Dq asymmetric three-phase power grid impedance facing identification method, device, equipment and medium

Abstract

This application discloses a method, apparatus, equipment, and medium for identifying the impedance of an asymmetrical three-phase power grid (dq), belonging to the field of power system technology. The method involves sampling voltage or current superimposed with zero-mean broadband small perturbations, obtaining the impedance through synchronous filtering and Park transform, performing DFT on the point records, establishing a generalized complex linear frequency domain relationship model, and performing local rational approximation or weighted least squares on the narrow-frequency neighborhood of each target frequency point to obtain the impedance matrix. The impedance matrix is ​​then recovered through analytical mapping, outputting the full-band impedance. This method eliminates the need for frequency-by-frequency sweeping and long steady-state recording, enabling the identification of the equivalent impedance of an asymmetrical three-phase power grid in a short time.

Description

Technical Field

[0001] This invention relates to the field of power system technology, specifically to a method, apparatus, equipment, and medium for identifying impedance in a three-phase power grid with dq asymmetry. Background Technology

[0002] In new energy power electronic power systems, the converter-grid coupling relationship is often described using small-signal impedance. Traditional identification paths are mostly based on the symmetrical three-phase assumption or frequency-by-frequency sweep: firstly, assuming... First, the cross-coupling is weak or negligible, resulting in a diagonally dominant impedance model. Second, the point-by-point solution by hardware frequency sweep injection is time-consuming, has high requirements for steady-state operation, and is sensitive to non-stationary / short-term disturbances. Third, discrete ARX-like methods require higher-order models and long-term data to cover a wide frequency band, and are prone to systematic bias when facing asymmetric / negative-sequence coupling.

[0003] On the other hand, actual distribution networks and grid connection points often suffer from factors such as three-phase imbalance, line resistance / load asymmetry, and differences in control parameters, making... The strong coupling and non-conjugate nature of the two axes cause omissions in the "symmetric"-based identification. Therefore, an identification method is needed that can balance asymmetric coupling and broadband frequency domain characteristics under short time windows and non-periodic small perturbation conditions, and provide a verifiable, noise-resistant, and online-applicable implementation path. Summary of the Invention

[0004] Based on this, it is necessary to propose a method, device, equipment and medium for identifying the impedance of dq asymmetrical three-phase power grids, in order to solve the problems of existing technologies that rely on symmetry assumptions, have long identification time and poor anti-interference ability under asymmetrical three-phase or strongly coupled conditions.

[0005] This application provides a method for identifying the impedance of a dq-asymmetric three-phase power grid, the method comprising: The three-phase voltage instantaneous sampling signal and the three-phase current instantaneous sampling signal are synchronously acquired by superimposing a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter at a preset sampling frequency. The synchronous rotation electrical angle is obtained by using a phase-locked loop. The instantaneous sampling signals of the three-phase voltage and the three-phase current are subjected to Park transformation to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and the complex voltage signal and complex current signal are constructed respectively. After applying a window function to the voltage complex signal and current complex signal with a preset recording length, a discrete Fourier transform is performed to obtain the spectral values ​​of the voltage complex signal and the current complex signal. A generalized complex linear frequency domain relationship model of voltage and current is then established. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions. Within the narrow frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model. A weighted sum of squared residuals is constructed, and the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. Based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined by using the analytical mapping relationship of the dq impedance matrix.

[0006] Furthermore, the step of using a phase-locked loop to obtain the synchronous rotational electrical angle, and performing a Park transform on the acquired instantaneous sampling signals of the three-phase voltage and the three-phase current to obtain the d-axis and q-axis components in the dq synchronous rotating coordinate system includes: according to Determine the d-axis and q-axis components in the dq coordinate system; in, Let be the physical quantity to be transformed, when Time represents voltage, when Time represents current; Three-phase stationary coordinate systems The instantaneous components of phases A, B, and C; Synchronous rotating coordinate system The d-axis and q-axis components are shown below. The synchronous rotational electrical angle is obtained from the phase-locked loop (PLL).

[0007] Furthermore, the construction of the complex voltage signal and the complex current signal respectively includes: Define complex signals ; in, for Complex number representation in a coordinate system; represents the coefficient corresponding to the real part, corresponding to the d-axis component; These are the coefficients corresponding to the imaginary part, and the corresponding q-axis components; The imaginary unit satisfies ; according to Construct a complex voltage signal; according to Construct a complex current signal; in, and These represent voltage and current respectively. Complex space vector representation in a coordinate system.

[0008] Further, the step of applying a window function to the voltage complex signal and the current complex signal of a preset recording length and then performing a discrete Fourier transform to obtain the spectral values ​​of the voltage complex signal and the current complex signal includes: right Apply window function to point data Forming a windowed sequence ,in, It is a discrete-time sampling sequence; The discrete sequence after windowing; This is the sampling point number, and its value range is usually [value range missing]. ; according to Performing a Discrete Fourier Transform yields the spectral values ​​of the complex voltage signal and the complex current signal, where... For sequence In the Discrete Fourier transform results at each frequency point; Frequency point number; This represents the number of sampling points; The sampling frequency; For the first The angular frequency corresponding to each frequency point; Let X be the complex exponential basis function of the DFT, and let X be the physical quantity to be transformed. When X Time represents voltage, when Time represents current. Represents a complex voltage signal No. Spectral values ​​at each frequency point Current complex signal The spectral values.

[0009] Furthermore, a generalized complex linear frequency domain relationship model for voltage and current is established. This model includes generalized frequency domain coefficients for the same-frequency channel, generalized frequency domain coefficients for the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions, including: according to A generalized complex linear frequency domain model of the relationship between voltage and current is established. in, Represents a complex voltage signal No. Spectral values ​​at each frequency point Current complex signal spectral values; Mirror frequency The complex conjugate of the current spectrum reflects the mirror coupling effect in an asymmetric system; These are the generalized frequency domain coefficients of the same frequency channel, i.e., complex current. To complex voltage The main channel equivalent impedance function; These are the generalized frequency domain coefficients of the mirror channel, used to characterize the effects of asymmetric three-phase systems. arrive The coupling relationship; This is an additional term caused by transient components, spectral leakage, and truncation error under finite-duration measurement conditions.

[0010] Furthermore, within the narrow-frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model to construct a weighted residual sum of squares. The estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares, including: At the frequency point to be estimated Narrow frequency neighborhood Inside, to , and At the same time, a low-order rational model is adopted:

[0011]

[0012]

[0013] in, for The first of the numerator polynomials One coefficient, for The denominator polynomial of the first One coefficient, For molecular order, For the order of the denominator, These are the molecular coefficients of the mirror channel model. For the denominator coefficients of the mirror channel model, and These represent the numerator and denominator orders of the mirror channel model, respectively. The first polynomial approximation for transient / leakage terms One coefficient, Let the order of the transient term be the approximation. The term number is the polynomial term number. Normalize the denominator constant term to 1, substitute the DFT relation into, and... Inner superposition, according to Construct a weighted residual sum of squares, and obtain the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated by solving linear least squares or iterative weighted least squares. in, The vector of parameters to be estimated contains , For the first The weighting coefficients corresponding to each frequency point are obtained by solving linear least squares or iterative weighted least squares. Set the fitting residual threshold. , Then adjust the model order. With local bandwidth or weight Reassessment.

[0014] Further, based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel at the frequency to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined using the analytical mapping relationship of the dq impedance matrix, including: The four elements of the dq impedance matrix are determined using the analytical mapping relationship of the dq impedance matrix:

[0015] in, Indicates taking the real part, This indicates taking the imaginary part. The d-axis self-impedance represents the d-axis current. voltage on the d-axis The impact; The q-axis self-impedance represents the q-axis current. q-axis voltage The impact; The cross-coupling impedance represents the q-axis current. voltage on the d-axis The impact; The cross-coupling impedance represents the d-axis current. q-axis voltage The impact at each frequency point get Output amplitude and phase curves and data tables for the entire frequency band.

[0016] This application also provides a device for identifying the impedance of a dq-asymmetric three-phase power grid, the device comprising: The signal sampling unit is used to superimpose a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter, and synchronously acquire the instantaneous sampling signals of the three-phase voltage and the three-phase current at a preset sampling frequency. The Park transform unit is used to obtain the synchronous rotation electrical angle using a phase-locked loop, perform Park transform on the acquired instantaneous sampling signals of the three-phase voltage and the three-phase instantaneous sampling signals to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and construct the complex voltage signal and the complex current signal respectively. The model building unit is used to apply a window function to the voltage complex signal and the current complex signal with a preset recording length, and then perform a discrete Fourier transform to obtain the spectrum values ​​of the voltage complex signal and the current complex signal, and establish a generalized complex linear frequency domain relationship model of voltage and current. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite duration measurement conditions. The estimation and solution unit is used to simultaneously perform local rational approximation on the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions using a low-order rational model within the narrow-frequency neighborhood of the frequency point to be estimated. It constructs a weighted residual sum of squares and obtains the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated by solving linear least squares or iterative weighted least squares. The impedance determination unit is used to determine the equivalent impedance of the dq asymmetrical three-phase power grid based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, and by utilizing the analytical mapping relationship of the dq impedance matrix.

[0017] This application also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor causes the processor to perform the following steps: The three-phase voltage instantaneous sampling signal and the three-phase current instantaneous sampling signal are synchronously acquired by superimposing a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter at a preset sampling frequency. The synchronous rotation electrical angle is obtained by using a phase-locked loop. The instantaneous sampling signals of the three-phase voltage and the three-phase current are subjected to Park transformation to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and the complex voltage signal and complex current signal are constructed respectively. After applying a window function to the voltage complex signal and current complex signal with a preset recording length, a discrete Fourier transform is performed to obtain the spectral values ​​of the voltage complex signal and the current complex signal. A generalized complex linear frequency domain relationship model of voltage and current is then established. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions. Within the narrow frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model. A weighted sum of squared residuals is constructed, and the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. Based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined by using the analytical mapping relationship of the dq impedance matrix.

[0018] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the following steps: The three-phase voltage instantaneous sampling signal and the three-phase current instantaneous sampling signal are synchronously acquired by superimposing a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter at a preset sampling frequency. The synchronous rotation electrical angle is obtained by using a phase-locked loop. The instantaneous sampling signals of the three-phase voltage and the three-phase current are subjected to Park transformation to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and the complex voltage signal and complex current signal are constructed respectively. After applying a window function to the voltage complex signal and current complex signal with a preset recording length, a discrete Fourier transform is performed to obtain the spectral values ​​of the voltage complex signal and the current complex signal. A generalized complex linear frequency domain relationship model of voltage and current is then established. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions. Within the narrow frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model. A weighted sum of squared residuals is constructed, and the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. Based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined by using the analytical mapping relationship of the dq impedance matrix.

[0019] The embodiments of this application have the following beneficial effects: By establishing a generalized complex linear model that includes coefficients of both co-frequency and mirror channels, the mirror coupling under asymmetric operating conditions is explicitly characterized, and the complete 2×2 dq impedance matrix is ​​accurately recovered. Based on a local rational approximation strategy, multi-frequency parallel estimation can be achieved with a single preset short-time window recording, completely eliminating the dependence on traditional frequency sweeping and long steady-state conditions. The transient leakage term and impedance coefficient are jointly fitted and solved in a narrow frequency neighborhood, effectively suppressing spectral leakage and noise deviation caused by short-time aperiodic data. The converter itself is used as a micro-perturbation source to achieve low intrusion, and the calculation process is decoupled into small-scale independent operations with extremely low computing power requirements. It is easy to implement online evaluation on the edge controller, realizing wide-band fast grid impedance identification for asymmetric three-phase power systems without the need for symmetry assumptions in the dq synchronous rotating coordinate system. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application 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 this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] in: Figure 1 This is a flowchart illustrating a method for identifying the impedance of a dq-asymmetric three-phase power grid in one embodiment. Figure 2 This is another flowchart illustrating a method for identifying the impedance of a dq-asymmetric three-phase power grid in one embodiment. Figure 3 This is a structural diagram of a device for identifying the impedance of a three-phase power grid with dq asymmetry in one embodiment; Figure 4 This is a schematic diagram of the structure of a computer device in one embodiment; Figure 5This is a schematic diagram of the structure of a computer-readable storage medium in one embodiment. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0023] This application provides a method for identifying the impedance of a dq-asymmetric three-phase power grid. Please refer to [link to relevant documentation]. Figure 1 , Figure 1 This is a flowchart illustrating a method for identifying the impedance of a three-phase power grid with dq asymmetry in one embodiment; the method for identifying the impedance of a three-phase power grid with dq asymmetry includes steps S1 to S5.

[0024] Step S1: Superimpose a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter, and synchronously acquire the instantaneous sampling signals of the three-phase voltage and the three-phase current at a preset sampling frequency. Specifically, using the grid-connected converter itself as a broadband micro-perturbation source, a zero-mean small perturbation is superimposed on the voltage reference or current reference channel of the grid-connected converter. , This refers to a zero-mean broadband small disturbance signal (such as an RBS or multi-sine wave signal) superimposed on the voltage or current reference channel of the grid-connected converter, with a disturbance amplitude of... It meets the grid connection voltage and current limits and EMC constraints, and Normal amplitude. Based on sampling frequency. Synchronous acquisition , The instantaneous sampling signal of three-phase current can usually be written as: Set an anti-aliasing filter before sampling. ( (The transfer function of the filter). Record length. point( Phase alignment / delay calibration ensures synchronization error of voltage and current channels. Less than 10% of the sampling period.

[0025] Step S2: Use a phase-locked loop to obtain the synchronous rotation electrical angle, perform Park transformation on the collected instantaneous sampling signals of the three-phase voltage and the three-phase current to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and construct the complex voltage signal and the complex current signal respectively. In some embodiments, the step of using a phase-locked loop to obtain the synchronous rotational electrical angle, and performing a Park transform on the acquired instantaneous sampling signals of the three-phase voltage and the three-phase current to obtain the d-axis and q-axis components in the dq synchronous rotating coordinate system includes: First, obtain the synchronization angle from the phase-locked loop (PLL). Performing the Park transformation yields... Quantity: according to Determine the d-axis and q-axis components in the dq coordinate system; in, Let be the physical quantity to be transformed, when Time represents voltage, when Time represents current; Three-phase stationary coordinate systems The instantaneous components of phases A, B, and C; Synchronous rotating coordinate system The d-axis and q-axis components are shown below. The synchronous rotational electrical angle is obtained from the phase-locked loop (PLL).

[0026] In some implementations, the construction of the complex voltage signal and the complex current signal respectively includes: Define complex signals ; in, for Complex number representation in a coordinate system; represents the coefficient corresponding to the real part, corresponding to the d-axis component; These are the coefficients corresponding to the imaginary part, and the corresponding q-axis components; The imaginary unit satisfies ; according to Construct a complex voltage signal; according to Construct a complex current signal; in, and These represent voltage and current respectively. Complex space vector representation in a coordinate system.

[0027] Step S3: Apply a window function to the voltage complex signal and current complex signal with a preset recording length, and then perform a discrete Fourier transform to obtain the spectrum values ​​of the voltage complex signal and the current complex signal. Establish a generalized complex linear frequency domain relationship model between voltage and current. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions. In some implementations, the step of applying a window function to the complex voltage signal and complex current signal of a preset recording length and then performing a discrete Fourier transform to obtain the spectral values ​​of the complex voltage signal and the complex current signal includes: right Apply window function to point data Forming a windowed sequence ,in, It is a discrete-time sampling sequence; The discrete sequence after windowing; This is the sampling point number, and its value range is usually [value range missing]. ; according to Performing a Discrete Fourier Transform yields the spectral values ​​of the complex voltage signal and the complex current signal, where... For sequence In the Discrete Fourier transform results at each frequency point; Frequency point number; This represents the number of sampling points; The sampling frequency; For the first The angular frequency corresponding to each frequency point; Let X be the complex exponential basis function of the DFT, and let X be the physical quantity to be transformed. When X Time represents voltage, when Time represents current. Represents a complex voltage signal No. Spectral values ​​at each frequency point Current complex signal The spectral values.

[0028] Within a finite time window, there exist transient / leakage terms. Then we have a generalized complex linear frequency domain relationship model between voltage and current. ; in, Represents a complex voltage signal No. Spectral values ​​at each frequency point Current complex signal spectral values; Mirror frequency The complex conjugate of the current spectrum reflects the mirror coupling effect in an asymmetric system; These are the generalized frequency domain coefficients of the same frequency channel, i.e., complex current. To complex voltage The main channel equivalent impedance function; These are the generalized frequency domain coefficients of the mirror channel, used to characterize the effects of asymmetric three-phase systems. arrive The coupling relationship; This is an additional term caused by transient components, spectral leakage, and truncation error under finite-duration measurement conditions.

[0029] Set the signal-to-noise ratio threshold :like or If the value is too small, then increase the disturbance amplitude / extend the recording time or change the window function for resampling. In the above formula, For the first Signal-to-noise ratio at each frequency point; This is the signal-to-noise ratio threshold.

[0030] Step S4: Within the narrow frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model to construct a weighted residual sum of squares. The estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the generalized frequency domain coefficients of the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. In some implementations, the step of simultaneously applying a low-order rational model to locally rationally approximate the generalized frequency domain coefficients of the co-channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions within the narrow-band neighborhood of the frequency point to be estimated, constructing a weighted residual sum of squares, and obtaining the estimated values ​​of the generalized frequency domain coefficients of the co-channel and the mirror channel of the frequency point to be estimated by solving linear least squares or iterative weighted least squares includes: At the frequency point to be estimated Narrow frequency neighborhood Inside, to , and At the same time, a low-order rational model is adopted, in which... The angular frequency of the target frequency point to be identified; A narrow-frequency neighborhood established around the target frequency; For narrow-band neighborhood half-bandwidth:

[0031]

[0032]

[0033] in, for The first of the numerator polynomials One coefficient, for The denominator polynomial of the first One coefficient, For molecular order, For the order of the denominator, These are the molecular coefficients of the mirror channel model. For the denominator coefficients of the mirror channel model, and These represent the numerator and denominator orders of the mirror channel model, respectively. The first polynomial approximation for transient / leakage terms One coefficient, Let the order of the transient term be the approximation. The term number is the polynomial term number. Normalize the denominator constant term to 1, substitute the DFT relation into, and... Inner superposition, the DFT relation is also the Discrete Fourier Transform (DFT) relation. ,according to Construct a weighted residual sum of squares, and obtain the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated by solving linear least squares or iterative weighted least squares. in, The vector of parameters to be estimated; , , They are obtained through identification , and The estimated value; For the first The weighting coefficients corresponding to each frequency point are usually used to reflect the differences in signal-to-noise ratio or reliability at different frequency points; Represents the modulus of a complex number; The set of frequency points used for local fitting.

[0034] Include By multiplying the denominator and linearizing, it can be reduced to a set of terms. Linear least squares can be used, or iterative weighted least squares can be employed to improve robustness. This yields... Set the fitting residual threshold. ;like ,in, Target frequency The fitting residual index at the location; This is the residual threshold; when the residual exceeds the threshold, it indicates that the current model order, local bandwidth, or weight settings are insufficient, and the parameters need to be readjusted for identification. In this case, the model order should be adjusted. With bandwidth or weight Reassessment.

[0035] Step S5: Based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined using the analytical mapping relationship of the dq impedance matrix. In some implementations, the equivalent impedance of the dq-asymmetric three-phase power grid is determined by utilizing the analytical mapping relationship of the dq impedance matrix, based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel at the frequency to be estimated, including: The four elements of the dq impedance matrix are determined using the analytical mapping relationship of the dq impedance matrix:

[0036] in, Indicates taking the real part, This indicates taking the imaginary part. The d-axis self-impedance represents the d-axis current. voltage on the d-axis The impact; The q-axis self-impedance represents the q-axis current. q-axis voltage The impact; The cross-coupling impedance represents the q-axis current. voltage on the d-axis The impact; The cross-coupling impedance represents the d-axis current. q-axis voltage The impact at each frequency point get Output amplitude and phase curves and data tables for the entire frequency band.

[0037] In some implementations, it is set ,Depend on and It can be deduced that form ,in, The generalized impedance function of the main channel; For the generalized impedance function of the mirror channel; For complex current variables The complex conjugate of . and They are respectively composed of complex currents The restored real-valued components of the d and q axes; This indicates taking the real part of a complex number; This indicates taking the imaginary part of a complex number.

[0038] in, for The impedance matrix in coordinate system is generally written as ,in, The d-axis self-impedance represents the d-axis current. voltage on the d-axis The impact; The q-axis self-impedance, Represents q-axis current q-axis voltage The impact; The cross-coupling impedance represents the q-axis current. voltage on the d-axis The impact; The cross-coupling impedance represents the d-axis current. q-axis voltage The impact.

[0039] in,

[0040] In the above formula, Indicates taking the real part; Indicates taking the imaginary part, given by the coefficients of the generalized complex linear system. , recover The analytical mapping relationship of the four elements of the impedance matrix.

[0041] Therefore, at each frequency point get .in, , , , They represent the frequency points respectively. The estimated values ​​of the four impedance elements obtained at the location are as follows.

[0042] In some implementations, a generalized complex linear model is used as the core. It is defined that a complex signal for an asymmetric three-phase system, after a finite-time DFT, is transformed into an estimable linear equation, with the mirror term reflected at the frequency point. and The coupling. A local rational approximation is used within the narrow frequency neighborhood of each target frequency point. , With transient terms Simultaneously modeling, and independently solving for each frequency point using weighted least squares. Then, using parsing mapping, it is restored to its original state. The impedance matrix has four elements. The entire process includes SNR thresholding, residual thresholding, and passivity / consistency checks. Regularization smoothing and reweighted iteration are performed as needed to output full-band results. See details in [link to documentation]. Figure 2 Another flowchart illustrating a method for identifying impedance in a dq-asymmetric three-phase power grids in one embodiment is shown.

[0043] It should be noted that existing dq-frame impedance measurement methods are based on the three-phase balance assumption. They employ linear regression to solve for the 2×2 impedance matrix by injecting multiple independent angle disturbance sequences into the DQ plane, using frequency-by-frequency analysis. However, these methods implicitly assume weak coupling between the dq axes and system symmetry, failing to explicitly consider the mirror coupling effect caused by asymmetrical three-phase systems. In actual distribution networks, three-phase imbalance and line resistance asymmetry result in strong coupling and non-conjugate between the dq axes. In this case, the voltage-current relationship cannot be fully described by a traditional 2×2 real-valued matrix. Ignoring mirror coupling terms will lead to systematic deviations in cross-terms such as Zdq and Zqd, and may even misjudge system stability. Other methods, while eliminating frequency sweep dependence and considering dq asymmetry, directly perform ratio calculations or simple averaging on frequency domain sampling points, failing to explicitly include transient and spectral leakage terms caused by finite-duration measurements in the estimation model. Under aperiodic, short-time-window (≤1s) conditions, transient components and leakage errors will couple with the true impedance response, leading to "frequency aliasing" bias in wide-band estimation results, especially noticeable in low- and high-frequency bands. Furthermore, this type of method lacks a closed-loop verification mechanism for estimation quality (such as residual thresholds or passive checks), making it difficult to provide reliable impedance results in engineering settings.

[0044] This invention aims to solve the triple coupling problem in impedance identification of asymmetrical three-phase power grids: mirror coupling caused by asymmetry (i →v) Structural contradictions with traditional dq models; transient-leakage coupling caused by short-window measurements (T and G± aliasing in the frequency domain); estimation bias coupling caused by noise and model mismatch. Existing technologies often address these problems independently or only partially, resulting in an inability to simultaneously meet the requirements of speed, bandwidth, and accuracy under asymmetric, short-window, and noisy conditions.

[0045] To address the aforementioned triple coupling, this invention proposes a collaborative framework of "generalized complex linear modeling—local rational approximation—analytic mapping recovery": Generalized complex linear frequency domain model: Models the voltage-current relationship of an asymmetric three-phase system as G The T-term explicitly captures the mirror coupling, and the T-term explicitly captures the finite-duration effect, thus decoupling the triple coupling from the model structure. Local rational approximation and joint estimation: Unlike global high-order fitting or frequency-by-frequency direct division, this invention performs local rational approximation and joint estimation within a narrow-frequency neighborhood Ω0 of each target frequency point, for G... + G Both T and low-order rational models are used for local approximation. The physical essence of this strategy is to approximate the true frequency response of the impedance by utilizing the local smoothness of rational functions, while absorbing transient leakage through the polynomial term T to avoid contaminating the estimation of G±. Parsing and mapping recovery: after obtaining G± After that, it is not directly equated with Z. dq Instead of elements, it uses a strict complex-to-real analytic mapping. This mapping guarantees that even if G ± There is a small estimation error, which is mapped to Z. dq It maintains linear propagation, thus avoiding direct estimation. , , , Nonlinear coupling amplification may occur at this time.

[0046] Through the aforementioned collaborative framework, this invention achieves the identification of the impedance matrix of an asymmetric three-phase power grid across the entire frequency band within a short time window of ≤1s, without requiring symmetry assumptions. Compared to direct frequency domain division, the local rational approximation suppresses the bias caused by spectral leakage; compared to ignoring G... Compared with the traditional dq model, it reduces the estimation error of cross-coupling impedance under asymmetric operating conditions.

[0047] The technical solution adopted in this embodiment aims to solve the following key problems: Under asymmetrical three-phase or strongly coupled operating conditions, it can accurately identify dq impedance without relying on symmetry assumptions, and effectively handle situations such as... Under common conditions such as short time windows and non-periodic measurement, wide-band information can be obtained with only a limited duration of recording, thus avoiding the dependence of traditional methods on long-term steady-state or dedicated frequency sweep signals. Under complex conditions such as real-world noise, synchronization errors, and frequency leakage, it still has strong robustness and accuracy, and the identification quality can be verified based on indicators such as residuals and passivity. At the same time, it has online and low-intrusion characteristics. Through small-amplitude perturbations and flexible injection strategies, the impact on the operation of the system under test can be controlled within an acceptable range, meeting the engineering needs of online or on-site rapid evaluation.

[0048] By introducing a generalized complex linear model, the mirror coupling induced by the asymmetric three-phase system is explicitly characterized in the derivation and estimation stages, allowing for the recovery of the complete system without the need for symmetry assumptions. Impedance matrix Secondly, based on the exact relation and local rational approximation of finite-duration DFT, a single short recording (≤1s) can be used for parallel estimation at multiple frequency points. Compared to frequency-by-frequency sweeping, this significantly shortens the recognition time and reduces dependence on steady-state operating conditions. Furthermore, a weighted least-squares problem is constructed within a narrow-frequency neighborhood of each frequency point, while simultaneously fitting transient terms. This method avoids frequency leakage bias caused by aperiodic data. Combined with SNR / residual thresholds and passive projection, the identification results maintain stable accuracy under low disturbance amplitude and moderate noise conditions. Using the converter itself as a broadband injection source, the hardware only requires sampling channels and synchronization / anti-aliasing processing. The calculations are all small-scale, independent least squares, which can be quickly implemented at edge controllers or maintenance sites. Furthermore, the method includes a complete quality control and uncertainty assessment process, and can output indicators such as residuals, SNR, and passivity, along with confidence bands, facilitating integration with grid connection specifications and stability criteria (Nyquist / eigenvalues). Moreover, the disturbance amplitude can be dynamically adjusted according to grid and equipment limitations, supporting rapid detection during operation and facilitating grid commissioning, post-fault reassessment, and daily grid monitoring. In this application embodiment, an identification device for dq-asymmetric three-phase power grid impedance is provided. Please refer to [link to relevant documentation]. Figure 3 , Figure 3 This is a structural diagram of an identification device for dq asymmetrical three-phase power grid impedance in one embodiment. The identification device for dq asymmetrical three-phase power grid impedance includes: a signal sampling unit 201, a Park transformation unit 202, a model building unit 203, an estimation and solution unit 204, and an impedance determination unit 205.

[0049] Among them, the signal sampling unit 201 is configured to superimpose a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter, and synchronously acquire the instantaneous sampling signals of the three-phase voltage and the three-phase current at a preset sampling frequency. Park transform unit 202 is configured to obtain synchronous rotation electrical angle using phase-locked loop, perform Park transform on the acquired instantaneous sampling signals of three-phase voltage and three-phase current to obtain d-axis and q-axis components in dq synchronous rotating coordinate system, and construct complex voltage and complex current signals respectively. The model building unit 203 is configured to apply a window function to the voltage complex signal and the current complex signal of a preset recording length and then perform a discrete Fourier transform to obtain the spectral values ​​of the voltage complex signal and the current complex signal, and to establish a generalized complex linear frequency domain relationship model between voltage and current. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite duration measurement conditions. The estimation and solution unit 204 is configured to simultaneously perform local rational approximation on the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions using a low-order rational model within the narrow-frequency neighborhood of the frequency point to be estimated, construct a weighted residual sum of squares, and obtain the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated by solving linear least squares or iterative weighted least squares. Impedance determination unit 205 is configured to determine the equivalent impedance of the dq asymmetrical three-phase power grid by using the analytical mapping relationship of the dq impedance matrix, based on the generalized frequency domain coefficient estimates of the same-frequency channel and the generalized frequency domain coefficient estimates of the mirror channel at the frequency point to be estimated.

[0050] For further details on how each unit in the device for identifying dq-asymmetric three-phase power grid impedance implements the above-mentioned technical solution, please refer to the description in the above-mentioned method for identifying dq-asymmetric three-phase power grid impedance, which will not be repeated here.

[0051] In this application embodiment, a computer device is provided; please refer to... Figure 4 , Figure 4 The diagram below illustrates the structure of a computer device in one embodiment. The device includes a memory 301 and a processor 302. The memory 301 stores a computer program. When the computer program is executed by the processor 302, the processor 302 performs the following steps: The three-phase voltage instantaneous sampling signal and the three-phase current instantaneous sampling signal are synchronously acquired by superimposing a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter at a preset sampling frequency. The synchronous rotation electrical angle is obtained by using a phase-locked loop. The instantaneous sampling signals of the three-phase voltage and the three-phase current are subjected to Park transformation to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and the complex voltage signal and complex current signal are constructed respectively. After applying a window function to the voltage complex signal and current complex signal with a preset recording length, a discrete Fourier transform is performed to obtain the spectral values ​​of the voltage complex signal and the current complex signal. A generalized complex linear frequency domain relationship model of voltage and current is then established. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions. Within the narrow frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model. A weighted sum of squared residuals is constructed, and the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. Based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined by using the analytical mapping relationship of the dq impedance matrix.

[0052] The processor 302 can also be called a CPU (Central Processing Unit). The processor 302 may be an integrated circuit chip with signal processing capabilities. The processor 302 can also be a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. The general-purpose processor can be a microprocessor, or the processor 302 can be any conventional processor.

[0053] In this application embodiment, a computer-readable storage medium is provided; please refer to [link to relevant documentation]. Figure 5 , Figure 5 This is a schematic diagram of the structure of a computer-readable storage medium in one embodiment, on which a readable computer program 401 is stored; wherein, the computer program 401 may be stored in the storage medium in the form of a software product, including a number of instructions to cause a computer device (which may be a personal computer, a server machine, or a network device, etc.) or a processor to perform the following steps: The three-phase voltage instantaneous sampling signal and the three-phase current instantaneous sampling signal are synchronously acquired by superimposing a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter at a preset sampling frequency. The synchronous rotation electrical angle is obtained by using a phase-locked loop. The instantaneous sampling signals of the three-phase voltage and the three-phase current are subjected to Park transformation to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and the complex voltage signal and complex current signal are constructed respectively. After applying a window function to the voltage complex signal and current complex signal with a preset recording length, a discrete Fourier transform is performed to obtain the spectral values ​​of the voltage complex signal and the current complex signal. A generalized complex linear frequency domain relationship model of voltage and current is then established. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions. Within the narrow frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model. A weighted sum of squared residuals is constructed, and the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. Based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined by using the analytical mapping relationship of the dq impedance matrix.

[0054] The aforementioned storage media include: USB flash drives, portable hard drives, magnetic disks or optical disks, ROM (Read-Only Memory), RAM (Random Access Memory), and other media that can store program code, or terminal devices such as computers, servers, mobile phones, and tablets.

[0055] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0056] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0057] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for identification of dq-asymmetrical three-phase grid impedance, characterized in that, Includes the following steps: The three-phase voltage instantaneous sampling signal and the three-phase current instantaneous sampling signal are synchronously acquired by superimposing a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter at a preset sampling frequency. The synchronous rotation electrical angle is obtained by using a phase-locked loop. The instantaneous sampling signals of the three-phase voltage and the three-phase current are subjected to Park transformation to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and the complex voltage signal and complex current signal are constructed respectively. After applying a window function to the voltage complex signal and current complex signal with a preset recording length, a discrete Fourier transform is performed to obtain the spectral values ​​of the voltage complex signal and the current complex signal. A generalized complex linear frequency domain relationship model of voltage and current is then established. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions. Within the narrow frequency neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model. A weighted sum of squared residuals is constructed, and the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. Based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, the equivalent impedance of the dq asymmetrical three-phase power grid is determined by using the analytical mapping relationship of the dq impedance matrix.

2. The method for identifying the dq-asymmetrical three-phase network impedance according to claim 1, characterized in that, The process of obtaining synchronous rotational electrical angles using a phase-locked loop, and performing Park transform on the acquired instantaneous sampling signals of the three-phase voltage and the three-phase current to obtain the d-axis and q-axis components in the dq synchronous rotating coordinate system includes: According to , the d-axis component and the q-axis component in the dq coordinate system are determined; wherein, represents a voltage when represents a current when ; are instantaneous components of the three-phase stationary coordinate system ; are d-axis and q-axis components in the synchronous rotating coordinate system ; is a synchronous rotating electric angle obtained by a phase-locked loop (PLL).

3. The method for identifying the dq-asymmetrical three-phase network impedance according to claim 2, characterized in that, The construction of the complex voltage signal and the complex current signal respectively includes: Defining complex signals ; wherein is a complex representation in a coordinate system; is a coefficient corresponding to the real part, corresponding to the d-axis component; is a coefficient corresponding to the imaginary part, corresponding to the q-axis component; is the imaginary unit, satisfying ; According to , constructing a voltage complex signal; According to , a complex current signal is constructed; wherein, and represent the complex space vector representation of voltage, current in coordinate system, respectively.

4. The method for identifying dq-asymmetrical three-phase network impedance according to claim 1, characterized in that, The step of applying a window function to the voltage complex signal and the current complex signal of a preset recording length and then performing a discrete Fourier transform to obtain the spectrum values ​​of the voltage complex signal and the current complex signal includes: To apply a window function to the point data , forming a windowed sequence wherein, is a discrete time sample sequence; is the windowed discrete sequence; is the sample point number, usually ranging from ; According to performing a discrete Fourier transform to obtain a spectrum value of the voltage complex signal and a spectrum value of the current complex signal, wherein, is a sequence is a sequence of discrete Fourier transform results at the th frequency point; is a frequency point sequence number; is a sampling point number; is a sampling frequency; is an angular frequency corresponding to the th frequency point; is a complex exponential base function of the DFT, X is a physical quantity to be transformed, when X represents the voltage, when represents the current, represents a spectrum value of the voltage complex signal at the th frequency point, represents a spectrum value of the current complex signal .

5. The method for identifying dq-asymmetrical three-phase network impedance according to claim 1, characterized in that, The generalized complex linear frequency domain relationship model of voltage and current is established. This model includes the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite-time measurement conditions, including: According to , a generalized complex linear frequency domain relationship model of voltage and current is established; in, Represents a complex voltage signal No. Spectral values ​​at each frequency point Current complex signal spectral values; Mirror frequency The complex conjugate of the current spectrum reflects the mirror coupling effect in an asymmetric system; These are the generalized frequency domain coefficients of the same frequency channel, i.e., complex current. To complex voltage The main channel equivalent impedance function; These are the generalized frequency domain coefficients of the mirror channel, used to characterize the effects of asymmetric three-phase systems. arrive The coupling relationship; This is an additional term caused by transient components, spectral leakage, and truncation error under finite-duration measurement conditions.

6. The method for identifying impedance of a dq-asymmetric three-phase power grid according to claim 1, characterized in that, Within the narrow-band neighborhood of the frequency point to be estimated, the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions are simultaneously approximated locally using a low-order rational model. A weighted sum of squared residuals is constructed, and the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated are obtained by solving linear least squares or iterative weighted least squares. At the frequency point to be estimated Narrow frequency neighborhood Inside, to , and At the same time, a low-order rational model is adopted: in, for The first of the numerator polynomials One coefficient, for The denominator polynomial of the first One coefficient, For molecular order, For the order of the denominator, These are the molecular coefficients of the mirror channel model. For the denominator coefficients of the mirror channel model, and These represent the numerator and denominator orders of the mirror channel model, respectively. The first polynomial approximation for transient / leakage terms One coefficient, Let the order of the transient term be the approximation. The term number is the polynomial term number. Normalize the denominator constant term to 1, substitute the DFT relation into, and... Inner superposition, according to Construct a weighted residual sum of squares, and obtain the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated by solving linear least squares or iterative weighted least squares. in, The vector of parameters to be estimated contains , For the first The weighting coefficients corresponding to each frequency point are obtained by solving linear least squares or iterative weighted least squares. Set the fitting residual threshold. , Then adjust the model order. With local bandwidth or weight Reassessment.

7. The method for identifying impedance of a dq-asymmetric three-phase power grid according to claim 1, characterized in that, The generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel based on the estimated frequency point are used to determine the equivalent impedance of the dq asymmetrical three-phase power grid using the analytical mapping relationship of the dq impedance matrix, including: The four elements of the dq impedance matrix are determined using the analytical mapping relationship of the dq impedance matrix: in, Indicates taking the real part, This indicates taking the imaginary part. The d-axis self-impedance represents the d-axis current. voltage on the d-axis The impact; The q-axis self-impedance represents the q-axis current. q-axis voltage The impact; The cross-coupling impedance represents the q-axis current. voltage on the d-axis The impact; The cross-coupling impedance represents the d-axis current. q-axis voltage The impact at each frequency point get Output amplitude and phase curves and data tables for the entire frequency band.

8. A device for identifying the impedance of a three-phase power grid with dq asymmetry, characterized in that, The device includes: The signal sampling unit is used to superimpose a zero-mean broadband small disturbance signal on the voltage or current of the grid-connected converter, and synchronously acquire the instantaneous sampling signals of the three-phase voltage and the three-phase current at a preset sampling frequency. The Park transform unit is used to obtain the synchronous rotation electrical angle using a phase-locked loop, perform Park transform on the acquired instantaneous sampling signals of the three-phase voltage and the three-phase instantaneous sampling signals to obtain the d-axis component and q-axis component in the dq synchronous rotating coordinate system, and construct the complex voltage signal and the complex current signal respectively. The model building unit is used to apply a window function to the voltage complex signal and the current complex signal with a preset recording length, and then perform a discrete Fourier transform to obtain the spectrum values ​​of the voltage complex signal and the current complex signal, and establish a generalized complex linear frequency domain relationship model of voltage and current. The generalized complex linear frequency domain relationship model includes the generalized frequency domain coefficients of the same frequency channel, the generalized frequency domain coefficients of the mirror channel, and additional terms caused by transient components, spectral leakage, and truncation errors under finite duration measurement conditions. The estimation and solution unit is used to simultaneously perform local rational approximation on the generalized frequency domain coefficients of the same-frequency channel, the generalized frequency domain coefficients of the mirror channel, and the additional terms caused by transient components, spectral leakage, and truncation errors under finite-duration measurement conditions using a low-order rational model within the narrow-frequency neighborhood of the frequency point to be estimated. It constructs a weighted residual sum of squares and obtains the estimated values ​​of the generalized frequency domain coefficients of the same-frequency channel and the mirror channel of the frequency point to be estimated by solving linear least squares or iterative weighted least squares. The impedance determination unit is used to determine the equivalent impedance of the dq asymmetrical three-phase power grid based on the generalized frequency domain coefficient estimates of the same-frequency channel and the mirror channel of the frequency point to be estimated, and by utilizing the analytical mapping relationship of the dq impedance matrix.

9. A computer device, characterized in that, It includes a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the method as described in any one of claims 1 to 7.

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

Images