0019] like Figure 1 As shown, the closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model provided in this embodiment of the invention includes the following steps: Step S1: Hardware-in-the-loop test platform setup and data acquisition; A hardware-in-the-loop test platform was built, which included an actual SVG controller and a real-time simulator (such as RTDS). Various operating conditions were set, and electrical quantity response data of the SVG grid connection point under each operating condition were collected to form a benchmark dataset.
[0020] Various operating conditions include: (1) Fault ride-through test conditions: including three-phase or two-phase symmetrical / asymmetrical faults, voltage drop or rise of 5%Un, 20%Un, 50%Un, 120%Un, 130%Un, and duration of 150ms, 625ms, 1214ms or 10s; (2) Continuous fault ride-through test conditions: the voltage drops to 5%Un for 150ms and then rises to 120%Un for 10s without delay, and the voltage drops to 5%Un for 150ms and then rises to 130%Un for 0.5s without delay; (3) Reactive power step test condition: The reactive power jumps from 0 to ±0.2Qn, ±0.5Qn, ±1.0Qn in sequence, and then steps back in the opposite direction; (4) Voltage reference value step test condition: The voltage reference value is subjected to multiple steps in the range of 525kV to 530kV; (5) Test conditions for switching between grid-connecting and grid-building modes: The switching between grid-building mode and grid-connecting mode is carried out under the conditions of 0 power, 20% capacitive reactive power and 20% inductive reactive power. (6) Short-circuit ratio adaptability test conditions: Set the short-circuit ratio SCR=1, 5, 15 and conduct fault ride-through test; (7) Frequency adaptability test conditions: The system frequency varies within the range of 46.5Hz to 51.5Hz, with a variation rate of ±2Hz / s; (8) Phase angle jump test conditions: The grid voltage phase jumps from 0° to 60°, 60° to 0°, 0° to 30°, and 30° to -30°.
[0021] The electrical quantities collected include at least: three-phase voltage at the grid connection point, three-phase current, positive sequence voltage, positive sequence active power, positive sequence reactive power, positive sequence active current, and positive sequence reactive current.
[0022] Step S2: Electromagnetic transient model establishment and simulation In electromagnetic transient simulation software (such as ADPSS), build an SVG electromagnetic transient model that is completely consistent with the primary topology and control strategy of the hardware-in-the-loop test platform, and run the simulation under the same operating conditions to collect the corresponding simulation response data. The data type is consistent with that in step S1.
[0023] A primary topology includes at least: SVG converter chain, connecting transformers, line impedance, etc. Control strategies include: virtual synchronous machine control, reactive voltage control, active frequency control, fault ride-through control, etc.
[0024] Step S3: Verification of multi-condition data and generation of multi-dimensional deviation matrix The simulation response data and the benchmark dataset are divided into multiple intervals in the time domain, including: the steady-state interval before the fault, the transient interval during the fault, the steady-state interval during the fault, the transient interval after the fault, and the steady-state interval after the fault.
[0025] Calculate the following deviation index for each interval: Maximum Error: The maximum absolute value of the difference between the simulated value and the reference value within the interval; Mean Error: The average of the absolute values of the differences between the simulated value and the reference value within the interval; Transient Mean Error: The average of the absolute values of the differences between the simulated value and the reference value within the transient interval.
[0026] The calculation should include at least the following electrical quantity deviations: voltage amplitude deviation ΔU, reactive power deviation ΔQ, reactive current deviation ΔIq, and active power deviation ΔP.
[0027] A four-dimensional deviation matrix M = {m} is formed, with operating condition type, time interval, electrical quantity, and deviation type as dimensions. i,j,k,l}, where i represents the working condition type index, j represents the time interval index, k represents the electrical quantity type index, and l represents the deviation type index.
[0028] Step S4: Deviation Judgment and Model Output Determine whether each deviation index in the multidimensional deviation matrix is less than a preset threshold. The threshold can be set according to engineering requirements, for example: voltage deviation ≤ 0.01pu, reactive current deviation ≤ 0.05pu, reactive power deviation ≤ 0.05pu.
[0029] If all deviation indicators meet the requirements, the current SVG electromagnetic transient model is output as the verified model; if any deviation indicator exceeds the threshold, the process proceeds to step S5 for parameter optimization.
[0030] Step S5: Parameter sensitivity analysis and automatic optimization First, sensitivity analysis is performed to identify key model parameters affecting the bias. The method involves applying small perturbations (e.g., ±5%) to each adjustable parameter and observing the rate of change of each index in the multidimensional bias matrix. Parameters whose rate of change exceeds a set sensitivity threshold (e.g., 10%) are identified as key parameters. Key model parameters include at least: virtual impedance and voltage loop proportional-integral coefficients. The proportional-integral coefficient of the current loop ( , ), current limit value, fault ride-through control coefficient, etc.
[0031] Then, an intelligent optimization algorithm is used to automatically optimize the key model parameters. The weighted sum of the deviation indices in the multidimensional deviation matrix is used as the objective function. :
[0032] in This is a weighting coefficient, which can be set according to the importance of the working condition and the engineering requirements; These are the deviation indices in the multidimensional deviation matrix. The physical feasible range of the model parameters is used as a constraint.
[0033] Intelligent optimization algorithms can employ particle swarm optimization (PSO), genetic algorithms, or differential evolution algorithms. Taking PSO as an example, key parameter combinations are used as particle positions, with the objective function... As a fitness function, iteratively updating particle velocity and position, it searches for a suitable... The smallest optimal parameter combination.
[0034] After updating the key parameters of the SVG electromagnetic transient model, repeat steps S2 to S4 until all deviation indicators meet the requirements.
[0035] Step S6: Model Validation and Application The optimized model was used for multi-condition simulation tests to verify its stability and response consistency under complex conditions such as different short-circuit ratios, frequency disturbances, and phase jumps. Finally, a high-precision grid-type SVG electromagnetic transient model was formed that can be used for power grid planning and operation analysis.
[0036] Compared with the prior art, the present invention has the following beneficial effects: 1. A closed-loop verification process from hardware-in-the-loop testing to electromagnetic model simulation was achieved, significantly improving model accuracy by using the actual controller response as a benchmark; 2. A multi-dimensional, multi-interval deviation analysis method is proposed, covering steady-state and transient processes, generating a four-dimensional deviation matrix, which can comprehensively evaluate the performance of the model under different working conditions; 3. An automatic parameter optimization mechanism combining sensitivity analysis and intelligent optimization algorithms is introduced to reduce the cost of manual parameter tuning and improve modeling efficiency; 4. The model has been verified under multiple operating conditions and has good short-circuit ratio adaptability, frequency adaptability and phase jump capability, making it suitable for simulation analysis and planning operation of high-proportion new energy power systems; 5. The optimized model can be used for electromagnetic transient simulation at the station level and system level, providing an accurate basic model for the study of the stability of new energy grid connection.
[0037] To better understand the above technical solutions, the following will provide a detailed explanation of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.
[0038] This embodiment uses a mesh-type SVG as an example to verify the effectiveness of the method of the present invention.
[0039] Step S1: Hardware-in-the-loop test platform setup and data acquisition A hardware-in-the-loop test platform was built, including the actual SVG controller and the RTDS real-time simulator. The primary topology was consistent with the field, including the SVG converter chain, connecting transformers, and line impedance.
[0040] Multiple operating conditions were set up for testing, including: (1) Fault ride-through test: The three-phase voltage drops to 20%Un for 625ms, and the initial reactive power output is inductive 0.2Qn; (2) Continuous fault ride-through test: Under the capacitive 0.2Qn condition, the three-phase voltage drops to 5%Un for 150ms and then rises to 120%Un for 10s without delay; (3) Reactive power step test: reactive power jumps from 0 to 20% capacitive reactive power; (4) Voltage reference value step test: The voltage reference value jumps from 525kV to 527kV; (5) Follow / Network Mode Switching Test: When the output power is 0, switch from network mode to follow mode.
[0041] Response data such as grid connection point voltage, active power, reactive power, and reactive current under various operating conditions are collected as a benchmark.
[0042] Step S2: Electromagnetic transient model establishment and simulation An SVG electromagnetic transient model, identical to the hardware-in-the-loop test platform, was built in the ADPSS electromagnetic transient simulation software. The control strategy adopted was a network-based control algorithm consistent with the actual device. The simulation was run under the same operating conditions, and the corresponding simulation response data were collected.
[0043] Step S3: Verification of multi-condition data and generation of multi-dimensional deviation matrix Taking the fault ride-through test (inductive 0.2Qn, three-phase voltage drop to 20%Un) as an example, the simulation data and HIL test data are compared in the time domain.
[0044] Divide the time intervals as follows: before the fault (0-0.5s), during the fault transient state (0.5-0.55s), during the fault steady state (0.55-1.0s), after the fault transient state (1.0-1.05s), and after the fault steady state (1.05-2.0s).
[0045] The deviation index for each interval was calculated, resulting in the following deviation table:
[0046] The deviations are calculated for other operating conditions in the same way, generating a four-dimensional deviation matrix.
[0047] Step S4: Deviation Judgment and Model Output Thresholds were set as follows: voltage deviation ≤ 0.01 pu, reactive current deviation ≤ 0.03 pu, and reactive power deviation ≤ 0.03 pu. During the fault, the maximum steady-state reactive current deviation of 0.039 pu exceeded the threshold (0.03 pu), therefore the model was deemed unsuitable and parameter optimization was required.
[0048] Step S5: Parameter sensitivity analysis and automatic optimization Sensitivity analysis was conducted: ±5% perturbations were applied to the virtual impedance, voltage loop PI parameters, and current loop PI parameters, and the rate of change of reactive current deviation was observed. It was found that the virtual impedance had the greatest impact on the steady-state reactive current deviation during faults (rate of change >20%), and was therefore identified as a key parameter.
[0049] The particle swarm optimization algorithm was used to automatically optimize the virtual impedance. The parameter range was set to 0.05-0.2 pu. The objective function was the weighted average deviation of the steady-state reactive current during the fault. After 50 iterations, the optimal virtual impedance of 0.11 pu was found.
[0050] Update the model parameters and rerun the simulation. After optimization, the maximum deviation of steady-state reactive current during the fault decreased from 0.039 pu to 0.012 pu, and the average deviation decreased from 0.023 pu to 0.008 pu, meeting the threshold requirements.
[0051] Step S6: Model Validation and Application The optimized model was validated under conditions including continuous fault ride-through, reactive power step, and voltage step, with deviations for each condition controlled within 0.02 pu. The validation results show that the optimized model accurately reflects the dynamic response characteristics of the actual SVG and can be used for subsequent station-level and system-level electromagnetic transient simulation analysis.
[0052] This invention specifically constructs a hardware-in-the-loop (HIL) test platform to collect response data of a network-type SVG under multiple operating conditions as a benchmark. An SVG electromagnetic transient model consistent with the HIL test is built in electromagnetic transient simulation software, and simulation data is obtained by running the same operating conditions. The simulation data and HIL test data are divided into multiple intervals in the time domain, and the maximum deviation, average deviation, and transient average deviation within each interval are calculated to generate a multi-dimensional deviation matrix. If the deviation exceeds a set threshold, key parameters are identified through sensitivity analysis, and intelligent optimization algorithms are used to automatically optimize the model parameters. The model is updated and re-verified until the accuracy requirements are met. Finally, a high-precision network-type SVG electromagnetic transient model is output. This invention achieves closed-loop verification and automatic optimization throughout the entire process from HIL testing to electromagnetic model simulation, significantly improving model accuracy and modeling efficiency, and is suitable for simulation analysis and planning operation of new energy power systems.
[0053] The present invention also discloses a computer-readable storage medium having a computer program stored thereon, the computer program performing the steps of the method described above when run by a processor.
[0054] This invention further discloses a closed-loop verification and parameter optimization system for a network-type SVG electromagnetic transient model, comprising an interconnected memory and a processor. The memory stores a computer program, which executes the steps of the method described above when run by the processor.
[0055] The medium and system of the present invention, corresponding to the methods described above, also have the advantages described above.
[0056] The present invention can implement all or part of the processes in the methods of the above embodiments, or it can be implemented by hardware related to computer program instructions. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, it can implement the steps of the above method embodiments. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer-readable storage medium includes: any entity or device capable of carrying computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. The memory is used to store computer programs and / or modules. The processor implements various functions by running or executing the computer programs and / or modules stored in the memory, and by calling data stored in the memory. The memory may include high-speed random access memory, as well as non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.
[0057] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.
Claims
1. A method for closed-loop verification and parameter optimization of a network-type SVG electromagnetic transient model, characterized in that, Including the following steps: Step S1: Build a hardware-in-the-loop test platform that includes an actual SVG controller and a real-time simulator, set up various operating conditions, collect electrical quantity response data of the SVG grid connection point under each operating condition, and form a benchmark dataset; Step S2: Build an SVG electromagnetic transient model in the electromagnetic transient simulation software that is consistent with the primary topology and control strategy of the hardware-in-the-loop test platform, and run the simulation under the same operating conditions to collect the corresponding simulation response data. Step S3: Divide the simulation response data and the benchmark dataset into multiple intervals in the time domain, calculate the deviation index in each interval, and generate a multidimensional deviation matrix; Step S4: Determine whether each deviation index in the multidimensional deviation matrix is less than a preset threshold. If yes, output the current SVG electromagnetic transient model as the verified model; otherwise, proceed to step S5. Step S5: Identify the key model parameters affecting the deviation through sensitivity analysis, use an intelligent optimization algorithm to automatically optimize the key model parameters, update the SVG electromagnetic transient model, and repeat steps S2 to S4 until all deviation indicators meet the requirements.
2. The closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model according to claim 1, characterized in that, The various operating conditions in step S1 include: fault ride-through test condition, continuous fault ride-through test condition, reactive power step test condition, voltage reference value step test condition, grid mode switching test condition, short-circuit ratio adaptability test condition, frequency adaptability test condition, and phase angle jump test condition.
3. The closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model according to claim 2, characterized in that, The fault ride-through test conditions include: three-phase or two-phase symmetrical / asymmetrical faults, with voltage drops or rises of 5%Un, 20%Un, 50%Un, 120%Un, and 130%Un, lasting for 150ms, 625ms, 1214ms, or 10s; the continuous fault ride-through test conditions include: voltage drops to 5%Un for 150ms followed by a rise to 120%Un for 10s without delay, and voltage drops to 5%Un for 150ms followed by a rise to 130%Un for 0.5s without delay; where Un is the rated voltage.
4. The closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model according to claim 1, 2, or 3, characterized in that, The multi-interval division in step S3 includes: the pre-fault steady-state interval, the fault transient interval, the fault steady-state interval, the post-fault transient interval, and the post-fault steady-state interval; the deviation index includes: the maximum deviation, the average deviation, and the transient average deviation; the multi-dimensional deviation matrix includes at least the voltage amplitude deviation, the reactive power deviation, the reactive current deviation, and the active power deviation.
5. The closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model according to claim 4, characterized in that, The method for generating the multidimensional deviation matrix in step S3 is as follows: for each time interval of each operating condition, calculate the maximum and average deviations of voltage amplitude, reactive power, reactive current and active power respectively, and form a four-dimensional deviation matrix with operating condition type, time interval, electrical quantity and deviation type as dimensions.
6. The closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model according to claim 1, 2, or 3, characterized in that, The key model parameters in step S5 include: virtual impedance, voltage loop proportional-integral coefficient, current loop proportional-integral coefficient, current limiting value, and fault ride-through control coefficient; the intelligent optimization algorithm includes particle swarm optimization, genetic algorithm, or differential evolution algorithm.
7. The closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model according to claim 6, characterized in that, The method for identifying key model parameters through sensitivity analysis in step S5 is as follows: apply a small perturbation to each adjustable parameter, observe the rate of change of each index in the multidimensional deviation matrix, and determine the parameters whose rate of change exceeds the set sensitivity threshold as key parameters.
8. The closed-loop verification and parameter optimization method for the network-type SVG electromagnetic transient model according to claim 7, characterized in that, In step S5, when using an intelligent optimization algorithm to automatically optimize key model parameters, the weighted sum of each deviation index in the multidimensional deviation matrix is used as the objective function, and the physical feasible range of the model parameters is used as the constraint condition.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1-8.
10. A closed-loop verification and parameter optimization system for a network-type SVG electromagnetic transient model, comprising an interconnected memory and a processor, wherein the memory stores a computer program, characterized in that, The computer program, when run by a processor, performs the steps of the method as described in any one of claims 1-8.
