Equivalent modeling method and system for network construction type cluster converter and storage medium

Abstract

The invention provides a network construction type cluster converter equivalent modeling method and system and a storage medium, and the method comprises the steps: applying preset standard disturbance to a network construction type cluster converter, and extracting feature points of a power angle dynamic response curve of each network type converter in the cluster converter during the standard disturbance; taking the feature points of the power angle dynamic response curve of each network-type converter in each network-type converter as feature vectors, taking feature point vectors as clustering indexes, and dividing the converters with dynamic characteristics of the feature vectors into coherent groups by adopting a clustering algorithm; and aggregating the parameters of the converters in the coherence groups, aggregating each coherence group into an equivalent converter, and forming a multi-machine equivalent model.

Description

Technical Field

[0001] This application relates to the energy field, and in particular to an equivalent modeling method, system, and storage medium for a grid-type cluster converter. Background Technology

[0002] New and renewable sources of energy (NRSE) are energy sources that are not traditional energy sources, such as coal, oil, and natural gas. They are also known as unconventional energy sources and refer to various forms of energy other than traditional energy sources. NRSEs are those that are just beginning to be developed or are under active research and development, awaiting wider adoption. They include solar energy, geothermal energy, wind energy, ocean energy, biomass energy, and nuclear fusion energy.

[0003] With the accelerated transformation of the energy structure, the penetration rate of new energy power generation in the power system continues to increase. However, the high proportion of power electronic equipment connected to the grid leads to a decrease in system inertia, weakening strength, and reduced recovery capability. Unlike traditional grid-following control, grid-forming control strategies (such as virtual synchronous generators, VSGs) can enable converters to simulate synchronous machine characteristics, providing voltage and frequency support to the grid. However, the complex interaction between multiple VSG converters in a converter cluster and the grid results in a significant increase in the computational load of electromagnetic transient simulation, necessitating efficient equivalent modeling methods. Existing equivalent methods are mostly designed for grid-following converters, but the synchronous machine characteristics of grid-forming VSG converters differ fundamentally from traditional grid-following control, causing existing methods to fail when applied to grid-forming converters. Summary of the Invention

[0004] This invention provides an equivalent modeling method and system for grid-type converter clusters, which can achieve the simulation efficiency advantage of hundreds of converter clusters and support real-time analysis of large-scale new energy power grids.

[0005] In a first aspect, embodiments of the present invention provide an equivalent modeling method for a grid-type cluster converter, the method comprising the following steps: A preset standard disturbance is applied to the grid-type cluster converter, and the characteristic points of the dynamic response curves of the power angle of each grid converter in the cluster converter are extracted under the standard disturbance. The characteristic points of the dynamic response curve of the power angle of each grid converter in each grid converter are used as feature vectors. The feature vectors are used as clustering indexes. The clustering algorithm is used to divide the converters with similar dynamic characteristics of the feature vectors into homogeneous groups. The parameters of each converter within the homogeneous group are aggregated, and each homogeneous group is aggregated into an equivalent converter to form a multi-machine equivalent model.

[0006] Secondly, a network-type cluster converter equivalent modeling system is provided, the system comprising: The disturbance extraction unit is used to apply a preset standard disturbance to the grid-type cluster converter and extract the feature points of the dynamic response curve of the power angle of each grid converter in the cluster converter under the standard disturbance. The clustering unit is used to take the feature points of the dynamic response curve of the power angle of each grid converter in each grid converter as feature vectors, use the feature point vectors as clustering indexes, and use a clustering algorithm to divide the converters with similar dynamic characteristics of the feature vectors into homogeneous groups. The aggregation construction unit is used to aggregate the parameters of each converter in the homogeneous group, and aggregate each homogeneous group into an equivalent converter to form a multi-machine equivalent model.

[0007] Thirdly, a computer-readable storage medium is provided that stores a program for electronic data interchange, wherein the program causes a terminal to execute the method provided in the first aspect.

[0008] Implementing the embodiments of the present invention has the following beneficial effects: This application applies a preset standard perturbation to a grid-type power converter cluster, extracting feature points from the dynamic response curves of the power angle of each grid-type power converter in the cluster under the standard perturbation. The feature points of the dynamic response curves of each grid-type power converter are used as feature vectors, and the feature vectors are used as clustering indices. A clustering algorithm is employed to divide power converters with similar dynamic characteristics into coherent groups. The parameters of each power converter within the coherent group are aggregated, and each coherent group is aggregated into an equivalent power converter, forming a multi-machine equivalent model. The technical solution of this application can preserve the dynamic differences within the cluster in a multi-machine equivalent architecture, significantly reducing simulation time compared to detailed models, demonstrating simulation efficiency advantages for clusters of hundreds of power converters, and supporting real-time analysis of large-scale new energy power grids. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 A schematic diagram of a typical converter cluster structure provided in this application for an equivalent modeling method of a network-type converter cluster. Figure 2 A schematic diagram of a virtual synchronous machine control system structure provided for an equivalent modeling method of a network-type cluster converter in this application; Figure 3A schematic diagram of a two-stage converter unit model for an equivalent modeling method of a network-type cluster converter provided in this application; Figure 4 A flowchart illustrating an equivalent modeling method for a network-type cluster converter provided in this application; Figure 5 A schematic diagram of the dynamic characteristic curve of the converter power angle of a network-type cluster converter provided in this application; Figure 6 A schematic diagram of the FCM clustering results for an equivalent modeling method for a network-type cluster converter provided in this application; Figure 7 A schematic diagram of an equivalent converter for a network-type cluster converter equivalent modeling method provided in this application; Figure 8 Active power comparison waveform diagram of an equivalent modeling method for a grid-type cluster converter provided in a specific embodiment of this application; Figure 9 Reactive power comparison waveform diagram of an equivalent modeling method for a grid-type cluster converter provided in this application; Figure 10 This is a structural schematic diagram of an equivalent modeling system for a network-type cluster converter provided in this application. Detailed Implementation

[0011] 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, not all, of the embodiments of the present invention. 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.

[0012] The terms "first," "second," "third," and "fourth," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0013] In this document, the term "embodiment" means that a particular feature, result, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0014] The following section introduces the concepts involved in this application.

[0015] Grid-based technology is a control technology that simulates the characteristics of traditional synchronous generators, enabling new energy equipment (such as wind turbines and energy storage systems) to autonomously build and stabilize the power grid. It is suitable for scenarios with a high proportion of new energy sources and poor grid stability.

[0016] Equivalent modeling is a technique used in power system analysis to simplify complex system models by preserving key dynamic characteristics and reducing computational complexity.

[0017] A converter is a power electronic device used to convert the voltage, frequency, number of phases, and other characteristics of electrical energy. Its core functions include rectification (AC to DC), inversion (DC to AC), and AC frequency conversion.

[0018] MATLAB is a commercial mathematical software that uses matrix operations as its core and integrates numerical computation, data analysis, visualization, and simulation modeling.

[0019] Simulink is an integrated visualization and simulation tool for MATLAB, primarily used for dynamic system modeling, simulation, automatic code generation, and embedded system verification.

[0020] RT-LAB is an industrial-grade real-time simulation platform, primarily used for real-time simulation and testing of control and power systems.

[0021] See Figure 1 , Figure 1 For a typical converter cluster structure provided in this application for an equivalent modeling method of a grid-type converter cluster, please refer to [reference needed]. Figure 1 , Figure 1 Multiple new energy power generation devices (such as photovoltaics) are connected to digital-to-analog converters (DC / AC) and then connected to the load.

[0022] Figure 2 This application provides a virtual synchronous machine control system structure based on an equivalent modeling method for a network-type cluster converter. (See also...) Figure 2 The angle information is processed through a series of steps to output the angle value θ, and the voltage information is processed to synthesize the voltage information. The above processing scheme is the existing processing method, which will not be elaborated here.

[0023] See Figure 2 The virtual synchronous generator control system structure, This refers to the three-phase stator voltage. It is a three-phase excitation electromotive force. For synchronization reactance, For three-phase stator current, For stator winding resistance, For mechanical torque, For electromagnetic torque, The moment of inertia of the rotor. For the mechanical angular velocity of the generator, The damping coefficient is... This refers to the synchronous angular velocity.

[0024] The equations for an ideal synchronous generator model are:

[0025] Figure 3 A schematic diagram of a two-stage converter unit model for an equivalent modeling method for a network-type cluster converter provided in this application is shown in the attached diagram. Figure 3 The new energy power generation equipment outputs power to the load after rectification and conversion.

[0026] See Figure 3 , Figure 3 The bipolar converter unit model includes: DC / DC converter and its control module, and DC / AC converter and its control module, etc.

[0027] The practical criterion of homology equivalence theory states that, within a given time period... Inside, the initial values ​​of the two generators relative to their respective power angles are as follows: and If using and The maximum absolute value of the difference between the two generators is used to reflect the degree of synchronization between them. When the calculated result is less than a given value ε, the synchronization is satisfied. These generators are said to be in sync. Furthermore, it is assumed that generators within a sync group can be simplified to an equivalent generator unit.

[0028] Based on the synchronous generator characteristics of the cluster converter employing a virtual synchronous generator control strategy, and drawing on the synchronization determination method for synchronous generators in AC systems, the synchronization of the cluster converter is judged by observing the power angle characteristics of each converter over a certain period. The specific process is essentially the same as the synchronization equivalence steps in AC systems. However, considering the inherent differences between power electronic converters and synchronous generators, certain corrections are necessary during the equivalence process.

[0029] See Figure 4 , Figure 4 This is a flowchart illustrating an equivalent modeling method for a network-type cluster converter provided in this application. The method can be used in applications such as... Figures 1-3 Implemented in the technical scenarios shown, such as Figure 4 As shown, the above method includes the following steps: Step S401: Apply a preset standard disturbance to the grid-type cluster converter and extract the characteristic points of the dynamic response curve of the power angle of each grid-type converter in the cluster converter under the standard disturbance. For standard disturbances, a clear disturbance reporting system and recovery evaluation framework have been established. For example, determining the peak difference between the start time and elimination time of the disturbance allows dispatchers to accurately grasp the grid status after the disturbance and rationally adjust dispatch strategies. Simultaneously, this standard can promote the introduction of automated control measures in grid dispatching, enabling the system to automatically initiate adjustment processes after a disturbance occurs, reducing the manual operation burden on dispatchers, and improving the operation mode and control efficiency of grid dispatching. Therefore, from a modeling perspective, the performance of standard disturbances simulates the actual disturbance situation in grid operation to a certain extent, thereby improving the simulation effect.

[0030] Step S402: Take the feature points of the dynamic response curve of the power angle of each grid converter in each grid converter as feature vectors, use the feature point vectors as clustering index, and use a clustering algorithm to divide the converters with similar dynamic characteristics of the feature vectors into homogeneous groups. The converter power angle dynamic response curve describes the characteristic curve of the converter's power angle changing over time when subjected to grid disturbances (such as voltage dips, load surges, etc.). It can intuitively reflect the converter's stability and dynamic regulation capability during transient processes. The technical solution of this application innovatively uses the difference between the first peak swing at the occurrence and elimination of the standard disturbance as the characteristic vector, and constructs the characteristic curve based on the characteristic vector. This makes the characteristic curve more closely match the converter's stability and regulation capability during disturbances.

[0031] Step S403: Aggregate the parameters of each converter in the harmonic group, and aggregate each harmonic group into an equivalent converter to form a multi-machine equivalent model.

[0032] In one optional technical scenario, the aforementioned preset standard disturbance specifically involves setting a three-phase ground fault at the converter's common connection point and eliminating it after a preset time.

[0033] In one optional technical scenario, the extraction of feature points from the dynamic response curves of the power angle of each grid converter in the cluster converter under the standard disturbance specifically includes: When the standard disturbance occurs, the first swing peak difference δ1-δ0 of the converter is extracted as a feature point; When the standard disturbance is eliminated, the first swing peak difference δ1-δ2 of the converter is extracted as another feature point.

[0034] The technical solution of this application extracts the peak difference of the first swing amplitude during the disturbance period. Difference between peak amplitude and first swing amplitude during recovery period It constitutes a dual feature point criterion, which fully covers the entire dynamic response process and solves the deficiency of insufficient representation of single working conditions; the transient peak error of the multi-machine equivalent model is ≤5%, which is more than 60% lower than the error of the single-machine equivalent model, thereby reducing the error of the model.

[0035] In one optional technical scenario, the above-mentioned use of a clustering algorithm to divide the dynamic characteristics of the feature vector into homogeneous groups specifically includes: Fuzzy C-means clustering algorithm is used to divide converters with similar dynamic characteristics into homogeneous groups.

[0036] In one alternative technical scenario, the above-mentioned use of fuzzy C-means clustering algorithm to divide converters with similar dynamic characteristics into homogeneous groups specifically includes: Extract the clustering indices of each converter in the detailed model and select the initial cluster centers; Update the membership matrix using the cluster centers according to Formula 1; Formula 1 in As a clustering metric, As the cluster index center point, Let Euclidean distance be the Euclidean distance between the i-th cluster center and the j-th converter cluster index. For fuzzy coefficients; This application utilizes the advantages of membership matrix and combines it with fuzzy clustering. It uses the Euclidean distance between the i-th cluster center and the j-th converter cluster index, and the Euclidean distance between the k-th cluster center and the j-th converter cluster index. The ratio of these two types of Euclidean distances is used to obtain the degree of belonging between the two types of Euclidean distances.

[0037] The cluster centers are updated using the membership matrix according to Formula 2; Formula 2 in, Let be the membership value of the j-th converter to the group corresponding to the i-th cluster center. Repeat the membership matrix update operation and cluster center update operation until the change in the FCM objective function between two consecutive iterations is less than the allowable value, at which point the iteration ends.

[0038] See Figure 6 , Figure 6A schematic diagram of the FCM clustering results of the equivalent modeling method for the network-type cluster converter in the technical solution provided in this application.

[0039] See Figure 7 , Figure 7 A schematic diagram of the converter's power angle dynamic characteristic curve provided in this application. The power angle is maintained during stable operation of the converter. Basically unchanged, At time 1, the system experiences a disturbance that persists until time 2. During this process, the converter power angle jumps to a maximum of It then enters an oscillating state and gradually tends towards a new stable value. ; At that moment, the system returned to normal from the disturbance state, and the converter power angle first jumped to... It then enters an oscillating state and gradually tends towards a new stable value. .parameter This represents the first peak value of the converter's power angle after a system fault occurs. (Parameter) The first peak value of the converter's power angle after the system returns to normal represents the first peak value of the power angle. The former reflects the dynamic response characteristics of the converter's power angle under disturbance conditions, while the latter reflects the dynamic response characteristics of the converter's power angle under normal operating conditions. By extracting these two parameters for each converter, the feature point vector corresponding to each converter can be constructed. , [ ] as its clustering index.

[0040] If the distance between two clustering indices is small, it indicates that the strain gauges they correspond to have relatively consistent dynamic characteristics. By applying a clustering algorithm, they should be classified into the same homology group; otherwise, they should be classified into different homology groups.

[0041] Based on the above-mentioned main ideas of homology equivalence, homology equivalence criteria and FCM algorithm principle, a clustering method for homology converters based on the FCM algorithm is summarized, and its specific steps are as follows. S1: Extract the clustering index of each converter in the detailed model. S2: Select initial cluster centers. S3: Update the membership matrix from the cluster centers according to the following formula.

[0042] in As a clustering metric, As the cluster index center point, Let Euclidean distance be the Euclidean distance between the i-th cluster center and the j-th converter cluster index. is the fuzzy coefficient.

[0043] S4: Update the cluster centers using the membership matrix according to the following formula.

[0044]

[0045] This represents the membership value of the j-th converter to the group corresponding to the i-th cluster center. S5: Repeat S3 and S4 until the change in the FCM objective function between the two iterations is less than the allowable value, then the iteration ends.

[0046] Using the FCM algorithm for, for example Figure 1 The converter cluster shown can be clustered to obtain c cluster centers and one cluster center. The fuzzy clustering matrices are used to describe the average characteristics of the converter clustering index in the corresponding group and the membership degree of each converter to each group.

[0047] Combining the parameter aggregation method in the synchronous generator group coherence equivalent theory of AC systems, it is assumed that the control parameters and circuit parameters in the detailed model can be aggregated independently.

[0048] Compared to the virtual rotor motion equation of the converter based on the virtual synchronous generator control strategy, the control parameter aggregation method in the converter control system, using the coherence equivalent theory in AC systems, is shown in the following equation:

[0049]

[0050] J represents the moment of inertia of the converter.

[0051] According to the equivalent constraint requirements, the active power and reactive power of the equivalent model are the same as those of the detailed model. In the control system, this is reflected as the aggregation of active power command values ​​and reactive power command values, as shown in the following formula:

[0052]

[0053] in, This is the active power command value for the equivalent converter; J is the active power command value of the j-th equivalent converter; , , Similarly to K, This is the reactive power command value for the equivalent converter; This is the reactive power command value for the j-th equivalent converter.

[0054] By adopting the parameter aggregation method for generator excitation system in the homology equivalent theory, a method for aggregating reactive voltage loop control parameters of virtual synchronous generators is derived.

[0055]

[0056]

[0057] The D above represents reactive power.

[0058] The circuit structure of n equivalent converters connected to the same busbar and Figure 1 Similarly, the difference lies in that each converter is an equivalent converter obtained by aggregating the original converters. The aggregation of the converter LC filter and output line parameters can be regarded as the parallel equivalent of the circuit elements, as shown in the following formula:

[0059]

[0060]

[0061] The parallel equivalent of the various circuit components provided in this application can simplify the circuit structure. For example, the parallel equivalent of resistors, capacitors and inductors can simplify multiple circuit devices. However, since the circuit is equivalent, it is consistent with setting them separately. Therefore, it can simplify the circuit structure, thereby reducing the number of parameters of each device, reducing the amount of analysis data, and improving the efficiency of model analysis data.

[0062] The equivalent model established using the above equivalent modeling method is basically consistent with the external characteristics of the power grid at the PCC node of the detailed model.

[0063] This application employs the Fuzzy C-means Clustering (FCM) algorithm to handle the nonlinear boundaries of the converter's dynamic response, and uses membership metrics to quantify cluster correlations; even in scenarios with discrete clustering indices, it can still accurately classify homogeneous groups. Control parameters are summed and aggregated according to synchronous machine characteristics, while circuit parameters are equivalent to electrical parallel topology; these two are processed independently, avoiding characteristic distortion caused by the mixed aggregation in existing technologies. The multi-machine equivalent architecture preserves the dynamic differences within the cluster, significantly reducing simulation time compared to detailed models. The simulation efficiency advantage of hundreds of converter clusters is further expanded, supporting real-time analysis of large-scale new energy power grids. It adapts to virtual synchronous generator control strategies, solving the problem of transplantation failure of traditional grid-connected converter equivalent methods; and it is compatible with electromagnetic transient simulation (Matlab / Simulink) and hardware-in-the-loop testing (RT-LAB).

[0064] To verify the feasibility and accuracy of the proposed equivalent modeling method for converters based on a grid-connected control strategy, a multi-converter parallel simulation platform based on the grid-connected control strategy was built in the Matlab / SIMULINK and RT-LAB real-time simulation environments. The simulation platform includes five grid-connected converters with different parameters, controlled by virtual synchronous generators, which operate in parallel on the same AC bus. During the modeling process, the converter-side region is simplified by equivalent modeling, while the grid-side region remains unchanged.

[0065] In the numerical analysis, clustering indices are extracted from the power angle curves of each converter to determine the homogeneity between them and divide them into cluster groups. This process then aggregates the parameters to obtain a multi-converter equivalent model. Finally, simulated disturbances are applied to the grid-connected system, and the dynamic response curves of the detailed model and the equivalent model are compared and analyzed to calculate the efficiency and accuracy of the equivalent model.

[0066] A multi-converter parallel simulation platform based on a grid-type control strategy was built. The grid rated voltage is 380V, and the parameters of each converter are shown in Table 1: Table 1:

[0067] During the simulation, all five converters entered a stable operating state. A three-phase ground fault was set at the PCC point, and the fault was subsequently cleared. During this process, due to changes in the system's operating state, the power angle curves of each converter will experience some fluctuations.

[0068] Figure 5 The diagram shows the dynamic characteristic curves of the converter power angle in the equivalent modeling method for a grid-type cluster converter provided in this application. As can be seen from Figure 5, the power angle curves of each converter change. When the converter is running stably, the power angle remains constant. Basically unchanged, When a disturbance occurs in the system, the converter power angle jumps to its highest value. Then it enters an oscillating state and gradually stabilizes. ; When the system returned to normal, the converter power angle first jumped to... Then it enters an oscillation state. During the construction of the clustering index, the primary focus is on the first peak value of the converter's power angle after a system fault occurs. And the first peak value of the converter power angle after the system returns to normal. By extracting the feature point sets of each converter separately, the vector corresponding to each converter can be constructed. The clustering index was used as the clustering index and normalized. The specific results are shown in Table 2: Table 2:

[0069] The clustering index vectors of each converter in Table 2 are plotted in a two-dimensional coordinate system. The cluster center and membership matrix are obtained iteratively by applying the FCM clustering algorithm. During the iteration process, at the 11th iteration, the difference between the objective function value and the objective function value at the 10th iteration begins to be less than the allowable value for the first time, and the iteration ends.

[0070] According to the parameter aggregation method proposed in this application, CON1, CON2, and CON3 are equivalent to a single converter, named CON_A, and CON4 and CON5 are equivalent to a single converter, named CON_B. The parameters of CON_A and CON_B obtained after parameter aggregation and equivalence are shown in Table 3: Table 3:

[0071] The results of the independent isometry analysis of homology group A and homology group B are shown in Table 4: Table 4:

[0072] During the simulation, the same fault disturbance as described above was set, and the dynamic response curves of active and reactive power at the boundary between the converter-side region and the grid-side region before and after the equivalent values ​​were compared. The results are as follows: Figure 8 and Figure 9 As shown. Figure 8 Active power comparison waveform diagram of an equivalent modeling method for a grid-type cluster converter provided in a specific embodiment of this application; Figure 9 The reactive power comparison waveform diagram is provided for the equivalent modeling method of a grid-type cluster converter provided in this application.

[0073] As can be seen from the above analysis, the dynamic response characteristics of both the active and reactive power in the equivalent model are highly similar to those in the detailed model. Compared with the detailed model, the equivalent error is very small, which is within the acceptable error range for power grid analysis.

[0074] This application uses the simulation time required for detailed models and equivalent models to reflect its analytical efficiency, as shown in Table 5: Table 5:

[0075] Compare the dynamic response curves of the equivalent model and the detailed model; it is necessary to meet the peak error of the transient process calculation, requiring active power error ≤5%, reactive power error ≤5%, output voltage error ≤5%, and output current error ≤5%.

[0076] In one optional technical scenario, the aggregation of parameters of each converter within the harmonic group specifically includes: Extract the control parameters and circuit parameters of each converter in the same tuning group; The control parameters are aggregated using a summation aggregation method; The circuit parameters are aggregated using a parallel equivalent aggregation method.

[0077] In one alternative technical scenario, the above method further includes: A detailed model of the network-type cluster converter is constructed, and the dynamic response curves of the equivalent model and the detailed model are compared and calculated.

[0078] In one optional technical scenario, the comparison metrics for the aforementioned dynamic response curves include: The transient trajectories of active power, reactive power, output voltage, and output current during the standard disturbance.

[0079] See Figure 10 , Figure 10 This application provides a structural schematic diagram of an equivalent modeling system for a network-type cluster converter, the system comprising: The disturbance extraction unit 1001 is used to apply a preset standard disturbance to the grid-type cluster converter and extract the feature points of the dynamic response curve of the power angle of each grid converter in the cluster converter under the standard disturbance. Clustering unit 1002 is used to take the feature points of the dynamic response curve of the power angle of each grid converter in each grid converter as feature vectors, use the feature point vectors as clustering indexes, and use a clustering algorithm to divide the converters with similar dynamic characteristics of the feature vectors into homogeneous groups. The aggregation construction unit 1003 is used to aggregate the parameters of each converter in the homogeneous group, and aggregate each homogeneous group into an equivalent converter to form a multi-machine equivalent model.

[0080] Optionally, the preset standard disturbance specifically involves setting a three-phase ground fault at the converter's common connection point and eliminating it after a preset time.

[0081] Optionally, the disturbance extraction unit 1001 is specifically used to extract the first swing peak difference δ1-δ0 of the converter as a feature point when the standard disturbance occurs; When the standard disturbance is eliminated, the first swing peak difference δ1-δ2 of the converter is extracted as another feature point.

[0082] Optionally, clustering unit 1002 is specifically used to divide converters with similar dynamic characteristics into homogeneous groups using a fuzzy C-means clustering algorithm.

[0083] Optionally, clustering unit 1002, specifically used for dividing converters with similar dynamic characteristics into homogeneous groups using the fuzzy C-means clustering algorithm, specifically includes: Extract the clustering indices of each converter in the detailed model and select the initial cluster centers; Update the membership matrix using the cluster centers according to Formula 1; Formula 1 in As a clustering metric, As the cluster index center point, Let Euclidean distance be the Euclidean distance between the i-th cluster center and the j-th converter cluster index. For fuzzy coefficients; The cluster centers are updated using the membership matrix according to the following formula;

[0084] in, Let be the membership value of the j-th converter to the group corresponding to the i-th cluster center. Repeat the membership matrix update operation and cluster center update operation until the change in the FCM objective function between two consecutive iterations is less than the allowable value, at which point the iteration ends.

[0085] The aggregation construction unit 1003 is specifically used to extract the control parameters and circuit parameters of each converter in the harmonic group; The control parameters are aggregated using a summation aggregation method; The circuit parameters are aggregated using a parallel equivalent aggregation method.

[0086] Optionally, the aggregation building unit 1003 is also used to build a detailed model of the network-type cluster converter, and to compare and calculate the dynamic response curves of the equivalent model and the detailed model.

[0087] Optionally, the comparison metrics for the dynamic response curves include: The transient trajectories of active power, reactive power, output voltage, and output current during the disturbance.

[0088] This application provides a computer-readable storage medium storing a computer program for electronic data interchange. The computer program includes execution instructions for performing some or all of the steps of any of the equivalent modeling methods for grid-type cluster converters described in the above model training method embodiments. The computer includes an electronic client device.

[0089] This application provides a computer program product, which includes a computer program operable to enable a computer to perform some or all of the steps of any model training method described in the above method embodiments. The computer program product may be a software installation package.

[0090] It should be noted that, for the sake of simplicity, all embodiments of the aforementioned equivalent modeling methods for grid-type power converters are described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions involved are not necessarily essential to this application.

[0091] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of a model training method, computer, storage medium, and program product of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and its core ideas of this application. At the same time, for those skilled in the art, based on the ideas of a model training method, computer, storage medium, and program product of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

[0092] This application is described with reference to flowchart illustrations and / or block diagrams of methods, hardware products, and computer program products according to embodiments of this application. It should be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable model training device to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable model training device, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0093] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable model training device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The functions specified in one or more boxes. Memory may include: flash drives, read-only memory (ROM), random access memory (RAM), hard disks or optical disks, etc.

[0094] Although this application has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the accompanying drawings, disclosure, and appended claims, will understand and implement other variations of the disclosed embodiments in carrying out the claimed application. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude multiple instances. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce a good effect.

[0095] Those skilled in the art will understand that all or part of the steps in the various method embodiments of any of the above-described model training methods can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage device, which may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a disk, or an optical disk, etc.

[0096] It is understood that any product that is controlled or configured to execute the processing method of the flowchart described in an embodiment of a model training method of this application, such as the apparatus and computer program product of the above flowchart, falls within the scope of the related products described in this application.

[0097] Obviously, those skilled in the art can make various modifications and variations to the model training method, computer, storage medium, and program product provided in this application without departing from the spirit and scope of this application. Therefore, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application also intends to include these modifications and variations.

Claims

1. A method for equivalent modeling of a network-type cluster converter, characterized in that, The method includes the following steps: A preset standard disturbance is applied to the grid-type cluster converter, and the characteristic points of the dynamic response curves of the power angle of each grid converter in the cluster converter are extracted under the standard disturbance. The characteristic points of the dynamic response curve of the power angle of each grid converter in each grid converter are used as feature vectors. The feature vectors are used as clustering indexes. The clustering algorithm is used to divide the converters with similar dynamic characteristics of the feature vectors into homogeneous groups. The parameters of each converter within the homogeneous group are aggregated, and each homogeneous group is aggregated into an equivalent converter to form a multi-machine equivalent model.

2. The equivalent modeling method for network-type cluster converters according to claim 1, characterized in that, The preset standard disturbance is specifically: a three-phase ground fault is set at the common connection point of the converter and eliminated after a preset time.

3. The equivalent modeling method for network-type cluster converters according to claim 2, characterized in that, The extraction of feature points from the dynamic response curves of the power angle of each grid converter in the cluster converter under the standard disturbance specifically includes: When the standard disturbance occurs, the first swing peak difference δ1-δ0 of the converter is extracted as a feature point; When the standard disturbance is eliminated, the first swing peak difference δ1-δ2 of the converter is extracted as another feature point.

4. The equivalent modeling method for network-type cluster converters according to claim 1, characterized in that, The step of using a clustering algorithm to divide the dynamic characteristics of the feature vectors into homogeneous groups specifically includes: Fuzzy C-means clustering algorithm is used to divide converters with similar dynamic characteristics into homogeneous groups.

5. The equivalent modeling method for network-type cluster converters according to claim 4, characterized in that, The method of using fuzzy C-means clustering algorithm to divide converters with similar dynamic characteristics into homogeneous groups specifically includes: Extract the clustering indices of each converter in the detailed model and select the initial cluster centers; Update the membership matrix using the cluster centers according to Formula 1; Official 1 in As a clustering metric, As the cluster index center point, Let Euclidean distance be the Euclidean distance between the i-th cluster center and the j-th converter cluster index. For fuzzy coefficients; The cluster centers are updated using the membership matrix according to Formula 2; Official 2 in, Let be the membership value of the j-th converter to the group corresponding to the i-th cluster center. Repeat the membership matrix update operation and cluster center update operation until the change in the FCM objective function between two consecutive iterations is less than the allowable value, at which point the iteration ends.

6. The equivalent modeling method for network-type cluster converters according to claim 1, characterized in that, The aggregation of parameters of each converter within the harmonic group specifically includes: Extract the control parameters and circuit parameters of each converter in the same tuning group; The control parameters are aggregated using a summation aggregation method; The circuit parameters are aggregated using a parallel equivalent aggregation method.

7. The equivalent modeling method for network-type cluster converters according to claim 1, characterized in that, The method further includes: A detailed model of the network-type cluster converter is constructed, and the dynamic response curves of the equivalent model and the detailed model are compared and calculated.

8. The equivalent modeling method for network-type cluster converters according to claim 7, characterized in that, The comparison metrics for the dynamic response curves include: The transient trajectories of active power, reactive power, output voltage, and output current during the standard disturbance are shown.

9. An equivalent modeling system for a network-type cluster converter, characterized in that, The system includes: The disturbance extraction unit is used to apply a preset standard disturbance to the grid-type cluster converter and extract the feature points of the dynamic response curve of the power angle of each grid converter in the cluster converter under the standard disturbance. The clustering unit is used to take the feature points of the dynamic response curve of the power angle of each grid converter in each grid converter as feature vectors, use the feature point vectors as clustering indexes, and use a clustering algorithm to divide the converters with similar dynamic characteristics of the feature vectors into homogeneous groups. The aggregation construction unit is used to aggregate the parameters of each converter in the homogeneous group, and aggregate each homogeneous group into an equivalent converter to form a multi-machine equivalent model.

10. A computer-readable storage medium storing a program for electronic data interchange, wherein, The program causes the terminal to perform the method as described in any one of claims 1-8.

Images