An electromagnetic transient simulation method, system, storage medium and device

Abstract

The application discloses an electromagnetic transient simulation method, comprising the following steps: collecting voltage and current response signals of nodes under wide-frequency disturbance, and determining cross-coupling impedance sensitivity matrix; establishing a full-order state matrix according to an original full-order model, decomposing the matrix to determine initial participation factors of each node, and dividing the nodes into full-order detailed nodes or aggregated equivalent nodes according to the participation factors, and establishing a hybrid equivalent network; collecting a common connection point voltage signal, determining a comprehensive voltage disturbance index, and when the index exceeds a threshold, determining real-time pseudo participation factors according to instantaneous apparent power fluctuation, initial participation factors and sensitivity matrix; dynamically switching node types according to a comparison result of the real-time pseudo participation factors and a model switching threshold; finally, simulating the aggregated equivalent nodes by using a reduced-order aggregation model, simulating the full-order detailed nodes by using a full-order detailed model, and realizing electromagnetic transient simulation through dynamic switching.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to an electromagnetic transient simulation method, system, storage medium, and device. Background Technology

[0002] For electromagnetic transient simulation of large-scale wind, solar, and energy storage clusters connected to the grid, two main modeling techniques are currently employed. The first is full-order detailed modeling, which requires establishing detailed models of all grid-connected inverters within the system, including complete control loops and switching devices, and then uniformly incorporating them into the solution of the entire network's electromagnetic transient equations. However, with a massive number of connected nodes, the dimension of the entire network admittance matrix expands exponentially, resulting in extremely long simulation times and failing to meet the needs for rapid analysis. The second is static equivalent modeling, which solidifies inverters with similar physical locations and parameters into a single equivalent model before simulation. However, when the system experiences severe faults or broadband oscillations, the operating conditions of the forcibly aggregated inverters deviate significantly from each other, exhibiting strong transient differences. The solidified equivalent model cannot reflect the details of transient interactions, leading to distorted simulation results. Summary of the Invention

[0003] Therefore, it is necessary to propose an electromagnetic transient simulation method to address the above problems.

[0004] An electromagnetic transient simulation method, the method comprising the following steps: Collect the voltage and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances; The cross-coupling impedance sensitivity matrix is ​​determined based on the voltage response signal and the current response signal. Based on the original full-order electromagnetic transient model of the system to be simulated, a full-order state matrix is ​​established. The full-order state matrix is ​​decomposed to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and the preset initial partitioning threshold, each grid-connected inverter node is divided into full-order detailed nodes or aggregated equivalent nodes to establish a hybrid equivalent network. The voltage signal of the common connection point in the hybrid equivalent network is collected, and the comprehensive voltage disturbance index is determined based on the voltage signal. When the comprehensive voltage disturbance index exceeds the preset trigger threshold, the real-time pseudo-participation factor is determined based on the instantaneous apparent power fluctuation of the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix. Based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold, the type of node is dynamically switched: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and the delay time window continues, the node is switched to an aggregated equivalence node. The reduced-order aggregation model is used to simulate the aggregated equivalent nodes, and the full-order detailed model is used to simulate the full-order detailed nodes. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0005] In the above scheme, determining the cross-coupling impedance sensitivity matrix based on the voltage response signal and the current response signal specifically includes: The voltage response signal and the current response signal are converted to the frequency domain and divided to obtain a wideband equivalent impedance array. Calculate the absolute value of the partial derivative of the wideband equivalent impedance array with respect to the frequency domain voltage response fluctuation, and use it as an element of the cross-coupled impedance sensitivity matrix; The elements of the cross-coupling impedance sensitivity matrix at frequency ω are determined according to the following formula: in, Let be the elements of the cross-coupling impedance sensitivity matrix of node k to node i at frequency ω. ΔZ i (ω) represents the change in equivalent impedance at node i at frequency ω, ΔU kj (ω) represents the voltage fluctuation at node i caused by the power disturbance injected at node k at frequency ω.

[0006] In the above scheme, the step of decomposing the full-order state matrix to determine the initial participation factor of each grid-connected inverter node, and dividing each grid-connected inverter node into a full-order detailed node or an aggregated equivalent node based on the comparison relationship between the initial participation factor and a preset initial partitioning threshold, and establishing a hybrid equivalent network, specifically includes: The full-order state matrix is ​​decomposed into eigenvalues, and the weakly damped mode is taken as the initial dominant oscillation mode. Based on the left and right eigenvectors of the full-order state matrix, calculate the initial participation factor of each node in the i-th mode:

[0007] Among them, w ik v is the k-th element in the left eigenvector corresponding to the i-th mode. ki Let k be the k-th element in the right eigenvector corresponding to the i-th mode. Let the initial participation factor of the k-th node in the i-th mode be denoted as ; When the maximum value of the initial participation factor is less than the preset initial division threshold, and the maximum value of the cross-coupling impedance sensitivity corresponding to the node is less than the preset impedance coupling threshold, the node is divided into an aggregated equal value node; otherwise, it is divided into a full-order detailed node.

[0008] In the above scheme, the step of collecting voltage signals at common connection points in the hybrid equivalence network and determining the comprehensive voltage disturbance index based on the voltage signals specifically includes: Collect voltage signals at common connection points in the hybrid equivalent network, and calculate the effective value deviation and time derivative rate of change of the voltage signals; The effective value deviation and the time derivative rate of change of the voltage signal are linearly weighted and summed to obtain the comprehensive voltage disturbance index: ; in, The preset linear weighting coefficients are used, and ΔU is the effective value deviation of the voltage signal. Let be the time derivative rate of change of the voltage signal.

[0009] In the above scheme, determining the real-time pseudo-participation factor based on the instantaneous apparent power fluctuation at the common junction point, the initial participation factor, and the cross-coupling impedance sensitivity matrix specifically includes: Calculate the real-time pseudo-participation factor for each node using the following formula:

[0010] in, Let i be the real-time pseudo-participation factor for node i. Let be the initial participation factor for node i. The total number of nodes. These are the normalized weighting coefficients. The dimension conversion coefficient is... The dominant oscillation frequency The elements of the cross-coupling impedance sensitivity matrix between node i and node k are shown below. Let be the instantaneous apparent power fluctuation at node k.

[0011] In the above scheme, the simulation of the aggregated equivalent nodes using a reduced-order aggregation model specifically includes: The discrete state-space parameter matrix is ​​determined based on the off-diagonal elements in the cross-coupling impedance sensitivity matrix. Construct the node admittance matrix of the hybrid isovalued network; By solving the node admittance matrix at the current moment, the instantaneous voltage deviation of each node at the current moment can be obtained. Collect the instantaneous voltage deviation corresponding to the aggregated equivalent nodes; Differential iterative calculations are performed based on the discrete state space parameter matrix and the instantaneous voltage deviation to generate a compensation current command; The compensation current command is converted into a controlled current source signal and injected into the node admittance matrix for solving in the next time step. Repeat the above solution steps until the preset simulation end time is reached to complete the electromagnetic transient simulation of aggregated equivalent nodes.

[0012] In the above scheme, the step of generating a compensation current command by performing differential iterative calculations based on the discrete state-space parameter matrix and the instantaneous voltage deviation specifically includes: The discrete system matrix is ​​determined based on the discrete state-space parameter matrix. Discrete input matrix Discrete output matrix Discrete direct feedforward matrix ; At the current discrete time Extract the previous time step instantaneous voltage deviation and the internal discrete state variables of the previous time step ; According to the discrete system matrix Discrete input matrix Update the internal discrete state variables at the current time step. : ; According to the discrete output matrix Discrete direct feedforward matrix Combined with the instantaneous voltage deviation at the current moment Calculate and generate compensation current command :

[0013] in, To compensate for current commands, Let this be the vector of internal state variables at the current moment. This represents the instantaneous voltage deviation at the current moment.

[0014] This application proposes an electromagnetic transient simulation system, which includes: a signal acquisition unit, a signal processing unit, a dynamic switching unit, and an electromagnetic transient simulation unit; The signal acquisition unit is used to acquire the voltage response signals and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances, as well as the voltage signal of the common connection point in the hybrid equivalent network. The signal processing unit is used to determine the cross-coupling impedance sensitivity matrix based on the voltage response signal and the current response signal; to establish the full-order state matrix based on the original full-order electromagnetic transient model of the system to be simulated; to decompose the full-order state matrix to determine the initial participation factor of each grid-connected inverter node; and to divide each grid-connected inverter node into full-order detailed nodes or aggregated equivalent nodes based on the comparison relationship between the initial participation factor and the preset initial partitioning threshold, thus establishing a hybrid equivalent network; to determine the comprehensive voltage disturbance index based on the voltage signal of the point of common coupling; and to determine the real-time pseudo-participation factor when the comprehensive voltage disturbance index exceeds the preset triggering threshold, based on the instantaneous apparent power fluctuation of the point of common coupling, the initial participation factor, and the cross-coupling impedance sensitivity matrix. The dynamic switching unit is used to dynamically switch the type of a node based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and is continuously delayed for a time window, the node is switched to an aggregated equivalence node. The electromagnetic transient simulation unit is used to simulate aggregated equivalent nodes using a reduced-order aggregated model and to simulate full-order detailed nodes using a full-order detailed model. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0015] This application proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: Collect the voltage and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances; The cross-coupling impedance sensitivity matrix is ​​determined based on the voltage response signal and the current response signal. Based on the original full-order electromagnetic transient model of the system to be simulated, a full-order state matrix is ​​established. The full-order state matrix is ​​decomposed to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and the preset initial partitioning threshold, each grid-connected inverter node is divided into full-order detailed nodes or aggregated equivalent nodes to establish a hybrid equivalent network. The voltage signal of the common connection point in the hybrid equivalent network is collected, and the comprehensive voltage disturbance index is determined based on the voltage signal. When the comprehensive voltage disturbance index exceeds the preset trigger threshold, the real-time pseudo-participation factor is determined based on the instantaneous apparent power fluctuation of the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix. Based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold, the type of node is dynamically switched: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and the delay time window continues, the node is switched to an aggregated equivalence node. The reduced-order aggregation model is used to simulate the aggregated equivalent nodes, and the full-order detailed model is used to simulate the full-order detailed nodes. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0016] This application discloses a computer device, including a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the following steps: Collect the voltage and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances; The cross-coupling impedance sensitivity matrix is ​​determined based on the voltage response signal and the current response signal. Based on the original full-order electromagnetic transient model of the system to be simulated, a full-order state matrix is ​​established. The full-order state matrix is ​​decomposed to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and the preset initial partitioning threshold, each grid-connected inverter node is divided into full-order detailed nodes or aggregated equivalent nodes to establish a hybrid equivalent network. The voltage signal of the common connection point in the hybrid equivalent network is collected, and the comprehensive voltage disturbance index is determined based on the voltage signal. When the comprehensive voltage disturbance index exceeds the preset trigger threshold, the real-time pseudo-participation factor is determined based on the instantaneous apparent power fluctuation of the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix. Based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold, the type of node is dynamically switched: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and the delay time window continues, the node is switched to an aggregated equivalence node. The reduced-order aggregation model is used to simulate the aggregated equivalent nodes, and the full-order detailed model is used to simulate the full-order detailed nodes. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0017] The embodiments of this invention have the following beneficial effects: First, by collecting response signals under broadband disturbances, the cross-coupling impedance sensitivity matrix is ​​determined, and a full-order state matrix is ​​established using the original full-order model, decomposing the initial participation factor of each node. Based on the comparison of this factor with a preset threshold, nodes are divided into full-order detailed nodes or aggregated equivalent nodes, thereby achieving optimal initial allocation for high-precision modeling of critical nodes and reduced-order modeling of non-critical nodes before simulation starts. Second, during the simulation, the comprehensive voltage disturbance index is calculated by collecting voltage signals at the common connection point, and when the index exceeds the threshold, the real-time pseudo-participation factor is determined using the instantaneous apparent power fluctuation, the initial participation factor, and the sensitivity matrix. Through dynamic comparison of the real-time pseudo-participation factor with the model switching threshold, adaptive switching of node types is achieved. Finally, a reduced-order aggregated model is used for simulation of aggregated equivalent nodes, and a full-order detailed model is used for simulation of full-order detailed nodes. By dynamically switching between the two types of models and solving them in a unified manner, reduced-order can be performed at steady-state or non-dominant nodes, and a detailed model can be used to restore physical details when severe transient disturbances occur locally, solving the mutual constraint between simulation speed and the degree of restoration of local transient details. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] in: Figure 1 This is a schematic diagram of the electromagnetic transient simulation method in one embodiment. Detailed Implementation

[0020] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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.

[0021] In the following description, numerous specific details are set forth in order to provide a more thorough understanding of the invention; however, it will be apparent to those skilled in the art that the invention may be practiced without one or more of these details; in other instances, certain technical features well-known in the art have not been described in order to avoid confusion with the invention. It should be understood that the invention can be practiced in different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided to make the disclosure thorough and complete and to fully convey the scope of the invention to those skilled in the art.

[0022] The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. When used herein, the singular forms “a,” “an,” and “the” are also intended to include the plural forms, unless the context clearly indicates otherwise. The terms “comprising” and / or “including,” when used in this specification, identify the presence of said features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups. When used herein, the term “and / or” includes any and all combinations of the associated listed items.

[0023] To fully understand the present invention, a detailed structure will be presented in the following description in order to illustrate the technical solution proposed by the present invention; optional embodiments of the present invention are described in detail below, however, in addition to these detailed descriptions, the present invention may have other embodiments.

[0024] like Figure 1 As shown, in one embodiment, an electromagnetic transient simulation method is provided, which includes steps S101 to S107, detailed below: S101. Collect the voltage response signal and current response signal of each grid-connected inverter node in the system to be simulated under broadband disturbance; This step provides a basic data source for the subsequent extraction of the system's wideband dynamic characteristics. Unlike analysis at only the power frequency or a single frequency, the response signal under wideband disturbances can comprehensively characterize the external characteristics of the inverter in various oscillation frequency bands, especially the subsynchronous / supersynchronous oscillation frequency bands, ensuring that the potential dynamic interaction risk characteristics of the system to be simulated are captured.

[0025] S102. Determine the cross-coupling impedance sensitivity matrix based on the voltage response signal and the current response signal; In large-scale wind, solar and energy storage clusters, a disturbance at one point may trigger broadband oscillations in other nodes. This sensitivity matrix accurately identifies which node has the greatest impact on which node, providing a mathematical basis for subsequent identification of key coupling paths and implementation of precise order reduction, and avoiding the loss of interactive features caused by blind aggregation.

[0026] In some embodiments, the cross-coupling impedance sensitivity matrix is ​​determined based on the voltage response signal and the current response signal, specifically including: The voltage response signal and the current response signal are converted to the frequency domain and divided to obtain a wideband equivalent impedance array. Calculate the absolute value of the partial derivative of the broadband equivalent impedance array with respect to the frequency domain voltage response fluctuation, and use it as an element of the cross-coupled impedance sensitivity matrix; The elements of the cross-coupling impedance sensitivity matrix at frequency ω are determined using the following formula: in, Let ΔZ be the cross-coupling impedance sensitivity matrix element of node k to node i at frequency ω. i (ω) represents the change in equivalent impedance at node i at frequency ω, ΔU kj (ω) represents the voltage fluctuation at node i caused by the power disturbance injected at node k at frequency ω.

[0027] Preferably, a wideband swept-frequency voltage disturbance signal is generated in an offline computing environment using a multi-frequency sinusoidal superposition or pseudo-random binary sequence (PRBS) generator. This signal is injected at the corresponding common connection point (PCC) of each grid-connected inverter node in the network to be equivalent, and the instantaneous voltage response signal and output current response signal are recorded. The voltage response signal and the output current response signal are converted to the frequency domain and divided to obtain the wideband equivalent impedance array. The absolute value of the partial derivative of each wideband equivalent impedance array with respect to the frequency domain voltage response fluctuation caused by the power disturbance is calculated using the numerical difference method and used as an element of the cross-coupling impedance sensitivity matrix.

[0028] Assuming there are nodes 1 and 2, at a frequency of 100Hz, a small power disturbance injected by node 2 causes a voltage fluctuation of ΔU = 0.01pu at node 1, and simultaneously, an equivalent impedance change of ΔZ = 0.05pu is observed at node 1. Then, at this frequency, the matrix element C... 12 (100HZ)=0.05 / 0.01=5.0. This is used to iterate through all node combinations to generate a complete cross-coupled impedance sensitivity matrix.

[0029] S103. Based on the original full-order electromagnetic transient model of the system to be simulated, establish the full-order state matrix; S104. Decompose the full-order state matrix to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and the preset initial division threshold, divide each grid-connected inverter node into a full-order detailed node or an aggregated equivalent node to establish a hybrid equivalent network. This application quantifies the contribution of each node to the dominant oscillation mode of the system from the perspective of the global inherent oscillation characteristics, i.e., the participation factor. Nodes with high participation factors are the core sources of system oscillations, while nodes with low participation factors are non-dominant nodes. This step strongly correlates network order reduction with the most dangerous dynamic characteristics of the system, ensuring that the order reduction does not lose its physical essence. Furthermore, by retaining the full-order detailed nodes with high participation factors, the control details of the core oscillation sources are not lost; while by aggregating and equivaluing the non-dominant nodes with low participation factors, the dimension of the initial admittance matrix of the entire network is greatly reduced, achieving a preliminary balance between maintaining accuracy and improving speed before the simulation starts.

[0030] In some embodiments, the full-order state matrix is ​​decomposed to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and a preset initial partitioning threshold, each grid-connected inverter node is divided into full-order detailed nodes or aggregated equivalent nodes to establish a hybrid equivalent network. Specifically, this includes: Eigenvalue decomposition is performed on the full-order state matrix, and the weakly damped mode is taken as the initial dominant oscillation mode; Based on the left and right eigenvectors of the full-order state matrix, calculate the initial participation factor of each node in the i-th mode:

[0031] Among them, w ik v is the k-th element in the left eigenvector corresponding to the i-th mode. ki Let k be the k-th element in the right eigenvector corresponding to the i-th mode. Let the initial participation factor of the k-th node in the i-th mode be denoted as ; When the maximum value of the initial participation factor is less than the preset initial partitioning threshold, and the maximum value of the cross-coupling impedance sensitivity of the node is less than the preset impedance coupling threshold, the node is partitioned into an aggregated equal value node; otherwise, it is partitioned into a full-order detailed node.

[0032] Preferably, eigenvalue decomposition is performed on the full-order state matrix, and weakly damped modes are selected as the initial dominant oscillation modes based on the cross-selection of damping ratio and oscillation frequency. The oscillation frequency corresponding to this initial dominant oscillation mode will serve as the frequency positioning benchmark for subsequently extracting specific elements of the cross-coupling impedance sensitivity matrix, thereby strongly correlating the network order reduction with the most dangerous dynamic interaction frequency of the system. Then, based on the left and right eigenvectors of the full-order state matrix, the initial participation factor of each node in the i-th order mode is calculated.

[0033] Furthermore, the nodes of the reduced-order aggregation model are constructed using the frequency domain Thevenin analytical expression to form an equivalent controlled voltage source model. The equivalent internal potential is obtained by calculating the steady-state voltage and steady-state injected current phasors, and the internal equivalent impedance is obtained by discretizing the broadband equivalent impedance array into a time-domain equivalent state-space branch.

[0034] S105. Collect the voltage signal of the common connection point in the hybrid equivalent network, and determine the comprehensive voltage disturbance index based on the voltage signal; when the comprehensive voltage disturbance index exceeds the preset trigger threshold, determine the real-time pseudo-participation factor based on the instantaneous apparent power fluctuation of the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix. This step constructs a real-time sensing probe of the power grid's operating status. The comprehensive voltage disturbance index integrates voltage offset, i.e., steady-state deviation and rate of change, i.e., transient change. It can sensitively and without hysteresis capture the moment when the system transitions from steady state to transient fault or broadband oscillation, serving as a reliable preliminary criterion for triggering dynamic model reconstruction.

[0035] In some embodiments, voltage signals at common connection points in a hybrid equivalent network are acquired, and a comprehensive voltage disturbance index is determined based on the voltage signals, specifically including: Collect voltage signals at common connection points in a hybrid equivalent network, and calculate the effective value deviation and time derivative rate of change of the voltage signals. The comprehensive voltage disturbance index is obtained by linearly weighting and summing the effective value deviation and the time derivative rate of change of the voltage signal. ; in, The preset linear weighting coefficients are used, and ΔU is the effective value deviation of the voltage signal. Let be the time derivative rate of change of the voltage signal.

[0036] In some embodiments, the real-time pseudo-participation factor is determined based on the instantaneous apparent power fluctuation at the point of common coupling, the initial participation factor, and the cross-coupling impedance sensitivity matrix, specifically including: Calculate the real-time pseudo-participation factor for each node using the following formula:

[0037] in, Let i be the real-time pseudo-participation factor for node i. Let be the initial participation factor for node i. The total number of nodes. These are the normalized weighting coefficients. The dimension conversion coefficient is... The dominant oscillation frequency The elements of the cross-coupling impedance sensitivity matrix between node i and node k are shown below. Let be the instantaneous apparent power fluctuation at node k.

[0038] S106. Based on the comparison results between the real-time pseudo-participation factor and the preset model switching threshold, dynamically switch the node type: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, switch the node to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and continues to be delayed for a time window, switch the node to an aggregated equivalence node. In reality, when a large disturbance occurs in a power system, the roles of each node shift, and previously non-dominant nodes may be activated to become core nodes. The real-time pseudo-participation factor combines real-time power disturbance with prior coupling sensitivity, reflecting both the current excitation level and the inherent network structure coupling, enabling accurate switching of the node type corresponding to each node. Specifically, at the moment of a large disturbance, the excited node is instantly unpacked into a full-order model, ensuring high-fidelity reproduction of transient interaction details; after the disturbance decays, a continuous delay time window is used to confirm that the power system has returned to stability before switching the node back to a reduced-order model. This effectively avoids frequent model switching due to signal fluctuations at the fault critical point, ensuring the numerical stability of the time-domain iteration.

[0039] S107. Simulation is performed using a reduced-order aggregation model for aggregated equivalent nodes and a full-order detailed model for full-order detailed nodes. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0040] When the power system is in steady state, the entire network is mainly simulated using a reduced-order model, resulting in extremely low computational load and overcoming the bottleneck of time consumption for full-order models. When the power system is in transient state, the core region is instantly restored to a full-order physical model, overcoming the persistent problem of static equivalent models becoming distorted under large disturbances. The seamless dynamic switching between the two types of models in the time domain ultimately achieves fast and accurate electromagnetic transient simulation in large-scale wind, solar, and energy storage multi-source integration scenarios.

[0041] In some embodiments, a reduced-order aggregation model is used to simulate the aggregated equivalent nodes, specifically including: The discrete state-space parameter matrix is ​​determined based on the off-diagonal elements in the cross-coupling impedance sensitivity matrix. Construct the node admittance matrix of the hybrid iso-network; By solving the node admittance matrix at the current moment, the instantaneous voltage deviation of each node at the current moment can be obtained. Collect the instantaneous voltage deviation corresponding to the aggregated equivalent nodes; Differential iterative calculations are performed based on the discrete state-space parameter matrix and the instantaneous voltage deviation to generate compensation current commands; The compensation current command is converted into a controlled current source signal and injected into the node admittance matrix for the next time step. Repeat the above solution steps until the preset simulation end time is reached to complete the electromagnetic transient simulation of aggregated equivalent nodes.

[0042] In some embodiments, determining the discrete state-space parameter matrix based on the off-diagonal elements of the cross-coupling impedance sensitivity matrix specifically includes: (1) Threshold filtering: Extract the off-diagonal elements of the cross-coupling impedance sensitivity matrix, set the intensity filtering threshold (e.g., 0.1), if the extracted off-diagonal element = 0.03 < 0.1, then force it to be zero; if = 0.5 ≥ 0.1, then retain it. In this way, weak coupling noise is filtered out and a sparsified residual coupling impedance spectrum is obtained; (2) Vector fitting process: Input the retained residual coupling impedance spectrum data into the vector fitting algorithm. First, a set of initial poles on the complex plane is preset. Using the least squares method based on partial fraction expansion, the linear equation system is solved iteratively to obtain the poles and residues of the rational fraction that approximates the original frequency response. By comparing the root mean square error of the fitting, if the error does not meet the standard, the fitting order is dynamically increased and iterated again until the error converges. (3) Forced stability test: Extract all poles of the fitted rational fraction and check the real part of each pole one by one. If there is a pole located in the right half-plane (i.e., real part > 0), perform a pole flipping operation to change it to a negative value and force it to be mapped to the left half-plane to ensure the absolute stability of subsequent time-domain simulations; (4) Finally, the rational fraction that has passed the stability test is transformed into a discrete state-space parameter matrix by combining the bilinear transformation.

[0043] In some embodiments, differential iterative calculations are performed based on the discrete state-space parameter matrix and the instantaneous voltage deviation to generate a compensation current command, specifically including: Determine the discrete system matrix based on the discrete state-space parameter matrix. Discrete input matrix Discrete output matrix Discrete direct feedforward matrix ; At the current discrete time Extract the previous time step instantaneous voltage deviation and the internal discrete state variables of the previous time step ; Based on the discrete system matrix Discrete input matrix Update the internal discrete state variables at the current time step. : ; Based on the discrete output matrix Discrete direct feedforward matrix Combined with the instantaneous voltage deviation at the current moment Calculate and generate compensation current command :

[0044] in, To compensate for current commands, Let this be the vector of internal state variables at the current moment. This represents the instantaneous voltage deviation at the current moment.

[0045] Furthermore, in order to solve the numerical ill-conditioned problem of solving large-scale multi-source networks, the compensation current command is converted into a controlled current source signal constructed by parallel connection of controlled current sources with minimal conductance Norton equivalent form, and injected into the network node admittance matrix to be equivalently solved.

[0046] During the injection solution process, when the principal element in the nodal admittance matrix decomposition is detected to be minimizing and tending to be singular, the conductance disturbance compensation logic is automatically triggered. The diagonal loading protection, which is no less than the lower limit of numerical truncation, is applied to the main diagonal elements to prevent matrix decomposition failure and calculation divergence caused by weak damping coupling in the scenario of massive access of multi-source power electronic equipment. The equivalent model of multi-source AC access is output for fast electromagnetic transient simulation.

[0047] Preferably, when the principal elements in the decomposition of the node admittance matrix are detected to be singular, the conductance disturbance compensation logic is automatically triggered to apply diagonal loading protection of not less than the lower limit of numerical truncation to the main diagonal elements of the node admittance matrix.

[0048] In summary, this application achieves adaptive dynamic switching of the dimensionality of the grid-connected inverter model in time-domain simulation through cross-coupling impedance sensitivity assessment and real-time pseudo-participation factor calculation techniques. It performs order reduction aggregation and discrete difference operations on non-dominant disturbance regions, significantly reducing the size of node equations participating in high-frequency iterations and increasing the transient simulation speed of large-scale wind-solar-storage clusters by several times. It also overcomes the persistent problem of traditional static equivalent models failing under large disturbances. At the moment of a grid fault, the affected core nodes are automatically unpacked and restored to full-order detailed physical models, capable of faithfully reproducing the transient interactions and electromagnetic response details of the internal controllers.

[0049] This application proposes an electromagnetic transient simulation system, which includes: a signal acquisition unit, a signal processing unit, a dynamic switching unit, and an electromagnetic transient simulation unit; The signal acquisition unit is used to acquire the voltage response signals and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances, as well as the voltage signal of the common connection point in the hybrid equivalent network. The signal processing unit is used to determine the cross-coupling impedance sensitivity matrix based on the voltage response signal and the current response signal; to establish the full-order state matrix based on the original full-order electromagnetic transient model of the system to be simulated; to decompose the full-order state matrix to determine the initial participation factor of each grid-connected inverter node; and to divide each grid-connected inverter node into full-order detailed nodes or aggregated equivalent nodes based on the comparison relationship between the initial participation factor and the preset initial partitioning threshold, thus establishing a hybrid equivalent network; to determine the comprehensive voltage disturbance index based on the voltage signal of the point of common coupling; and to determine the real-time pseudo-participation factor when the comprehensive voltage disturbance index exceeds the preset triggering threshold, based on the instantaneous apparent power fluctuation of the point of common coupling, the initial participation factor, and the cross-coupling impedance sensitivity matrix. The dynamic switching unit is used to dynamically switch the type of a node based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and is continuously delayed for a time window, the node is switched to an aggregated equivalence node. The electromagnetic transient simulation unit is used to simulate aggregated equivalent nodes using a reduced-order aggregated model and to simulate full-order detailed nodes using a full-order detailed model. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0050] This application proposes a readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: Collect the voltage and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances; The cross-coupling impedance sensitivity matrix is ​​determined based on the voltage response signal and the current response signal. Based on the original full-order electromagnetic transient model of the system to be simulated, a full-order state matrix is ​​established. The full-order state matrix is ​​decomposed to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and the preset initial partitioning threshold, each grid-connected inverter node is divided into full-order detailed nodes or aggregated equivalent nodes to establish a hybrid equivalent network. The voltage signal of the common connection point in the hybrid equivalent network is collected, and the comprehensive voltage disturbance index is determined based on the voltage signal. When the comprehensive voltage disturbance index exceeds the preset trigger threshold, the real-time pseudo-participation factor is determined based on the instantaneous apparent power fluctuation of the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix. Based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold, the type of node is dynamically switched: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and the delay time window continues, the node is switched to an aggregated equivalence node. The reduced-order aggregation model is used to simulate the aggregated equivalent nodes, and the full-order detailed model is used to simulate the full-order detailed nodes. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0051] This application discloses a computer device, including a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the following steps: Collect the voltage and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances; The cross-coupling impedance sensitivity matrix is ​​determined based on the voltage response signal and the current response signal. Based on the original full-order electromagnetic transient model of the system to be simulated, a full-order state matrix is ​​established. The full-order state matrix is ​​decomposed to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and the preset initial partitioning threshold, each grid-connected inverter node is divided into full-order detailed nodes or aggregated equivalent nodes to establish a hybrid equivalent network. The voltage signal of the common connection point in the hybrid equivalent network is collected, and the comprehensive voltage disturbance index is determined based on the voltage signal. When the comprehensive voltage disturbance index exceeds the preset trigger threshold, the real-time pseudo-participation factor is determined based on the instantaneous apparent power fluctuation of the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix. Based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold, the type of node is dynamically switched: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and the delay time window continues, the node is switched to an aggregated equivalence node. The reduced-order aggregation model is used to simulate the aggregated equivalent nodes, and the full-order detailed model is used to simulate the full-order detailed nodes. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

[0052] Those skilled in the art will understand that implementing all or part of the processes in the above embodiments can be accomplished by instructing related hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0053] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0054] The embodiments described above are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application's patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. The embodiments disclosed above are merely preferred embodiments of the present invention and should not be construed as limiting the scope of the present invention. Therefore, equivalent variations made according to the claims of this invention are still within the scope of this invention.

Claims

1. An electromagnetic transient simulation method, characterized in that, The method includes: Collect the voltage and current response signals of each grid-connected inverter node in the system to be simulated under broadband disturbances; The cross-coupling impedance sensitivity matrix is ​​determined based on the voltage response signal and the current response signal. Based on the original full-order electromagnetic transient model of the system to be simulated, a full-order state matrix is ​​established. The full-order state matrix is ​​decomposed to determine the initial participation factor of each grid-connected inverter node. Based on the comparison between the initial participation factor and the preset initial partitioning threshold, each grid-connected inverter node is divided into full-order detailed nodes or aggregated equivalent nodes to establish a hybrid equivalent network. The voltage signal of the common connection point in the hybrid equivalent network is collected, and the comprehensive voltage disturbance index is determined based on the voltage signal. When the comprehensive voltage disturbance index exceeds the preset trigger threshold, the real-time pseudo-participation factor is determined based on the instantaneous apparent power fluctuation of the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix. Based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold, the type of node is dynamically switched: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and is continuously delayed for a time window, the node is switched to an aggregated equivalence node. The reduced-order aggregation model is used to simulate the aggregated equivalent nodes, and the full-order detailed model is used to simulate the full-order detailed nodes. Electromagnetic transient simulation is achieved by dynamically switching between the two types of models.

2. The electromagnetic transient simulation method according to claim 1, characterized in that, The determination of the cross-coupling impedance sensitivity matrix based on the voltage response signal and the current response signal specifically includes: The voltage response signal and the current response signal are converted to the frequency domain and divided to obtain a wideband equivalent impedance array. Calculate the absolute value of the partial derivative of the wideband equivalent impedance array with respect to the frequency domain voltage response fluctuation, and use it as an element of the cross-coupled impedance sensitivity matrix; The elements of the cross-coupling impedance sensitivity matrix at frequency ω are determined according to the following formula: in, Let ΔZ be the cross-coupling impedance sensitivity matrix element of node k to node i at frequency ω. i (ω) represents the change in equivalent impedance at node i at frequency ω, ΔU kj (ω) represents the voltage fluctuation at node i caused by the power disturbance injected at node k at frequency ω.

3. The electromagnetic transient simulation method according to claim 2, characterized in that, The process of decomposing the full-order state matrix to determine the initial participation factor for each grid-connected inverter node, and then dividing each grid-connected inverter node into full-order detailed nodes or aggregated equivalent nodes based on the comparison between the initial participation factor and a preset initial partitioning threshold, thereby establishing a hybrid equivalent network, specifically includes: The full-order state matrix is ​​decomposed into eigenvalues, and the weakly damped mode is taken as the initial dominant oscillation mode. Based on the left and right eigenvectors of the full-order state matrix, calculate the initial participation factor of each node in the i-th mode: Among them, w ik v is the k-th element in the left eigenvector corresponding to the i-th mode. ki Let k be the k-th element in the right eigenvector corresponding to the i-th mode. Let the initial participation factor of the k-th node in the i-th mode be denoted as ; When the maximum value of the initial participation factor is less than the preset initial division threshold, and the maximum value of the cross-coupling impedance sensitivity corresponding to the node is less than the preset impedance coupling threshold, the node is divided into an aggregated equal value node; otherwise, it is divided into a full-order detailed node.

4. The electromagnetic transient simulation method according to claim 3, characterized in that, The process of collecting voltage signals from common connection points in the hybrid equivalent network and determining a comprehensive voltage disturbance index based on these voltage signals specifically includes: Collect voltage signals at common connection points in the hybrid equivalent network, and calculate the effective value deviation and time derivative rate of change of the voltage signals; The effective value deviation and the time derivative rate of change of the voltage signal are linearly weighted and summed to obtain the comprehensive voltage disturbance index: ； in, The preset linear weighting coefficients are used, and ΔU is the effective value deviation of the voltage signal. Let be the time derivative rate of change of the voltage signal.

5. The electromagnetic transient simulation method according to claim 4, characterized in that, The step of determining the real-time pseudo-participation factor based on the instantaneous apparent power fluctuation at the common connection point, the initial participation factor, and the cross-coupling impedance sensitivity matrix specifically includes: Calculate the real-time pseudo-participation factor for each node using the following formula: in, Let i be the real-time pseudo-participation factor for node i. Let be the initial participation factor for node i. The total number of nodes. These are the normalized weighting coefficients. The dimension conversion coefficient is... The dominant oscillation frequency The elements of the cross-coupling impedance sensitivity matrix between node i and node k are shown below. Let be the instantaneous apparent power fluctuation at node k.

6. The electromagnetic transient simulation method according to claim 5, characterized in that, The simulation of the aggregated equivalent nodes using a reduced-order aggregation model specifically includes: The discrete state-space parameter matrix is ​​determined based on the off-diagonal elements in the cross-coupling impedance sensitivity matrix. Construct the node admittance matrix of the hybrid isovalued network; By solving the node admittance matrix at the current moment, the instantaneous voltage deviation of each node at the current moment can be obtained. Collect the instantaneous voltage deviation corresponding to the aggregated equivalent nodes; Differential iterative calculations are performed based on the discrete state space parameter matrix and the instantaneous voltage deviation to generate a compensation current command; The compensation current command is converted into a controlled current source signal and injected into the node admittance matrix for solving in the next time step. Repeat the above solution steps until the preset simulation end time is reached to complete the electromagnetic transient simulation of aggregated equivalent nodes.

7. The electromagnetic transient simulation method according to claim 6, characterized in that, The step of generating a compensation current command by performing differential iterative calculations based on the discrete state-space parameter matrix and the instantaneous voltage deviation specifically includes: The discrete system matrix is ​​determined based on the discrete state-space parameter matrix. Discrete input matrix Discrete output matrix and discrete direct feedforward matrix ; At the current discrete time Extract the previous time step Instantaneous voltage deviation and the internal discrete state variables of the previous time step ; According to the discrete system matrix Discrete input matrix Update the internal discrete state variables at the current time step. : ； According to the discrete output matrix Discrete direct feedforward matrix Combined with the instantaneous voltage deviation at the current moment Calculate and generate compensation current command : in, To compensate for current commands, Let this be the vector of internal state variables at the current moment. This represents the instantaneous voltage deviation at the current moment.

8. An electromagnetic transient simulation system, characterized in that, The system includes: a signal acquisition unit, a signal processing unit, a dynamic switching unit, and an electromagnetic transient simulation unit; The signal acquisition unit is used to acquire the voltage response signal and current response signal of each grid-connected inverter node in the system to be simulated under broadband disturbance, as well as the voltage signal of the common connection point in the hybrid equivalent network. The signal processing unit is used to determine the cross-coupling impedance sensitivity matrix based on the voltage response signal and the current response signal; establish a full-order state matrix based on the original full-order electromagnetic transient model of the system to be simulated; decompose the full-order state matrix to determine the initial participation factor of each grid-connected inverter node; divide each grid-connected inverter node into full-order detailed nodes or aggregated equivalent nodes based on the comparison relationship between the initial participation factor and the preset initial partitioning threshold, and establish a hybrid equivalent network; determine the comprehensive voltage disturbance index based on the voltage signal of the point of common coupling; and when the comprehensive voltage disturbance index exceeds the preset triggering threshold, determine the real-time pseudo-participation factor based on the instantaneous apparent power fluctuation of the point of common coupling, the initial participation factor, and the cross-coupling impedance sensitivity matrix. The dynamic switching unit is used to dynamically switch the type of a node based on the comparison result between the real-time pseudo-participation factor and the preset model switching threshold: when the real-time pseudo-participation factor of the aggregated equivalence node is greater than the preset model switching threshold, the node is switched to a full-order detailed node; when the real-time pseudo-participation factor of the full-order detailed node is less than the preset model switching threshold and is continuously delayed for a time window, the node is switched to an aggregated equivalence node. The electromagnetic transient simulation unit is used to simulate the aggregated equivalent nodes using a reduced-order aggregated model and the full-order detailed nodes using a full-order detailed model, thereby realizing electromagnetic transient simulation through dynamic switching between the two types of models.

9. A readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, the processor performs the steps of the method as described in any one of claims 1 to 7.

10. A computer device, comprising a memory and a processor, characterized in that, The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

Images