Method, device and equipment for monitoring interactive coupling behavior of grid-forming converter and grid-following converter and medium

Abstract

This application discloses a method, apparatus, equipment, and medium for monitoring the interaction coupling behavior of grid-type and grid-connected converters, relating to the field of power system technology. The method includes: injecting a disturbance signal into the input signal of a target converter to determine the time-domain response of the output signal; discretizing the time-domain response to obtain data at each frequency point; the target converters include grid-type and grid-connected converters; generating state variables for each frequency point data and mapping each state variable to a preset high-dimensional observable space; generating local models for each target frequency point based on the mapping results; fusing the local models based on target weights to obtain a global model; determining the coupling relationship between the grid-type and grid-connected converters; analyzing the interaction between the grid-type and grid-connected converters based on the coupling relationship; and identifying potential unstable devices or channels based on the analysis results. This application is capable of performing interaction coupling analysis between grid-type and grid-connected converters.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method, device, equipment and medium for monitoring the interactive coupling behavior of grid-connected and grid-linked converters. Background Technology

[0002] With the continuous expansion of power electronic devices for new energy sources such as wind power and photovoltaics, the proportion of traditional synchronous machines has decreased, leading to a reduction in the inertia and damping level of the power system and an increase in grid strength fluctuations. To improve the grid support capacity, converter control methods such as GFM (Grid-Forming) and GFL (Grid-Following) have been developed: Grid-Forming converters actively establish voltage and frequency through voltage source characteristics and power synchronization mechanisms, providing equivalent inertia and damping support; Grid-Following converters typically rely on PLL (Phase-Locked Loop) to achieve synchronous grid connection and follow the grid operation. When different types of converters coexist in the same AC system and interact through a common coupling point, multiple control links such as the power outer loop, voltage outer loop, current inner loop, and PLL will form dynamic interactions across links and devices through grid impedance and power and voltage coupling. This makes the system exhibit dynamic characteristics of multiple time scales, strong coupling, and parameter sensitivity, which can easily trigger low-frequency, subsynchronous, and other oscillation risks at specific operating points or frequency bands, thereby affecting the grid connection and absorption capacity of new energy sources and the safe and stable operation of the system.

[0003] Among existing related technologies, some methods focus on the interaction analysis between specific devices, which limits the research objects and scope and makes it difficult to characterize the complex coupling mechanism of hybrid systems with multiple types of converters. Other technologies focus on the coupling evaluation of static or typical operating conditions, which makes it difficult to reflect the dynamic coupling evolution law under different frequency bands. At the same time, traditional analytical modeling methods are highly dependent on the modeling accuracy of system parameters, topology and controller components, the derivation process is cumbersome and the computational efficiency is low, and the accuracy is difficult to guarantee under nonlinear scenarios such as frequent changes in operating conditions and weak grid fluctuations.

[0004] In summary, how to perform interactive coupling analysis between grid-type converters and mesh-type converters is a technical problem that urgently needs to be solved. Summary of the Invention

[0005] In view of this, the purpose of this invention is to provide a method, apparatus, device, and medium for monitoring the interaction coupling behavior of grid-type and grid-connected converters, capable of performing interaction coupling analysis between grid-type and grid-connected converters. The specific solution is as follows:

[0006] Firstly, this application provides a method for monitoring the interactive coupling behavior of grid-type and grid-connected converters, including:

[0007] Acquire the input and output signals of the target converter; the target converter includes grid-connected converters and grid-linked converters in the power system;

[0008] The input signal is injected with perturbation signals at each target frequency point to determine the time domain response corresponding to the output signal, and the time domain response is discretized to obtain the data at each frequency point.

[0009] The state variables of the power system corresponding to each frequency point data are generated, and each state variable is mapped to a preset high-dimensional observable space. Based on the mapping results, a local model corresponding to each target frequency point is generated.

[0010] The target weights are determined by constraining the target quadratic programming problem, and the global model is obtained by convex combination fusion of the local models of each target frequency point based on the target weights.

[0011] Based on the global model, the target transmission relationship between the input signal and the output signal is determined, and the target coupling relationship between the grid-type converter and the follow-grid converter is determined based on the target transmission relationship. The interaction between the grid-type converter and the follow-grid converter is analyzed based on the target coupling relationship, and potential unstable devices or channels are identified based on the analysis results.

[0012] Optionally, injecting perturbation signals at each target frequency point into the input signal to determine the time-domain response corresponding to the output signal, and discretizing the time-domain response to obtain data at each frequency point, includes:

[0013] Set a preset perturbation frequency and determine each of the target frequency points in the preset perturbation frequency;

[0014] A single-frequency sinusoidal signal at each of the target frequencies is injected into the input signal, and the time-domain response corresponding to the output signal is measured simultaneously.

[0015] The time-domain response is discretized to obtain the sampling sequence of each target frequency point, and the data of each frequency point is determined based on the sampling sequence of each target frequency point.

[0016] Optionally, before mapping each of the state variables to a preset high-dimensional observable space, the method further includes:

[0017] Determine the higher-order terms and coupling terms corresponding to the state variables, and construct an observable function based on the state variables, the higher-order terms and the coupling terms, so as to map the state variables to the preset high-dimensional observable space based on the observable function;

[0018] The state variables, higher-order terms, and coupling terms are each subsets of the observable function.

[0019] Optionally, generating the local model corresponding to each target frequency point based on the mapping result includes:

[0020] For any target frequency, construct a shift matrix and an input matrix based on the mapping result;

[0021] A regression matrix is ​​constructed based on the shift matrix and the input matrix, and singular value decomposition is performed on the regression matrix by truncated SVD. The first orthogonal matrix, the second orthogonal matrix, and the diagonal matrix are determined based on the singular value decomposition results.

[0022] The first orthogonal matrix is ​​divided into blocks, and the local model corresponding to any target frequency point is generated based on the block division result, the second orthogonal matrix, and the diagonal matrix.

[0023] Optionally, the step of dividing the first orthogonal matrix into blocks and generating the local model corresponding to any of the target frequency points based on the block division results, the second orthogonal matrix, and the diagonal matrix includes:

[0024] The first orthogonal matrix is ​​divided into blocks according to the spatial dimension of the preset high-dimensional observable space to obtain the first matrix;

[0025] The first orthogonal matrix is ​​divided into blocks according to the dimension of the input signal to obtain the second matrix;

[0026] The least squares closed-form solution is determined based on the first matrix, the second orthogonal matrix, and the diagonal matrix to obtain the local state space matrix;

[0027] The least squares closed-form solution is determined based on the second matrix, the second orthogonal matrix, and the diagonal matrix to obtain the local input matrix, and the local model is determined based on the local state space matrix and the local input matrix.

[0028] The local state space matrix is ​​used to reflect the coupling and evolution characteristics of the power system state in the preset high-dimensional observable space, and the local input matrix is ​​used to reflect the effect of the disturbance signal on the power system state.

[0029] Optionally, determining the target weights through a constrained quadratic programming problem includes:

[0030] Determine the target correlation between the predicted state variables output by each of the local models, and determine the target matching degree between the predicted state variables output by each of the local models and the state variables corresponding to the mapping results;

[0031] The target quadratic programming problem is constructed based on the target relevance and target matching degree, so that the target weights can be determined by constraining the target quadratic programming problem;

[0032] The target weights are in the form of vectors, and each component of the target weights corresponds one-to-one with the local model of each target frequency point. The constraints are used to control that each component is non-negative and that the sum of each component is 1.

[0033] Optionally, both the target transfer relationship and the target coupling relationship are in the form of matrices;

[0034] Accordingly, determining the target transmission relationship between the input signal and the output signal based on the global model includes:

[0035] Based on the input signal, the output signal, and the global model, a preset feedforward matrix is ​​determined, as well as the target mapping relationship between the output signal and the state variables corresponding to the mapping result;

[0036] The target transfer relationship is determined based on the global model, the target mapping relationship, and the preset feedforward matrix.

[0037] Secondly, this application provides a monitoring device for the interaction coupling behavior of grid-type and grid-connected converters, comprising:

[0038] The signal acquisition module is used to acquire the input and output signals of the target converter; the target converter includes grid-connected converters and grid-linked converters in the power system.

[0039] The disturbance signal injection module is used to inject disturbance signals at each target frequency point into the input signal to determine the time domain response corresponding to the output signal, and to discretize the time domain response to obtain data at each frequency point.

[0040] The local model generation module is used to generate the state variables of the power system corresponding to each frequency point data, map each state variable to a preset high-dimensional observable space, and generate a local model corresponding to each target frequency point based on the mapping result.

[0041] The global model determination module is used to determine the target weights through a constrained target quadratic programming problem, and to perform convex combination fusion of the local models of each target frequency point based on the target weights to obtain a global model;

[0042] The interaction analysis module is used to determine the target transmission relationship between the input signal and the output signal based on the global model, and to determine the target coupling relationship between the grid-type converter and the follow-grid converter based on the target transmission relationship. The interaction between the grid-type converter and the follow-grid converter is analyzed based on the target coupling relationship, and potential unstable devices or channels are identified based on the analysis results.

[0043] Thirdly, this application provides an electronic device, comprising:

[0044] Memory, used to store computer programs;

[0045] A processor is used to execute the computer program to implement the aforementioned method for monitoring the interactive coupling behavior of grid-type and grid-connected converters.

[0046] Fourthly, this application provides a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned method for monitoring the interactive coupling behavior of grid-type and grid-connected converters.

[0047] In this application, the input and output signals of the target converters are first acquired; the target converters include grid-connected converters and grid-connected converters in the power system; then, disturbance signals at each target frequency point are injected into the input signals to determine the time-domain response corresponding to the output signals, and the time-domain response is discretized to obtain data for each frequency point; subsequently, the state variables of the power system corresponding to each frequency point data are generated, and each state variable is mapped to a preset high-dimensional observable space, and a local model corresponding to each target frequency point is generated based on the mapping result; then, the target weights are determined by constrained target quadratic programming problems, and the local models of each target frequency point are convexly combined and fused based on the target weights to obtain a global model; finally, the target transmission relationship between the input signals and the output signals is determined based on the global model, and the target coupling relationship between the grid-connected converters and the grid-connected converters is determined based on the target transmission relationship, so as to analyze the interaction between the grid-connected converters and the grid-connected converters based on the target coupling relationship, and to identify potential unstable devices or channels based on the analysis results. As can be seen from the above, this application first collects the input and output signals of two types of converters, injects disturbance signals at each target frequency into the input signal, obtains the time-domain response of the output signal and performs discretization processing to obtain data at each frequency point; then, the state variables of the power system are generated from the data at each frequency point and mapped to a preset high-dimensional observable space, and a local model of each target frequency point is constructed based on the mapping result; subsequently, the target weights are solved by constrained target quadratic programming problem, and the local models of each target frequency point are convexly combined and fused based on the target weights to obtain a global model adapted to the entire frequency; finally, the target transmission relationship between the input and output signals is determined based on the global model, and the target coupling relationship between the grid-type converter and the grid-connected converter is further derived. Based on the target coupling relationship, the interaction between the grid-type converter and the grid-connected converter is analyzed, so as to accurately locate potential unstable devices or unstable channels in the power system based on the analysis results. In this way, by combining frequency division modeling with global fusion, this application overcomes the limitations of traditional methods in characterizing multi-timescale and strongly coupled characteristics, significantly improving the accuracy of converter coupling mechanism analysis. Furthermore, the data-driven modeling approach reduces the reliance on analytical system models, adapting to complex scenarios such as changing operating conditions and grid fluctuations. It can efficiently identify oscillation risks caused by converter interactions, providing precise technical support for power system stability control and instability prevention, and effectively ensuring the grid connection and consumption of new energy sources and the safe and stable operation of the power system. Attached Figure Description

[0048] 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0049] Figure 1 A flowchart of a method for monitoring the interactive coupling behavior of grid-type and grid-following type converters provided in this application;

[0050] Figure 2 This application provides a flowchart for monitoring the interaction coupling behavior of grid-type and grid-connected converters.

[0051] Figure 3 This application provides a specific topology diagram of a power distribution network testing system;

[0052] Figure 4 A schematic diagram of the interactive coupling behavior monitoring device for grid-type and grid-following type converters provided in this application;

[0053] Figure 5 This application provides a structural diagram of an electronic device. Detailed Implementation

[0054] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0055] With the continuous expansion of power electronic devices for new energy sources such as wind power and photovoltaics, the proportion of traditional synchronous machines is declining, leading to a decrease in the inertia and damping level of the power system and an increase in grid strength fluctuations. To improve the grid's support capacity, converter control methods such as GFM and GFL have been developed. When different types of converters coexist in the same AC system and interact through a common coupling point, multiple control links such as the power outer loop, voltage outer loop, current inner loop, and PLL will form dynamic interactions across links and devices through grid impedance and power / voltage coupling. This causes the system to exhibit dynamic characteristics of multi-timescale, strong coupling, and parameter sensitivity, which can easily trigger low-frequency and subsynchronous oscillation risks at specific operating points or frequency bands, thereby affecting the grid integration and absorption capacity of new energy sources and the safe and stable operation of the system. Among the existing related technologies, some methods focus on the interaction analysis between specific devices, which limits the research objects and scope and makes it difficult to characterize the complex coupling mechanism of multi-type converter hybrid systems; other technologies focus on the coupling evaluation of static or typical operating conditions, which makes it difficult to reflect the dynamic coupling evolution law under different frequency bands. Meanwhile, traditional analytical modeling methods are highly dependent on the accuracy of system parameters, topology, and controller components. The derivation process is cumbersome and computationally inefficient, and accuracy is difficult to guarantee under nonlinear scenarios such as frequent changes in operating conditions and weak grid fluctuations. To address this, this application provides a monitoring scheme for the interactive coupling behavior of grid-connected and grid-linked converters, enabling interactive coupling analysis between them.

[0056] See Figure 1 As shown in the figure, this invention discloses a method for monitoring the interactive coupling behavior of grid-type and grid-following type converters, which may include:

[0057] Step S11: Obtain the input and output signals of the target converter; the target converter includes grid-connected converters and grid-linked converters in the power system.

[0058] In this embodiment, the input signal used to excite the dynamics of the power system is acquired in a power system containing multiple converters, namely grid-connected converters and grid-linked converters. u(t) With the output signal characterizing the system response y(t) As shown below:

[0059] ;

[0060] in, This is the active power reference value for grid-type converters. This is the reactive power reference value for grid-type converters. This is a reference value for the active power of the grid-type converter. This is the reactive power reference value for grid-type converters.

[0061] ;

[0062] in, This refers to the terminal voltage of a grid-type converter. This refers to the terminal current of the grid-type converter. The angular frequency of the grid-type converter. To match the phase-locked loop (PLL) phase of the grid converter, To match the terminal voltage of the grid converter, This represents the active power of the load.

[0063] Step S12: Inject perturbation signals at each target frequency point into the input signal to determine the time domain response corresponding to the output signal, and discretize the time domain response to obtain data at each frequency point.

[0064] In this embodiment, a wideband perturbation generation and injection module is constructed to inject perturbation signals at each target frequency point into the input signal to determine the time-domain response corresponding to the output signal, and to discretize the time-domain response to obtain data for each frequency point. The specific process may include: first, setting a preset perturbation frequency and determining each target frequency point in the preset perturbation frequency; then, injecting a single-frequency sine wave signal at each target frequency point into the input signal and simultaneously measuring the time-domain response corresponding to the output signal; finally, discretizing the time-domain response to obtain the sampling sequence of each target frequency point, and determining the data for each frequency point based on the sampling sequence of each target frequency point.

[0065] Specifically, the perturbation frequency is set as follows:

[0066] ;

[0067] in, Let k be the frequency of the target frequency point. Let be the angular frequency of the k-th target frequency point.

[0068] Then, a perturbation is formed by injecting single-frequency sinusoidal signals at each target frequency point within the perturbation frequency range into the input signal, as shown below:

[0069] ;

[0070] in, Let be the input signal at the k-th target frequency point after the perturbation is injected. The input signal is the one without any disturbance. For the perturbation signal component at the k-th target frequency, To inject the disturbance amplitude into the two input channels; The phase of the disturbance signal injected into the two input channels.

[0071] In one specific implementation, during the injection of the first... When a target frequency point is disturbed, the time-domain response of the output signal at that frequency point is measured synchronously, and the time-domain response is discretized into the corresponding sampling sequence. Therefore, the data segment sets for each frequency point can be obtained, as shown below:

[0072] ;

[0073] in, Let k be the target frequency point, and n be the input signal at time n. For the k-th target frequency point, the output signal at time n is... This represents the number of sampling points for the k-th target frequency.

[0074] Step S13: Generate the state variables of the power system corresponding to each frequency point data, map each state variable to a preset high-dimensional observable space, and generate a local model corresponding to each target frequency point based on the mapping results.

[0075] In this embodiment, in order to approximate the nonlinear dynamics of the power system as a controlled linear form in a preset high-dimensional observable space, and to identify the local state space matrix at each frequency point, the state variables of the power system corresponding to the data at each frequency point are first generated. For the first... k Individual frequency point data D k The state variables are constructed based on the output signal, as shown below:

[0076] ;

[0077] in, Let k be the target frequency point, and n be the state variable at time n. For observable functions, The spatial dimension of the state variable.

[0078] Next, before mapping each state variable to a preset high-dimensional observable space, the process may further include: determining the higher-order terms and coupling terms corresponding to the state variables, and constructing an observable function based on the state variables, the higher-order terms, and the coupling terms, so as to map the state variables to the preset high-dimensional observable space based on the observable function; wherein the state variables, the higher-order terms, and the coupling terms are each a subset of the observable function.

[0079] Specifically, an observable function g(x) is designed, which can be divided into two subsets: g1(x) using the original state variables of the power system as the observable function subsets, and g2(x) and g3(x) using the state variables with explicit coupling relationships and the higher-order terms of the coupling variables as the observable function subsets based on the differential equation expression in the nonlinear model. The original state variables g1(x), the higher-order terms of the state variables g2(x), the coupling terms of the state variables g3(x), and the observable function g(x) are as follows:

[0080] ;

[0081] ;

[0082] ;

[0083] ;

[0084] in, To predetermine the dimension of the high-dimensional observable space, the observable function g(x) must contain at least linear terms, delayed embedding terms, and polynomial or cross terms to enhance the observability and identifiability of coupled dynamics under weak power grids.

[0085] Then, the constructed observable function is used to map the state variables to the corresponding preset high-dimensional observable space to obtain the corresponding high-dimensional state variables. Next, in order to generate the local model corresponding to each target frequency point based on the mapping result, the specific process may include: first, for any target frequency point, constructing a shift matrix and an input matrix based on the mapping result; then, constructing a regression matrix based on the shift matrix and the input matrix, and performing singular value decomposition on the regression matrix using truncated SVD (Singular Value Decomposition), determining the first orthogonal matrix, the second orthogonal matrix, and the diagonal matrix based on the singular value decomposition result; then, dividing the first orthogonal matrix into blocks, and generating the local model corresponding to any target frequency point based on the block division result, the second orthogonal matrix, and the diagonal matrix.

[0086] It should be noted that the above-mentioned process of dividing the first orthogonal matrix into blocks and generating the local model corresponding to any target frequency point based on the block division result, the second orthogonal matrix, and the diagonal matrix can include: firstly, dividing the first orthogonal matrix into blocks according to the spatial dimension of the preset high-dimensional observable space to obtain a first matrix; simultaneously, dividing the first orthogonal matrix into blocks according to the dimension of the input signal to obtain a second matrix; then, determining the least squares closed-form solution based on the first matrix, the second orthogonal matrix, and the diagonal matrix to obtain a local state space matrix; finally, determining the least squares closed-form solution based on the second matrix, the second orthogonal matrix, and the diagonal matrix to obtain a local input matrix, and determining the local model based on the local state space matrix and the local input matrix; wherein, the local state space matrix is ​​used to reflect the coupling and evolution characteristics of the power system state in the preset high-dimensional observable space, and the local input matrix is ​​used to reflect the effect of the disturbance signal on the power system state.

[0087] Specifically, construct the shift matrix. , and the input matrix This is used to reflect the evolution dynamics of the power system in a high-dimensional linear space, as shown below:

[0088] ;

[0089] ;

[0090] in, , q The input dimension.

[0091] Next, a local linear model for each target frequency point is established, as shown below:

[0092] ;

[0093] in, For the first The local state-space matrix of each target frequency point reflects the coupling and evolution characteristics between power system states in a high-dimensional observable space. For the first The local input matrix of each target frequency point reflects the channel and intensity of the effect of the input disturbance on the observable state.

[0094] It should be noted that by introducing a controlled local linear model, the nonlinear response of the original power system in a specified frequency band can be approximated as an identifiable linear controlled system, thereby achieving frequency band dynamic modeling.

[0095] Then, sub-model identification is performed for each frequency, based on the data of each frequency. Construct the regression matrix As shown below:

[0096] ;

[0097] Subsequently, to improve the numerical stability and robustness under conditions of noise disturbance, strong sample correlation, or ill-conditioned regression matrix, the regression matrix was modified... Perform singular value decomposition as follows:

[0098] ;

[0099] in, For the first orthogonal matrix, It is the second orthogonal matrix. It is a diagonal matrix, and its diagonal elements These are singular values, representing the energy distribution of the regression matrix in different principal directions.

[0100] Then According to the spatial dimension of high-dimensional observable space and input dimension q Divide the data into blocks to obtain the first matrix corresponding to the state subspace. and the second matrix corresponding to the input subspace Based on this, the local state space matrix With local input matrix It can be obtained from the least-squares closed-form solution of truncated SVD, as shown below:

[0101] ;

[0102] .

[0103] In this way, by analyzing all target frequencies... Repeating the above process, a set of local models corresponding to each target frequency point can be obtained:

[0104] .

[0105] Step S14: Determine the target weights by constraining the target quadratic programming problem, and perform convex combination fusion of the local models of each target frequency point based on the target weights to obtain the global model.

[0106] In this embodiment, a global fusion module is constructed to build the optimal target weights for solving a constrained quadratic programming problem using a globally observable sequence, and to perform convex combination fusion of the local models at each target frequency point to obtain a global model. That is, the global fusion module is used to fuse the locally controlled linear models obtained at each perturbation frequency point into a unified global model. This approach aims to achieve a continuous and consistent dynamic description across frequency bands, avoiding the fragmented modeling problem caused by local models only being applicable to a single frequency band. Specifically, non-negative weights can be introduced into the local models of each target frequency point and convex combinations can be performed to minimize the one-step prediction error of the fused global model on global samples, thereby determining the optimal target weights and obtaining the global model.

[0107] It should be noted that the global fusion module can determine the target weights by constraining a quadratic programming problem. The specific process may include: first, determining the target correlation between the predicted state variables output by each of the local models, and determining the target matching degree between the predicted state variables output by each of the local models and the state variables corresponding to the mapping results; then, constructing the target quadratic programming problem based on the target correlation and target matching degree, so as to determine the target weights by constraining the target quadratic programming problem; wherein, the target weights are in the form of a vector, and each component of the target weights corresponds one-to-one with the local model of each target frequency point, and the constraints are used to control that each component is non-negative and the sum of each component is 1.

[0108] Specifically, let's assume that for the first... k Local model weights are assigned to each target frequency point. Construct the target weights in vector form as follows:

[0109] ;

[0110] Apply convex combination constraints to the target weights as follows:

[0111] ;

[0112] Among them, constraints This is used to ensure that the global model is a convex combination of the local models, thereby maintaining the physical interpretability and numerical stability of the parameters; constraints This is used to ensure that the fused global model does not introduce additional scale bias and to ensure that the fusion result has a consistent amplitude scale across different frequency points.

[0113] Next, the QP (Quadratic Programming) method is introduced to perform weighted fusion of local models at each frequency point, with the goal of minimizing the global prediction error:

[0114] ;

[0115] in, Let be the true state vector at time n+1 in the high-dimensional observable space. This is a weighted sum of all local model predictions.

[0116] right Expanding the equation yields a quadratic form, which can be rewritten as a standard quadratic programming problem, as shown below:

[0117] ;

[0118] Where H represents the target correlation between different local model prediction terms, i.e., the output predicted state variables. f It represents the degree of matching between the local model predictions and the real samples, that is, the high-dimensional state variables corresponding to the mapping results.

[0119] By solving the quadratic programming problem, the optimal target weights that minimize the global prediction error can be obtained. Finally, convex combination fusion is performed on the local models to obtain the global model, as shown below:

[0120] .

[0121] Step S15: Determine the target transmission relationship between the input signal and the output signal based on the global model, and determine the target coupling relationship between the grid-type converter and the follow-grid converter based on the target transmission relationship. Analyze the interaction between the grid-type converter and the follow-grid converter based on the target coupling relationship, and identify potential unstable devices or channels based on the analysis results.

[0122] In this embodiment, the target transfer relationship between the input signal and the output signal is determined based on a global model. The specific process may include: first, determining a preset feedforward matrix and a target mapping relationship between the output signal and the state variables corresponding to the mapping result based on the input signal, the output signal, and the global model; then, determining the target transfer relationship based on the global model, the target mapping relationship, and the preset feedforward matrix.

[0123] Specifically, the output equation is first constructed based on the global model, as shown below:

[0124] ;

[0125] Where C is the state determined based on the output signal, serving as the output mapping matrix; and D is a preset feedforward matrix. In one specific implementation, D is set to 0 when the output quantity does not contain an algebraic pass-through term to the input quantity; otherwise, non-zero elements of D are set according to the system structure correspondence.

[0126] The output mapping matrix C and the preset feedforward matrix D are obtained through the output equation. Based on these, the target transfer relationship in matrix form, i.e., the input-output transfer matrix, is constructed as follows:

[0127] ;

[0128] Where s is a complex frequency variable and I is the identity matrix.

[0129] Subsequently, based on the target transfer relationship, the target coupling relationship between the grid-type converter and the root-grid type converter is determined in matrix form, as the relative coupling matrix, as shown below:

[0130] ;

[0131] in, Represents element-wise product. It is a Moore–Penrose pseudo-inverse.

[0132] It is understandable that the relative coupling matrix The description refers to the situation when other controls are added. The degree of influence of the control loop is analyzed in this embodiment through frequency scanning using the relative coupling matrix. Middle elements The amplitude variation within a specific frequency range is used to determine the strength of electrical and control interactions between grid-connected converters and grid-connected converters, thereby identifying the range of strong coupling frequency bands and their corresponding coupling control channels, and locating potential unstable devices or channels in the power system.

[0133] In one specific implementation, see Figure 2 As shown, the specific monitoring process for the interaction coupling behavior of grid-connected and grid-linked converters can be described as follows:

[0134] (1) Select the input signal used to excite the system dynamics and the output signal used to characterize the system response in the power system.

[0135] Input signals: voltage reference values, active and reactive power reference values ​​for grid-type converters, and voltage reference values ​​and active power reference values ​​for integrated grid-type converters, etc.

[0136] The output signals include the virtual angular frequency, the AC bus voltage and current of the grid converter, the phase-locked loop phase of the grid converter, and the active and reactive power of the power grid lines.

[0137] (2) Construct a wideband disturbance generation and injection module, inject disturbances at each frequency point, and measure the time domain response of the output signal.

[0138] The "wideband disturbance generation and injection module" injects sinusoidal disturbance signals of disturbance frequency into the input signal one after another, while measuring the time domain response of the output signal.

[0139] (3) Observation function design and frequency-by-frequency local model identification

[0140] First, construct the state variables; then design the observable function to map the state variables to a high-dimensional observable space; then, for the data at each disturbance frequency point, construct the shift matrix and the input matrix, and then establish a frequency-by-frequency controlled linear model, i.e., a local model.

[0141] (4) Construct a global fusion module, solve for the target weights and perform convex combination to obtain the global model.

[0142] The target QP problem is constructed based on the globally observable sequence and combined with convex combination constraints. The optimal target weights are obtained by solving the constrained target QP problem. The local models of each target frequency point are combined according to the target weights to obtain a global model covering the entire frequency.

[0143] (5) Construction of transfer matrix and interaction analysis

[0144] Construct an input-output transfer matrix, i.e., the target transfer relationship; construct a relative coupling matrix, i.e., the target coupling relationship; finally, determine the frequency band range of strong coupling and the coupling control loop through the relative coupling matrix, providing a basis for subsequent identification of potentially unstable devices or channels.

[0145] As can be seen from the above, in this embodiment, the input and output signals of the target converter are first acquired; the target converter includes grid-type converters and grid-connected converters in the power system; then, disturbance signals at each target frequency point are injected into the input signal to determine the time-domain response corresponding to the output signal, and the time-domain response is discretized to obtain data at each frequency point; subsequently, the state variables of the power system corresponding to each frequency point data are generated, and each state variable is mapped to a preset high-dimensional observable space, and a local model corresponding to each target frequency point is generated based on the mapping result; then, the target weight is determined by constrained target quadratic programming problem, and the local model of each target frequency point is convexly combined and fused based on the target weight to obtain a global model; finally, the target transmission relationship between the input signal and the output signal is determined based on the global model, and the target coupling relationship between the grid-type converter and the grid-connected converter is determined based on the target transmission relationship, so as to analyze the interaction between the grid-type converter and the grid-connected converter based on the target coupling relationship, and to identify potential unstable devices or channels based on the analysis results. As can be seen from the above, in this embodiment, the input and output signals of two types of converters are first collected. Disturbance signals at each target frequency point are injected into the input signal, and the time-domain response of the output signal is obtained and discretized to obtain data at each frequency point. Then, the state variables of the power system are generated from the data at each frequency point and mapped to a preset high-dimensional observable space. Based on the mapping results, a local model of each target frequency point is constructed. Subsequently, the target weights are solved by constrained target quadratic programming problem. Based on the target weights, the local models of each target frequency point are convexly combined and fused to obtain a global model adapted to the entire frequency range. Finally, the target transmission relationship between the input and output signals is determined based on the global model. The target coupling relationship between the grid-type converter and the grid-connected converter is further derived. Based on the target coupling relationship, the interaction between the grid-type converter and the grid-connected converter is analyzed to accurately locate potential unstable devices or unstable channels in the power system based on the analysis results. In this way, by combining frequency division modeling with global fusion, this embodiment overcomes the limitations of traditional methods in characterizing multi-timescale and strongly coupled characteristics, significantly improving the accuracy of converter coupling mechanism analysis. Furthermore, the data-driven modeling approach reduces the reliance on analytical system models, adapting to complex scenarios such as changing operating conditions and grid fluctuations. It can efficiently identify oscillation risks caused by converter interactions, providing precise technical support for power system stability control and instability prevention, and effectively ensuring the grid connection and consumption of new energy sources and the safe and stable operation of the power system.

[0146] Next, this embodiment will describe the monitoring process of the interaction coupling behavior of grid-type and grid-following type converters in combination with specific implementation scenarios.

[0147] Simulation tests were conducted based on a distribution network test system. The system includes a hybrid energy storage station (12MW) connected to the grid and a grid-connected station (2MW) installed on the 35kV busbar at Xinmin station; a grid-connected energy storage station (5MW) installed on the 10kV busbar at Hongxing station; and a 3.5MW centralized photovoltaic system installed on the 35kV busbar at Hongxing station, with a total network load of 19.195MW. The equivalent controlled linear model, i.e., the global model, was identified using simulation data, and then interactive coupling analysis was performed on different control loops. The topology of the distribution network test system is as follows: Figure 3 As shown in Table 1, the parameters are used, and the simulation platform is DIgSLENT / PowerFactory.

[0148] Table 1 Distribution Network Test System Parameters

[0149]

[0150] The interaction between the grid-connected converter GFM1 and the grid-connected converter GFL in the distribution network test system is analyzed. The angular frequency and terminal voltage of GFM1 and the phase angle and terminal voltage of GFL are selected as output signals. The active power reference values ​​of the grid-connected converter and the grid-connected converter are selected as input signals. The RCA (Relative Coupling Array) of the system is shown below:

[0151] ;

[0152] in, and For input signal, This is the active power reference value for the grid-type converter GFM1. To provide a reference value for the active power of the grid-connected converter GFL, , , , For output signal, This is the terminal voltage of the grid-connected converter GFM1. The angular frequency of the grid-type converter GFM1, The phase-locked loop phase of the grid-type converter GFM1, The terminal voltage of the grid converter GFL Indicates at a frequency of At that time, the system's first The input signal pairs the first The coupling gain of each output signal, such as Indicates the input signal For output signal The coupling gain. When the coupling gain of a certain input-output pair... A value close to 1 indicates that the input signal has a positive influence on the output signal, and the coupling effect between the current input / output pair and other input / output pairs is relatively small; when the coupling gain of a certain input / output pair is close to 1, it indicates that the input signal has a positive influence on the output signal, and the coupling effect between the current input / output pair and other input / output pairs is relatively small. A value less than 0 indicates that the input signal has a negative impact on the output signal, and that the current input-output pair has a strong coupling effect with other input-output pairs; when the coupling gain of the input-output pair is less than 0, it indicates that the input signal has a negative impact on the output signal, and that the current input-output pair has a strong coupling effect with other input-output pairs; A value greater than 1 indicates that the current input / output pair has a strong coupling effect with other input / output pairs.

[0153] By injecting multi-frequency disturbances into the active power reference values ​​of the grid-type converter GFM1 and the grid-connected converter GFL, the time-domain response of the output signal is obtained, a pre-defined high-dimensional observable space is constructed, and the global model is identified. Experiments are conducted under disturbances at frequencies of 0.8 Hz, 3.5 Hz, 6.5 Hz, and 45 Hz. Based on the time-domain response relationship between the predicted values ​​of the equivalent controlled linear model and the original model regarding the angular frequency, port AC voltage, port effective line AC current of the grid-type converter, and the phase angle and port AC voltage of the phase-locked loop of the grid-type converter, it can be seen that the equivalent controlled linear model can accurately reproduce the time-domain response of the key state variables of the original model under different disturbance frequencies, and the dynamic trajectories of the two models match well. This indicates that the equivalent controlled linear model can be used as a linear surrogate model for the original system under small disturbance conditions for subsequent small-signal stability and modal characteristic analysis.

[0154] The interaction between the grid-type converter GFM1 and the grid-connected converter GFL in the distribution network test system is analyzed: Based on the frequency variation of each element in the RCA matrix of the simulated distribution network test system, it can be seen that the input-output pairs of the RCA are stable in the 10-100Hz range, and the RCA matrix... and The average value approaches 1, which means and The smaller the interaction coupling between the control loop and other input / output control loops, the better. The RCA variation is mainly located in the low-frequency range, i.e., 0.1-10Hz, with the peak values ​​of different input / output pairs of the RCA appearing around 1.6Hz, indicating strong interaction coupling between the input and output signals. Therefore, the frequency range of 0.1-10Hz falls within the strongly coupled frequency range of the distribution network testing system.

[0155] To verify the correctness of the RCA principle analysis, reference power disturbances of different frequencies were injected into the grid-type converter. dP GFM1 A simulation of the power system under the given conditions was performed, and the dynamic response curve of the system was obtained through the simulation. It can be seen that applying a reference power disturbance to the power system... dP GFM1 ,when dP GFM1Severe oscillations can be observed at frequencies of 0.1Hz and 1.5Hz. At 0.1Hz, strong oscillations occur with the phase-locked loop (PLL) phase angle mode of the grid converter. And when... dP GFM1 At a frequency of 1.5Hz, both the grid-type converter and the integrated grid-type converter experience severe oscillations, i.e. U GFM1 Virtual angular frequency , and U GFL Oscillations occurred at all frequencies. In contrast, at other frequencies... dP GFM1 The disturbances did not induce severe oscillations, especially at frequencies above 10 Hz. This indicates that in the case of strong coupling (0.1-10 Hz), the interaction effects of the system are significant, leading to severe oscillations and an inability to maintain stability, consistent with the analysis results of RCA.

[0156] As can be seen from the above, this embodiment identifies the equivalent controlled linear model of the system in a preset high-dimensional observable space using a data-driven approach. This eliminates the need to derive complex input-output transfer function matrices channel by channel, significantly reducing the modeling workload and mathematical derivation difficulty, and improving computational efficiency. Furthermore, by obtaining a consistent global model across frequency bands through frequency-point identification and global fusion mechanisms, it can effectively characterize the dynamic evolution of coupling relationships under different frequency bands, improving the applicability for evaluating interactions under complex operating conditions such as weak grid fluctuations and changes in operating points. In addition, this embodiment constructs an RCA matrix based on the global model to achieve quantitative identification of strongly coupled frequency bands and corresponding coupled control loops, and can further locate potential unstable devices or channels, improving the stability and security of the power grid system and providing important support for the coordinated control of multiple converters.

[0157] Accordingly, see Figure 4 As shown in the illustration, this application also provides a device for monitoring the interaction coupling behavior of grid-type and grid-connected converters, which may include:

[0158] The signal acquisition module 11 is used to acquire the input and output signals of the target converter; the target converter includes grid-connected converters and grid-linked converters in the power system.

[0159] The disturbance signal injection module 12 is used to inject disturbance signals at each target frequency point into the input signal to determine the time domain response corresponding to the output signal, and to discretize the time domain response to obtain data at each frequency point.

[0160] The local model generation module 13 is used to generate the state variables of the power system corresponding to each frequency point data, map each state variable to a preset high-dimensional observable space, and generate a local model corresponding to each target frequency point based on the mapping result.

[0161] The global model determination module 14 is used to determine the target weights through a constrained target quadratic programming problem, and to perform convex combination fusion of the local models of each target frequency point based on the target weights to obtain a global model;

[0162] The interaction analysis module 15 is used to determine the target transmission relationship between the input signal and the output signal based on the global model, and to determine the target coupling relationship between the grid-type converter and the follow-grid converter based on the target transmission relationship, so as to analyze the interaction between the grid-type converter and the follow-grid converter based on the target coupling relationship, and to identify potential unstable devices or channels based on the analysis results.

[0163] In some specific embodiments, the disturbance signal injection module 12 may include:

[0164] The target frequency point determination unit is used to set a preset disturbance frequency and determine each of the target frequency points in the preset disturbance frequency;

[0165] The disturbance signal injection unit is used to inject single-frequency sinusoidal signals at each of the target frequencies into the input signal and simultaneously measure the time-domain response corresponding to the output signal.

[0166] The frequency point data determination unit is used to discretize the time domain response to obtain the sampling sequence of each target frequency point, and determine the frequency point data according to the sampling sequence of each target frequency point.

[0167] In some specific embodiments, the monitoring device for the interaction coupling behavior of grid-type and grid-connected converters may further include:

[0168] An observable function construction module is used to determine the higher-order terms and coupling terms corresponding to the state variables, and to construct observable functions based on the state variables, the higher-order terms and the coupling terms, so as to map the state variables to the preset high-dimensional observable space based on the observable functions; wherein the state variables, the higher-order terms and the coupling terms are each a subset of the observable functions.

[0169] In some specific embodiments, the local model generation module 13 may include:

[0170] The input matrix construction submodule is used to construct a shift matrix and an input matrix based on the mapping result for any target frequency point;

[0171] The singular value decomposition submodule is used to construct a regression matrix based on the shift matrix and the input matrix, and to perform singular value decomposition on the regression matrix by truncated SVD, and to determine the first orthogonal matrix, the second orthogonal matrix and the diagonal matrix based on the singular value decomposition result.

[0172] The local model generation submodule is used to divide the first orthogonal matrix into blocks, and generate the local model corresponding to any of the target frequency points based on the block division results, the second orthogonal matrix, and the diagonal matrix.

[0173] In some specific implementations, the local model generation submodule may include:

[0174] The first matrix determination unit is used to divide the first orthogonal matrix into blocks according to the spatial dimension of the preset high-dimensional observable space to obtain the first matrix.

[0175] The second matrix determination unit is used to divide the first orthogonal matrix into blocks according to the dimension of the input signal to obtain the second matrix;

[0176] The local state space matrix determination unit is used to determine the least squares closed solution based on the first matrix, the second orthogonal matrix and the diagonal matrix to obtain the local state space matrix;

[0177] A local model generation unit is used to determine a least-squares closed-form solution based on the second matrix, the second orthogonal matrix, and the diagonal matrix to obtain a local input matrix, and to determine the local model based on the local state space matrix and the local input matrix; wherein, the local state space matrix is ​​used to reflect the coupling and evolution characteristics of the power system state in the preset high-dimensional observable space, and the local input matrix is ​​used to reflect the effect of the disturbance signal on the power system state.

[0178] In some specific embodiments, the global model determination module 14 may include:

[0179] The target matching degree determination unit is used to determine the target correlation between the predicted state variables output by each of the local models, and to determine the target matching degree between the predicted state variables output by each of the local models and the state variables corresponding to the mapping result.

[0180] The target weight determination unit is used to construct the target quadratic programming problem based on the target relevance and target matching degree, so as to determine the target weight by constraining the target quadratic programming problem; wherein, the target weight is in the form of a vector, each component of the target weight corresponds one-to-one with the local model of each target frequency point, and the constraint is used to control that each component is non-negative and the sum of each component is 1.

[0181] In some specific implementations, both the target transfer relationship and the target coupling relationship are in the form of matrices;

[0182] Accordingly, the interaction analysis module 15 may include:

[0183] The target mapping relationship determination unit is used to determine a preset feedforward matrix and a target mapping relationship between the output signal and the state variables corresponding to the mapping result based on the input signal, the output signal and the global model.

[0184] The target transfer relationship determination unit is used to determine the target transfer relationship based on the global model, the target mapping relationship, and the preset feedforward matrix.

[0185] Furthermore, embodiments of this application also disclose an electronic device, Figure 5 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content of the diagram should not be construed as limiting the scope of this application. Specifically, the electronic device 20 may include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 stores a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the interaction coupling behavior monitoring method for grid-type and grid-connected converters disclosed in any of the foregoing embodiments. Furthermore, the electronic device 20 in this embodiment may specifically be a computer.

[0186] In this embodiment, the power supply 23 is used to provide operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and external devices, and the communication protocol it follows can be any communication protocol applicable to the technical solution of this application, and is not specifically limited here; the input / output interface 25 is used to acquire external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0187] In addition, the memory 22, as a carrier for resource storage, can be a read-only memory, random access memory, disk or optical disk, etc. The resources stored thereon can include operating system 221, computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0188] The operating system 221 is used to manage and control the various hardware devices on the electronic device 20 and the computer program 222, which may be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program capable of performing the monitoring method for the interaction coupling behavior of network-type and grid-connected converters executed by the electronic device 20 as disclosed in any of the foregoing embodiments, the computer program 222 may further include computer programs capable of performing other specific tasks.

[0189] Furthermore, this application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, it implements the aforementioned method for monitoring the interactive coupling behavior of grid-type and grid-connected converters. Specific steps of this method can be found in the corresponding content disclosed in the foregoing embodiments, and will not be repeated here.

[0190] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0191] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0192] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0193] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0194] The technical solutions provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for monitoring the interactive coupling behavior of grid-type and grid-following type converters, characterized in that, include: Acquire the input and output signals of the target converter; The target converters include grid-connected converters and grid-linked converters in the power system; The input signal is injected with perturbation signals at each target frequency point to determine the time domain response corresponding to the output signal, and the time domain response is discretized to obtain the data at each frequency point. The state variables of the power system corresponding to each frequency point data are generated, and each state variable is mapped to a preset high-dimensional observable space. Based on the mapping results, a local model corresponding to each target frequency point is generated. The target weights are determined by constraining the target quadratic programming problem, and the global model is obtained by convex combination fusion of the local models of each target frequency point based on the target weights. Based on the global model, the target transmission relationship between the input signal and the output signal is determined, and the target coupling relationship between the grid-type converter and the follow-grid converter is determined based on the target transmission relationship. The interaction between the grid-type converter and the follow-grid converter is analyzed based on the target coupling relationship, and potential unstable devices or channels are identified based on the analysis results.

2. The method for monitoring the interactive coupling behavior of grid-type and grid-following type converters according to claim 1, characterized in that, The process of injecting perturbation signals at each target frequency point into the input signal to determine the time-domain response corresponding to the output signal, and discretizing the time-domain response to obtain data at each frequency point, includes: Set a preset perturbation frequency and determine each of the target frequency points in the preset perturbation frequency; A single-frequency sinusoidal signal at each of the target frequencies is injected into the input signal, and the time-domain response corresponding to the output signal is measured simultaneously. The time-domain response is discretized to obtain the sampling sequence of each target frequency point, and the data of each frequency point is determined based on the sampling sequence of each target frequency point.

3. The method for monitoring the interactive coupling behavior of grid-type and grid-following type converters according to claim 1, characterized in that, Before mapping each of the state variables to a preset high-dimensional observable space, the method further includes: Determine the higher-order terms and coupling terms corresponding to the state variables, and construct an observable function based on the state variables, the higher-order terms and the coupling terms, so as to map the state variables to the preset high-dimensional observable space based on the observable function; The state variables, higher-order terms, and coupling terms are each subsets of the observable function.

4. The method for monitoring the interactive coupling behavior of grid-type and grid-following type converters according to claim 1, characterized in that, The process of generating local models corresponding to each target frequency point based on the mapping results includes: For any target frequency, construct a shift matrix and an input matrix based on the mapping result; A regression matrix is ​​constructed based on the shift matrix and the input matrix, and singular value decomposition is performed on the regression matrix by truncated SVD. The first orthogonal matrix, the second orthogonal matrix, and the diagonal matrix are determined based on the singular value decomposition results. The first orthogonal matrix is ​​divided into blocks, and the local model corresponding to any target frequency point is generated based on the block division result, the second orthogonal matrix, and the diagonal matrix.

5. The method for monitoring the interactive coupling behavior of grid-type and grid-following type converters according to claim 4, characterized in that, The step of dividing the first orthogonal matrix into blocks and generating the local model corresponding to any target frequency point based on the block division result, the second orthogonal matrix, and the diagonal matrix includes: The first orthogonal matrix is ​​divided into blocks according to the spatial dimension of the preset high-dimensional observable space to obtain the first matrix; The first orthogonal matrix is ​​divided into blocks according to the dimension of the input signal to obtain the second matrix; The least squares closed-form solution is determined based on the first matrix, the second orthogonal matrix, and the diagonal matrix to obtain the local state space matrix; The least squares closed-form solution is determined based on the second matrix, the second orthogonal matrix, and the diagonal matrix to obtain the local input matrix, and the local model is determined based on the local state space matrix and the local input matrix. The local state space matrix is ​​used to reflect the coupling and evolution characteristics of the power system state in the preset high-dimensional observable space, and the local input matrix is ​​used to reflect the effect of the disturbance signal on the power system state.

6. The method for monitoring the interactive coupling behavior of grid-type and grid-following type converters according to claim 1, characterized in that, The determination of target weights through a constrained quadratic programming problem includes: Determine the target correlation between the predicted state variables output by each of the local models, and determine the target matching degree between the predicted state variables output by each of the local models and the state variables corresponding to the mapping results; The target quadratic programming problem is constructed based on the target relevance and target matching degree, so that the target weights can be determined by constraining the target quadratic programming problem; The target weights are in the form of vectors, and each component of the target weights corresponds one-to-one with the local model of each target frequency point. The constraints are used to control that each component is non-negative and that the sum of each component is 1.

7. The method for monitoring the interactive coupling behavior of grid-type and grid-following type converters according to any one of claims 1 to 6, characterized in that, Both the target transfer relationship and the target coupling relationship are in the form of matrices; Accordingly, determining the target transmission relationship between the input signal and the output signal based on the global model includes: Based on the input signal, the output signal, and the global model, a preset feedforward matrix is ​​determined, as well as the target mapping relationship between the output signal and the state variables corresponding to the mapping result; The target transfer relationship is determined based on the global model, the target mapping relationship, and the preset feedforward matrix.

8. A monitoring device for the interactive coupling behavior of grid-type and grid-following type converters, characterized in that, include: The signal acquisition module is used to acquire the input and output signals of the target converter. The target converters include grid-connected converters and grid-linked converters in the power system; The disturbance signal injection module is used to inject disturbance signals at each target frequency point into the input signal to determine the time domain response corresponding to the output signal, and to discretize the time domain response to obtain data at each frequency point. The local model generation module is used to generate the state variables of the power system corresponding to each frequency point data, map each state variable to a preset high-dimensional observable space, and generate a local model corresponding to each target frequency point based on the mapping result. The global model determination module is used to determine the target weights through a constrained target quadratic programming problem, and to perform convex combination fusion of the local models of each target frequency point based on the target weights to obtain a global model; The interaction analysis module is used to determine the target transmission relationship between the input signal and the output signal based on the global model, and to determine the target coupling relationship between the grid-type converter and the follow-grid converter based on the target transmission relationship. The interaction between the grid-type converter and the follow-grid converter is analyzed based on the target coupling relationship, and potential unstable devices or channels are identified based on the analysis results.

9. An electronic device, characterized in that, The electronic device includes a processor and a memory; wherein the memory is used to store a computer program, which is loaded and executed by the processor to implement the method for monitoring the interactive coupling behavior of grid-type and grid-following converters as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, Used to store a computer program, which, when executed by a processor, implements the method for monitoring the interactive coupling behavior of grid-type and grid-following type converters as described in any one of claims 1 to 7.

Images