Grid-side converter fault ride-through control structure modeling method, system, device and medium

Abstract

This invention belongs to the field of new energy power generation grid connection technology, and discloses a modeling method, system, equipment, and medium for grid-side converter fault ride-through control structure. It addresses the problems of difficult modeling, high cost, poor extrapolation, and lack of versatility in existing technologies due to reliance on manufacturer-internal logic or massive experimental data. The method acquires disturbance input and current output data under finite operating conditions, constructs an up-dimensional mapping function to map the input to a high-dimensional feature space, establishes a global linear mapping model based on ridge regression to solve the Koopman operator, predicts active and reactive current reference components, and finally generates the current transient response curve during the fault period using the inner-loop time-domain model of the input current. This invention requires only a small number of samples to achieve high-precision modeling and prediction, has good extrapolation and versatility, and can be adapted to various fault ride-through strategies.

Description

Technical Field

[0001] This invention belongs to the field of new energy power generation grid connection technology, specifically relating to the modeling method, system, equipment and medium for grid-side converter fault ride-through control structure. Background Technology

[0002] As the penetration rate of new energy power generation in the power system continues to increase, the grid-connected operation characteristics of new energy generator units (such as wind turbine generator units and photovoltaic power generation systems) are having an increasingly significant impact on the safety and stability of the power system. Grid connection guidelines require that new energy generator units must have fault ride-through capability when the grid experiences fault disturbances such as voltage dips or surges, that is, they must not disconnect from operation during the fault period and inject necessary reactive current into the grid to support voltage recovery according to voltage changes.

[0003] Currently, the modeling and performance evaluation of fault ride-through control structures for new energy generating units typically employ the following two technical approaches:

[0004] The first category is based on mechanism modeling. This method requires obtaining the detailed internal logic of the fault ride-through control strategy of the grid-side converter, including limiting elements, dead-zone characteristics, and proportional-integral (PI) parameter switching conditions, and then constructing an accurate mathematical model to reproduce its dynamic behavior. However, in practical applications, the fault ride-through control strategy, as the core control algorithm, is usually encapsulated and protected by the equipment manufacturer, making its internal logic difficult to obtain. This results in significant limitations of this method in practical engineering applications.

[0005] The second category is based on purely data-driven modeling methods. This method does not rely on the internal logic of the controller, but instead collects a large amount of experimental data and uses machine learning algorithms such as neural networks and support vector machines to fit the nonlinear mapping relationship between the input (e.g., voltage dip depth) and the output (e.g., current reference value). The modeling accuracy of this type of method heavily depends on the completeness of the training samples. It typically requires conducting experiments covering the entire operating range to obtain thousands or even tens of thousands of sets of response data under different voltage dip amplitudes and different combinations of control parameters, resulting in long experimental cycles, high hardware consumption, and high costs. More importantly, when the input operating conditions exceed the coverage of the training samples, the model's prediction accuracy drops sharply, meaning its extrapolation ability is poor, making it difficult to meet the diverse and uncertain requirements of actual power grid fault conditions.

[0006] Therefore, how to achieve unified and high-precision modeling of different fault ride-through control strategies and possess good operating condition extrapolation capabilities without relying on the manufacturer's internal control logic and using only limited externally measurable data is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0007] Based on the aforementioned shortcomings and deficiencies in the prior art, one of the objectives of this invention is to at least solve one or more of the aforementioned problems in the prior art. In other words, one of the objectives of this invention is to provide a grid-side converter fault ride-through control structure modeling method, system, device, and medium that meets one or more of the aforementioned requirements, so as to solve the problems of modeling difficulties, high costs, poor extrapolation, and lack of universality caused by reliance on the manufacturer's internal logic or massive experimental data in the prior art.

[0008] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0009] In a first aspect, the present invention provides a method for modeling the fault ride-through control structure of a grid-side converter, comprising the following steps:

[0010] S1. Obtain disturbance input data and corresponding current output data under finite operating conditions;

[0011] S2. Construct a dimension-upgrading mapping function to map the perturbation input data to a high-dimensional feature space to obtain a high-dimensional feature vector;

[0012] S3. Based on the high-dimensional feature vector and the current output data, solve the Koopman operator through the ridge regression optimization framework to establish a global linear mapping model from the high-dimensional feature space to the current output data;

[0013] S4. Input the new disturbance input data into the global linear mapping model to obtain the predicted active current reference component and reactive current reference component.

[0014] S5. Input the predicted active current reference component and reactive current reference component into the current inner loop time domain model, and combine the proportional-integral controller parameters and filter inductor parameters to solve the differential equation and generate the current transient response curve during the fault ride-through process.

[0015] As a preferred option:

[0016] The disturbance input data includes voltage drop amplitude characterizing the degree of grid fault, and characteristic parameters characterizing the fault ride-through control strategy.

[0017] The current output data includes active current reference components and reactive current reference components.

[0018] As a preferred embodiment, the characteristic parameters representing the fault ride-through control strategy are determined according to the control strategy type, specifically as follows:

[0019] When the control strategy is maximum current control, the characteristic parameters include the maximum current limit and the reactive power support coefficient.

[0020] When the control strategy is constant active power control, the characteristic parameters include the reactive power support coefficient and the constant active power setpoint.

[0021] When the control strategy is constant active current control, the characteristic parameters include the reactive power support coefficient and the constant active current setpoint.

[0022] As a preferred option:

[0023] The updimensional mapping function transforms the nonlinear input-output relationship in the original space into a linear relationship in the higher-dimensional space, including linear terms, quadratic terms, cross terms, combination terms, nonlinear function terms, and constant terms.

[0024] As a preferred embodiment, the objective function of the ridge regression optimization framework is:

[0025] ,

[0026] In the formula, For current output data matrix, Let be the Koopman operator matrix to be solved. It is a high-dimensional feature matrix. For regularization parameters, It is the square of the Frobenius norm of the matrix.

[0027] As a preferred option:

[0028] The regularization parameter The value range is

[10] -4 10 -1 ].

[0029] As a preferred option, it also includes:

[0030] The current transient response curve is compared with the grid fault ride-through standard limit. If the limit is exceeded, the proportional-integral controller parameters are adjusted and the process of generating the current transient response curve is repeated until the standard requirements are met.

[0031] In a second aspect, the present invention provides a grid-side converter fault ride-through control structure modeling system for implementing the grid-side converter fault ride-through control structure modeling method as described in the first aspect, comprising:

[0032] The data acquisition module is used to acquire disturbance input data and its corresponding current output data under limited operating conditions;

[0033] The dimension-up mapping module is used to construct a dimension-up mapping function to map the perturbation input data to a high-dimensional feature space to obtain a high-dimensional feature vector.

[0034] The Koopman modeling module is used to solve the Koopman operator based on the high-dimensional feature vector and the current output data through the ridge regression optimization framework, so as to establish a global linear mapping model from the high-dimensional feature space to the current output data.

[0035] The prediction module is used to input new disturbance input data into the global linear mapping model to obtain the predicted active current reference component and reactive current reference component.

[0036] The time-domain simulation module is used to input the predicted active current reference components and reactive current reference components into the current inner loop time-domain model, and solve the differential equations by combining the proportional-integral controller parameters and the filter inductor parameters to generate the current transient response curve during the fault ride-through process.

[0037] Thirdly, the present invention provides an electronic device, the computer device including a memory, a processor and a computer program, wherein when the computer program is executed by the processor, it implements the grid-side converter fault ride-through control structure modeling method as described in the first aspect.

[0038] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the grid-side converter fault ride-through control structure modeling method as described in the first aspect.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1. This invention does not require obtaining the internal logic of the fault ride-through control strategy packaged by the manufacturer. It can achieve equivalent modeling only through externally measurable input and output data. This solves the problem that traditional detailed modeling methods are difficult to apply in practice because the core algorithm is not disclosed. It provides third-party testing agencies, power grid operators and research institutes with a technical evaluation method that is independent of the equipment manufacturer.

[0041] 2. This invention transforms nonlinear relationships into linear relationships in a high-dimensional space through dimensionality-up mapping and employs ridge regression to solve the Koopman operator. This allows for the establishment of a high-precision global linear model with only tens to hundreds of sets of sample data under limited operating conditions. Compared to traditional pure data-driven methods that require thousands or even tens of thousands of samples, this invention significantly reduces data acquisition costs and shortens the experimental cycle considerably.

[0042] 3. The Koopman linear model constructed in this invention exhibits excellent generalization performance, maintaining high prediction accuracy even in conditions not covered by the training samples. Experimental results show that the prediction error is less than 3% when the input variables change individually, and remains within 5% even when multiple variables change jointly. This characteristic enables this invention to effectively address the diversity and uncertainty of actual power grid fault conditions, overcoming the shortcomings of traditional data-driven methods that suffer a sharp drop in prediction accuracy under extrapolation scenarios.

[0043] 4. This invention provides a control strategy-independent universal modeling framework that can be adapted to various fault ride-through strategies such as maximum current control, constant active power control, and constant active current control simply by replacing the disturbance input parameters. This feature enables the invention to be widely applied to new energy units of different types and manufacturers, without the need to develop dedicated modeling methods for each control strategy, thus exhibiting good universality and scalability.

[0044] 5. This invention combines the current reference value predicted by the Koopman model with the current inner-loop time-domain model to generate a transient current response curve throughout the entire fault process, intuitively presenting dynamic characteristics such as the current overshoot amplitude, decay time, and steady-state convergence. Furthermore, by comparing the transient curve with grid connection standards, compliance testing of the unit's fault ride-through performance can be achieved; adjusting the PI controller parameters based on the deviation and performing iterative simulations enables rapid optimization of the control parameters. This closed-loop process provides a complete technical solution for the grid connection performance evaluation, parameter tuning, and optimization design of new energy units.

[0045] Further or more detailed beneficial effects will be described in conjunction with specific embodiments in the detailed implementation. Attached Figure Description

[0046] 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.

[0047] Figure 1 This is a flowchart illustrating the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention.

[0048] Figure 2 This is a schematic diagram of the experimental data acquisition process in the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention.

[0049] Figure 3This is a schematic diagram of the dimensional mapping and feature construction in the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention.

[0050] Figure 4 This is a schematic diagram of the Koopman operator modeling and solving process in the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention.

[0051] Figure 5 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. Active current during single-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0052] Figure 6 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. Reactive current during single-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0053] Figure 7 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. Active current during single-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0054] Figure 8 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. Reactive current during single-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0055] Figure 9 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. Active current during single-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0056] Figure 10 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. Reactive current during single-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0057] Figure 11 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. and Active current during two-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0058] Figure 12 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. and Reactive current during dual-variable extrapolation A schematic diagram showing the comparison between the model's predicted values ​​and the actual analytical values.

[0059] Figure 13 This is a schematic diagram of the prediction error distribution under individual variable changes in the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention.

[0060] Figure 14 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. and Simultaneously changing active current Schematic diagram of prediction error distribution.

[0061] Figure 15 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. and Simultaneous change in reactive current Schematic diagram of prediction error distribution.

[0062] Figure 16 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. A schematic diagram of the active current fault-crossing transient curve during changes.

[0063] Figure 17 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. A schematic diagram of the reactive current fault ride-through transient curve under changing conditions.

[0064] Figure 18 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. A schematic diagram of the active current fault-crossing transient curve during changes.

[0065] Figure 19 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. A schematic diagram of the reactive current fault ride-through transient curve under changing conditions.

[0066] Figure 20 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. A schematic diagram of the active current fault-crossing transient curve during changes.

[0067] Figure 21 This is the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention. A schematic diagram of the reactive current fault ride-through transient curve under changing conditions.

[0068] Figure 22 This is a schematic diagram of the replacement of disturbance quantities for different control strategies in the grid-side converter fault ride-through control structure modeling method provided in Embodiment 1 of the present invention.

[0069] Figure 23 This is a structural diagram of the electronic device provided in Embodiment 3 of the present invention.

[0070] Icon labels:

[0071] 2300. Electronic equipment;

[0072] 2301. Processor; 2302. Communication bus; 2303. User interface; 2304. Network interface; 2305. Memory. Detailed Implementation

[0073] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0074] In the following description, several embodiments of the present invention are provided. Different embodiments can be substituted or combined. Therefore, the present invention can also be considered to include all possible combinations of the same and / or different embodiments described. Thus, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present invention should also be considered to include embodiments containing one or more other possible combinations of A, B, C, and D, even if such embodiments are not explicitly described in the following text.

[0075] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made to the function and arrangement of the described elements without departing from the scope of the invention. Various processes or components may be appropriately omitted, substituted, or added to the various examples. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined into other examples.

[0076] To facilitate a better understanding of the embodiments of the present invention, its application scenarios will be explained before providing a detailed explanation of the specific implementation methods.

[0077] The grid-side converter fault ride-through control structure modeling method described in this specification is applied to the field of new energy power generation grid connection technology, specifically involving the grid-side converter control performance evaluation scenarios of new energy units such as wind power, photovoltaic power, and energy storage systems. In these scenarios, the application of the grid-side converter fault ride-through control structure modeling method aims to: without relying on the internal control logic of equipment manufacturers, and using only limited externally measurable data, quickly and accurately establish equivalent mathematical models of different fault ride-through control strategies, thereby enabling the prediction, evaluation, and optimization of the current response characteristics of the unit during grid faults. This provides technical support for the grid connection compliance detection, control parameter tuning, and operational performance improvement of new energy units.

[0078] The following is a brief explanation of the perturbation input data, current output data, up-dimensional mapping function, high-dimensional feature space, high-dimensional feature vector, ridge regression optimization, Koopman operator, global linear mapping model, and current inner loop time-domain model involved in several embodiments of this specification:

[0079] Disturbance input data refers to the excitation signals or control parameters applied externally to the grid-side converter control system when a grid fault occurs. Specifically, this includes voltage sag amplitude characterizing the severity of the grid fault, and control parameters reflecting the characteristics of the fault ride-through control strategy, such as maximum current limits, reactive power support coefficients, and active power setpoints. These data serve as input variables for modeling, driving the model to generate the corresponding current response output.

[0080] Current output data refers to the current command value output by the grid-side converter under a given disturbance input condition, based on its fault ride-through control strategy. Specifically, it includes active current reference components and reactive current reference components. These two components together determine the magnitude and phase of the current injected into the grid by the unit during a fault, and are the core indicators for evaluating the fault ride-through performance of the unit.

[0081] Dimensional Upgrading Mapping Function: A mathematical transformation function that transforms low-dimensional raw input data to a high-dimensional feature space. By introducing nonlinear transformations such as linear, quadratic, cross, trigonometric, and hyperbolic function terms, the original input variables are expanded into high-dimensional feature vectors. This allows the complex nonlinear relationships in the original space to appear as approximately linear relationships in the high-dimensional space, laying the foundation for subsequent linear operator modeling.

[0082] A high-dimensional feature space refers to the mathematical space in which the original input data resides after being transformed by an up-dimensional mapping function. The dimension of this space is much higher than that of the original input space. In the high-dimensional space, the originally complex nonlinear input-output relationship can be approximated as a linear relationship, thus enabling linear modeling methods (such as the Koopman operator) to be effectively applied to the modeling of nonlinear systems.

[0083] A high-dimensional feature vector is a vector obtained after transforming the original input data through a dimension-up mapping function; it is a point in a high-dimensional feature space. This vector contains the linear terms, nonlinear terms, and coupling terms between the original input variables, comprehensively characterizing the multi-dimensional features of the perturbed input, and serves as the input feature for modeling with the Koopman operator.

[0084] Ridge Regression Optimization: A linear regression method with L2 regularization. By introducing a regularization parameter into the objective function of minimizing prediction error, the magnitude of the model parameters is penalized, thereby effectively suppressing overfitting while ensuring model fitting accuracy and enhancing the model's generalization ability and extrapolation stability. This method is used to solve for the optimal estimate of the Koopman operator.

[0085] Koopman operator: an infinite-dimensional linear operator used to describe the linear evolution of nonlinear dynamic systems in an increased-dimensional space. In this method, a matrix form of the Koopman operator is obtained through finite-dimensional approximation. This operator establishes a linear mapping relationship between the high-dimensional feature space and the current output data, thereby achieving equivalent linear modeling of unknown nonlinear control strategies.

[0086] The global linear mapping model refers to an approximately linear mathematical model based on the Koopman operator, which represents the relationship between a disturbance input and a current output. This model describes the relationship between the input and output in the form of matrix operations. It can quickly and accurately predict the current reference value under any operating condition without knowing the internal logic of the control strategy, and it has good extrapolation performance.

[0087] The current inner loop time-domain model is a mathematical model describing the dynamic process of the current inner loop control of the grid-side converter. This model uses a proportional-integral controller as its core, combined with filter inductor parameters, and describes the tracking dynamic process between the current reference value and the actual current through differential equations. By inputting the current reference value predicted by the Koopman model into this model, the transient response curve of the current during a fault can be simulated, including dynamic characteristics such as the current overshoot amplitude, decay time, and steady-state error.

[0088] Example 1:

[0089] like Figure 1 As shown in the figure, this embodiment provides a method for modeling the fault ride-through control structure of a grid-side converter, including six steps: data acquisition, feature dimensionality enhancement, Koopman operator modeling, extrapolation prediction, inner-loop time-domain simulation, and compliance detection. The specific steps are as follows:

[0090] Step S1, Experimental Data Acquisition:

[0091] like Figure 2As shown, this embodiment first collects disturbance input and current output data under limited operating conditions. Traditional methods often require thousands or even tens of thousands of samples to cover various combinations of power grid disturbances and control strategy parameters, which is extremely costly in actual experiments and simulations. This embodiment, through the dimensionality-upgrading modeling capability of the Koopman operator, only requires tens to hundreds of samples to achieve prediction of the entire operating condition, significantly reducing the experimental burden.

[0092] This embodiment preferably uses a maximum current limiting control strategy, and the defined disturbance input variables include:

[0093] Voltage sag amplitude This indicates the percentage decrease in voltage relative to the rated value when a grid fault occurs. In the experiment, the range was set to [0.1pu, 0.3pu], which conforms to the fault voltage range requirements for low-voltage ride-through in GB / T 19963.1-2021 Technical Regulations for Wind Farm Access to Power Systems. Its engineering significance lies in reflecting the intensity of grid disturbances. The larger the voltage, the deeper the voltage drop, and the stronger the impact on the unit's current.

[0094] Maximum current limit This indicates the maximum allowable output current amplitude of the grid-side converter. In the experiment, the range was set to [1.0pu, 1.2pu] to reflect the capabilities under different capacities and protection strategies. Its engineering significance lies in constraining the current amplitude to prevent overcurrent damage.

[0095] reactive power support coefficient This represents the proportionality coefficient between voltage dips and reactive current injection, with a value range of [1.0, 2.0]. Its engineering significance lies in reflecting the unit's ability to support grid voltage during fault periods.

[0096] For different , , A total of two hundred sets of operating condition data were collected as training samples. The corresponding output current components satisfy:

[0097] (1),

[0098] In equation (1), This refers to the active current during fault ride-through. This refers to the reactive current during fault ride-through. The active current setpoint before the fault is given. This formula conforms to the current distribution principle under the power grid specification: when the voltage drops, the unit prioritizes providing reactive current support, while being limited by... .

[0099] Data can be obtained through RT-LAB hardware-in-the-loop simulation or a power electronics test bench. Hardware-in-the-loop simulation can realistically reproduce controller logic and power grid fault disturbances; the test bench is used to verify the characteristics of real devices. Compared with traditional methods, this embodiment requires significantly less data and shortens the experiment and simulation cycle.

[0100] Step S2, Dimensional Upgrading Map Construction:

[0101] like Figure 3 As shown, in order to make the nonlinear input-output relationship approximately linearized by the Koopman operator, it is necessary to increase the dimensionality of the original input variables.

[0102] The expression for the input vector is:

[0103] (2).

[0104] Define a dimension-up mapping function, whose expression is:

[0105] (3),

[0106] In equation (3), The first-order term is used to retain the original perturbation information. The quadratic term is used to capture nonlinear squared effects. Cross terms are used to characterize the coupling relationship between input variables. and For the combined term, the corresponding current limiting characteristic is... The nonlinear function term reflects nonlinear dynamics and saturation effects, and 1 is a constant term used for the bias in regression.

[0107] Through this dimensional mapping, the originally complex nonlinear relationship is... It is transformed into a linear approximation in a high-dimensional space, making subsequent modeling possible using matrix operations.

[0108] Step S3, Koopman operator modeling:

[0109] like Figure 4 As shown, the core of this invention is to use the Koopman operator to model the control structure.

[0110] Let the matrix formed by the input and output of the collected m sets of samples be:

[0111] (4),

[0112] (5),

[0113] In equations (4) and (5), for d ×m eigenmatrix d The feature dimension after dimensionality increase. Y 2× m The output matrix, corresponding to Id and Iq Components. The goal is to find the Koopman matrix. K (2×) d ), so that:

[0114] (6),

[0115] Right now:

[0116] (7),

[0117] In equation (7), This represents the square of the Frobenius norm of the matrix.

[0118] To avoid overfitting, an L2 regularization term is introduced into the objective function:

[0119] (8),

[0120] In equation (8), λ is the regularization coefficient, and its value range is

[10] . -4 10 -1 The range of values ​​has been experimentally verified to suppress overfitting while ensuring that the extrapolation error of the Koopman operator is less than 5%.

[0121] Expanding on this optimization problem:

[0122] (9),

[0123] right K Differentiate:

[0124] (10)

[0125] Set the derivative to zero:

[0126] (11),

[0127] Solving for:

[0128] (12)

[0129] Received K That is, the Koopman matrix, with a dimension of 2× d This describes the approximate linear mapping relationship between the dimensionality-upgrading characteristics of the input perturbation and the output current reference value.

[0130] The final modeling equation is:

[0131] (13)

[0132] In equation (13), This is the reference component of the active current during fault ride-through. This is the reference component of reactive current during fault ride-through. This is the mean bias term, which can be obtained by correcting the mean of the training data.

[0133] This process demonstrates that this embodiment can directly obtain the control structure model through mathematical optimization methods without relying on control logic, and has clear physical interpretability.

[0134] Step S4, Extrapolation Prediction and Error Analysis:

[0135] Using the Koopman matrix obtained during training K Predicting new disturbance inputs:

[0136] (14)

[0137] In equation (14), This refers to the predicted active current component during fault ride-through. This refers to the reactive current prediction component during fault ride-through. This is the input for a new disturbance.

[0138] Prediction results are as follows Figures 5-12 As shown, where, Figure 5 and Figure 6 As shown Extrapolation prediction results Figure 7 and Figure 8 As shown Extrapolation prediction results Figure 9 and Figure 10 As shown Extrapolation prediction results Figure 11 and Figure 12 As shown and Joint extrapolation prediction results.

[0139] To evaluate model accuracy, relative error is defined as follows:

[0140] (15)

[0141] In equation (15), and These are the actual values ​​of active current and reactive current, derived from analytical formulas or experimental measurements.

[0142] like Figure 13The figure shows the extrapolation prediction error results for each variable individually, such as... Figure 14 and Figure 15 As shown and Joint extrapolation prediction error surface. Results show that: when , , When changing alone, and The prediction error is less than 5%; when and Simultaneously, the error remains at a low level when changes occur; the surface is smooth and continuous, indicating that the model has stable extrapolation performance and can quickly meet the requirements for grid fault ride-through compliance testing.

[0143] Therefore, this embodiment not only has an error of less than 3% in univariate prediction, but also maintains high accuracy in joint extrapolation, making it suitable for compliance testing under all operating conditions.

[0144] Step S5, Inner Loop Time Domain Simulation:

[0145] like Figures 16-21 As shown, in this embodiment, the predicted reference current is input into the inner loop PI controller to obtain the current transient curve.

[0146] The dynamic equation of the controller is:

[0147] (16)

[0148] Its analytical solution is:

[0149] (17)

[0150] In equation (17), I ref This refers to the active or reactive current during fault ride-through. , and For PI parameters, For inductance, This is the initial current value.

[0151] Simulation results show that the current surges rapidly upon fault occurrence, then gradually decays to a steady state. This dynamic process conforms to the power grid standards' requirements for transient overcurrent and response time. Furthermore, by changing... and These parameters can optimize response speed and overshoot, further improving the grid connection performance of the unit.

[0152] Step S6, Judgment and Optimization:

[0153] The current transient response curve is compared with the grid fault ride-through standard limit. If the limit is exceeded, the proportional-integral controller parameters are adjusted and the process of generating the current transient response curve is repeated until the standard requirements are met.

[0154] In addition, such as Figure 22 As shown, the method in this embodiment has good versatility, does not depend on specific control strategy logic, and can be adapted to different strategies simply by replacing the disturbance input:

[0155] Under constant active power control, the disturbance input is , and P set Constant active power parameters; under constant active power current control, the disturbance input is... , and I d_set Constant current parameters; under maximum current control (in this embodiment), the disturbance input is: , and If the control strategy is unknown, it can be... , , P set and I d_set All are considered as disturbance inputs. If there are other control parameters, you can add them yourself.

[0156] The remaining steps (dimensionality mapping, Koopman modeling, and inner-loop simulation) remain consistent. Therefore, this embodiment provides a control strategy-independent general modeling framework that can be widely applied to fault ride-through performance modeling, prediction, and compliance testing of new energy units.

[0157] Example 2:

[0158] This embodiment provides a grid-side converter fault ride-through control structure modeling system to implement the grid-side converter fault ride-through control structure modeling method as described in Embodiment 1, including:

[0159] The data acquisition module is used to acquire disturbance input data and its corresponding current output data under limited operating conditions;

[0160] The dimension-up mapping module is used to construct a dimension-up mapping function to map the perturbation input data to a high-dimensional feature space to obtain a high-dimensional feature vector.

[0161] The Koopman modeling module is used to solve the Koopman operator based on the high-dimensional feature vector and the current output data through the ridge regression optimization framework, so as to establish a global linear mapping model from the high-dimensional feature space to the current output data.

[0162] The prediction module is used to input new disturbance input data into the global linear mapping model to obtain the predicted active current reference component and reactive current reference component.

[0163] The time-domain simulation module is used to input the predicted active current reference components and reactive current reference components into the current inner loop time-domain model, and solve the differential equations by combining the proportional-integral controller parameters and the filter inductor parameters to generate the current transient response curve during the fault ride-through process.

[0164] Example 3:

[0165] like Figure 23 As shown, this embodiment provides an electronic device, which may include: at least one processor, at least one network interface, a user interface, a memory, and at least one communication bus.

[0166] The communication bus can be used to enable communication between the various components mentioned above.

[0167] The user interface may include buttons, and optional user interfaces may also include standard wired interfaces and wireless interfaces.

[0168] The network interface may include, but is not limited to, Bluetooth modules, NFC modules, Wi-Fi modules, etc.

[0169] The processor may include one or more processing cores. It connects various parts of the electronic device via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in memory, and accessing data stored in memory to perform various functions and process data. Optionally, the processor can be implemented using at least one hardware form of DSP, FPGA, or PLA. The processor may integrate one or more of the following: CPU, GPU, and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also be implemented as a separate chip without being integrated into the processor.

[0170] The memory may include RAM or ROM. Optionally, the memory may include a non-transitory computer-readable medium. The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor. The memory, as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a modeling application. The processor can be used to call the modeling application stored in the memory and execute the steps of the grid-side converter fault ride-through control structure modeling method mentioned in the foregoing embodiments.

[0171] Example 4:

[0172] This embodiment provides a computer-readable storage medium storing instructions that, when executed on a computer or processor, cause the computer or processor to perform the above-described instructions. Figure 1 One or more steps in the illustrated embodiment. If the constituent modules of the above-described electronic device are implemented as software functional units and sold or used as independent products, they can be stored in the computer-readable storage medium.

[0173] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this specification are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0174] Those skilled in the art will understand that all or part of the processes in the method of Embodiment 1 described above can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks. Unless otherwise specified, the technical features of this embodiment and the implementation scheme can be combined arbitrarily.

[0175] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0176] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0177] The above description is merely an exemplary embodiment of the present invention and should not be construed as limiting the scope of the invention. Any equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of embodiments of the invention upon considering the specification and practicing the disclosure herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of the invention are defined by the claims.

Claims

1. A method for modeling grid side converter fault ride through control structure, characterized in that, Including the following steps: S1. Obtain disturbance input data and corresponding current output data under finite operating conditions; S2. Construct a dimension-upgrading mapping function to map the perturbation input data to a high-dimensional feature space to obtain a high-dimensional feature vector; S3. Based on the high-dimensional feature vector and the current output data, solve the Koopman operator through the ridge regression optimization framework to establish a global linear mapping model from the high-dimensional feature space to the current output data. S4. Input the new disturbance input data into the global linear mapping model to obtain the predicted active current reference component and reactive current reference component. S5. Input the predicted active current reference component and reactive current reference component into the current inner loop time domain model, and combine the proportional-integral controller parameters and filter inductor parameters to solve the differential equation and generate the current transient response curve during the fault ride-through process.

2. The method for modeling a fault ride-through control structure for a grid-side converter according to claim 1, characterized in that: The disturbance input data includes voltage drop amplitude characterizing the degree of grid fault, and characteristic parameters characterizing the fault ride-through control strategy. The current output data includes active current reference components and reactive current reference components.

3. The method of claim 2, wherein, The characteristic parameters representing the fault ride-through control strategy are determined according to the control strategy type, specifically: When the control strategy is maximum current control, the characteristic parameters include the maximum current limit and the reactive power support coefficient. When the control strategy is constant active power control, the characteristic parameters include the reactive power support coefficient and the constant active power setpoint. When the control strategy is constant active current control, the characteristic parameters include the reactive power support coefficient and the constant active current setpoint.

4. The method for modeling the fault ride-through control structure of a grid-side converter according to claim 1, characterized in that: The updimensional mapping function transforms the nonlinear input-output relationship in the original space into a linear relationship in the higher-dimensional space, including linear terms, quadratic terms, cross terms, combination terms, nonlinear function terms, and constant terms.

5. The method of claim 1, wherein, The objective function of the ridge regression optimization framework is: ， wherein is the current output data matrix, is the Koopman operator matrix to be solved, is the high-dimensional feature matrix, is the regularization parameter, is the square of the Frobenius norm of the matrix.

6. The method for modeling the fault ride-through control structure of a grid-side converter according to claim 5, characterized in that: The regularization parameter The value range is [10]. -4 10 -1 ].

7. The method for modeling the fault ride-through control structure of a grid-side converter according to claim 1, characterized in that, Also includes: The current transient response curve is compared with the grid fault ride-through standard limit. If the limit is exceeded, the proportional-integral controller parameters are adjusted and the process of generating the current transient response curve is repeated until the standard requirements are met.

8. A fault ride-through control structure modeling system for grid-side converters, characterized in that, The method for modeling the fault ride-through control structure of a grid-side converter as described in any one of claims 1 to 7 includes: The data acquisition module is used to acquire disturbance input data and its corresponding current output data under limited operating conditions; The dimension-up mapping module is used to construct a dimension-up mapping function to map the perturbation input data to a high-dimensional feature space to obtain a high-dimensional feature vector. The Koopman modeling module is used to solve the Koopman operator through the ridge regression optimization framework based on the high-dimensional feature vector and the current output data, so as to establish a global linear mapping model from the high-dimensional feature space to the current output data. The prediction module is used to input new disturbance input data into the global linear mapping model to obtain the predicted active current reference component and reactive current reference component. The time-domain simulation module is used to input the predicted active current reference components and reactive current reference components into the current inner loop time-domain model, and solve the differential equations by combining the proportional-integral controller parameters and the filter inductor parameters to generate the current transient response curve during the fault ride-through process.

9. A computer device, the computer device comprising a memory, a processor, and a computer program, characterized in that, When the computer program is executed by the processor, it implements the grid-side converter fault ride-through control structure modeling method as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the grid-side converter fault ride-through control structure modeling method as described in any one of claims 1 to 7.

Images