A grid sequence domain parameter dynamic identification method and system based on a converter port

Abstract

The application discloses a grid sequence domain parameter dynamic identification method and system based on a converter port, and comprises the following steps: S1, acquiring real-time port voltage and port current at an AC grid-connected port of a grid-connected converter, and converting the port voltage and the port current into positive, negative and zero sequence components to weaken the influence of unbalanced voltage on parameter identification; S2, equivalent external grid to a Thevenin model containing positive, negative and zero sequence equivalent impedance and an equivalent voltage source, wherein the equivalent voltage source is expressed under the condition of real-time angular frequency and phase angle of the system to adapt to grid frequency fluctuation; S3, constructing a parameter regression model between the positive, negative and zero sequence lower port voltage, current, equivalent impedance and equivalent voltage source based on the model, and taking equivalent impedance parameters and equivalent voltage source dq components as to-be-identified parameters; S4, adopting a recursive least square algorithm with an adaptive forgetting factor to perform online parameter identification, and dynamically updating parameter estimation values according to system working condition changes to realize online identification of grid parameters.

Description

Technical Field

[0001] This invention relates to the field of power grid technology, specifically to a method for estimating port impedance in power systems, and in particular to a method and system for dynamic identification of grid sequence domain parameters based on converter ports. Background Technology

[0002] With the widespread application of power electronic converters in renewable energy grid-connected systems, the electrical coupling between the converter and the external power grid has significantly impacted system stability. For grid-connected converters, their operational stability, control parameter configuration, and operational strategy selection are highly dependent on the accurate assessment of the grid strength characteristics at the connection point. Among these, the equivalent impedance presented at the converter's grid connection port is a key parameter characterizing the electrical characteristics of the external power grid, comprehensively reflecting the overall electrical characteristics of the grid-connected line and the distant power grid. However, existing technologies primarily rely on assumptions of grid voltage balance and a fixed angular frequency for identification, making it difficult to provide accurate assessments under conditions of unbalanced voltage and grid frequency fluctuations.

[0003] In engineering practice, the grid short-circuit ratio is an important indicator for measuring the grid's support capacity for converters. Its magnitude directly affects the converter's stability analysis, control parameter tuning, and the selection of grid-connected operation mode. The calculation of the short-circuit ratio usually depends on the grid's equivalent impedance or short-circuit capacity parameters at the grid connection point. Therefore, accurately obtaining the converter's overall port impedance is a fundamental prerequisite for short-circuit ratio evaluation.

[0004] In existing technologies, methods for obtaining port impedance mainly fall into the following categories: 1. Measurement methods based on small disturbance signal injection: This type of method typically injects a small disturbance signal of a specific frequency or amplitude into the converter control system and calculates the impedance parameters based on the response relationship of port voltage and current. This method is highly dependent on the system operating state. When multiple converters are connected to the grid or when operating conditions change frequently, it can easily affect the normal operation of the system, and the stability and consistency of the measurement results are difficult to guarantee. Furthermore, its measurement accuracy is also prone to decrease under conditions of grid voltage imbalance or grid frequency fluctuations.

[0005] 2. Impedance identification methods based on steady-state or quasi-steady-state operating data: These methods typically require the system to be under steady-state or quasi-steady-state conditions. By collecting voltage and current information from the grid connection port and the other end of the line, and combining this with a network model, the impedance parameters of the line or grid can be deduced. This method often relies on synchronous measurement conditions at multiple points and grid structure information, which is difficult to maintain stably in actual distribution networks over long periods. Furthermore, the identification accuracy decreases significantly when grid operating conditions change rapidly. Additionally, the method depends on steady-state conditions and usually assumes a fixed angular frequency, making it difficult to adapt to changes in grid operating conditions.

[0006] 3. Method based on parallel impedance measurement device at the port: This method achieves port impedance measurement by injecting a specific subharmonic signal into the power grid through a dedicated impedance measurement device connected in parallel at the grid-connected port of the converter. Although it can directly obtain port impedance information, it requires additional hardware, increasing system cost and engineering complexity. Furthermore, the injected signal may interfere with power grid operation, and it cannot dynamically adjust parameters online, nor can it consider the effects of positive sequence, negative sequence, and zero sequence impedances.

[0007] Furthermore, most existing technologies focus on measuring or estimating the port impedance itself, making it difficult to simultaneously obtain the overall equivalent impedance and equivalent voltage characteristics of the external power grid based solely on converter port voltage and current measurements without introducing additional devices or affecting system operation. In application scenarios where a single converter is connected to the power grid, existing methods typically still require grid-side voltage measurements or known grid parameter models to further determine the operating status and stability of the remote power grid.

[0008] Therefore, how to simultaneously identify positive-sequence, negative-sequence, and zero-sequence parameters online while relying solely on measurements at the grid-connected converter port, taking into account the system's real-time angular frequency and phase angle, and introducing an adaptive forgetting factor to cope with changes in grid operating conditions, is a problem that urgently needs to be solved in existing technologies. Summary of the Invention

[0009] The main objective of this invention is to address the shortcomings of existing technologies by proposing a dynamic identification method and system for grid sequence domain parameters based on converter ports. This method can obtain the equivalent impedance and equivalent voltage source parameters of the converter's grid-connected ports online and identify them independently in positive-sequence, negative-sequence, and zero-sequence coordinates, thus eliminating the influence of unbalanced voltage on identification accuracy. Furthermore, the method employs a real-time angular frequency and phase angle update regression model and introduces an adaptive forgetting factor to improve the response of parameter identification to changes in grid operating conditions. The acquired port impedance and equivalent voltage can be used to assess grid strength and the operating status of remote grids, thereby assisting the converter in stable operation and control parameter tuning.

[0010] To achieve the above objectives, one aspect of the present invention provides the following technical solution: A dynamic identification method for grid sequence domain parameters based on converter ports includes the following steps: S1, acquiring the real-time port voltage and port current at the AC grid-connected port of the grid-connected converter, and converting them to positive-sequence, negative-sequence, and zero-sequence components to overcome the influence of unbalanced voltage on port impedance identification; S2, equating the external grid of the grid-connected converter to a Thevenin equivalent model containing equivalent impedances and equivalent voltage sources under positive-sequence, negative-sequence, and zero-sequence conditions, wherein the equivalent impedances include equivalent resistances and equivalent inductances, and the equivalent voltage sources are represented under the real-time angular frequency and phase angle conditions of the system to adapt to the grid frequency fluctuations. S3. Based on the Thevenin equivalent model, construct parameter regression models for port voltage, port current, equivalent impedance, and equivalent voltage source under positive sequence, negative sequence, and zero sequence conditions, and use the resistance value of the equivalent resistance, the inductance value of the equivalent inductance, and the dq component of the equivalent voltage source as parameters to be identified; S4. Use a recursive least squares parameter identification algorithm with an adaptive forgetting factor to perform online parameter identification on the parameter regression model, so that the identification process can adapt to changes in system operating conditions, and output the latest estimated values ​​of each parameter to be identified for the regression model update and online parameter update in the next sampling period.

[0011] Further, step S1 includes: acquiring the three-phase voltage signal and three-phase current signal at the AC grid-connected port of the grid-connected converter in real time and converting them to the dq coordinate system corresponding to the positive sequence, negative sequence and zero sequence respectively, to obtain the d-axis component and q-axis component of the port voltage in the positive sequence, negative sequence and zero sequence, and the d-axis component and q-axis component of the port current in the positive sequence, negative sequence and zero sequence.

[0012] Further, step S3 includes: constructing the parameter regression model in the dq coordinate system corresponding to the positive sequence, negative sequence, and zero sequence, wherein the voltage of the equivalent voltage source in the parameters to be identified in the positive sequence, negative sequence, and zero sequence includes d-axis voltage components and q-axis voltage components; the specific steps of constructing the parameter regression model in step S3 include: according to the port voltage equation in the dq coordinate system, expressing the d and q-axis components of the port voltage in the positive sequence, negative sequence, and zero sequence as a linear combination of the d and q-axis components of the port current in the positive sequence, negative sequence, and zero sequence and their differential terms with the parameters to be identified in the positive sequence, negative sequence, and zero sequence, thus obtaining the parameter regression model; the parameter regression model The system includes observation vectors, regression matrices, and parameter estimation vectors for positive, negative, and zero sequences. The observation vectors contain the d-axis and q-axis components of the port voltage for positive, negative, and zero sequences. The regression matrix contains the d-axis and q-axis components of the port current for positive, negative, and zero sequences, as well as their differential terms. The parameter estimation vectors contain the resistance values ​​of the equivalent resistance, the inductance values ​​of the equivalent inductance, and the d-axis and q-axis voltage components of the equivalent voltage source for positive, negative, and zero sequences. The system is updated online during parameter identification using an adaptive forgetting factor to adapt to changes in system operating conditions.

[0013] Furthermore, the mathematical expression of the parametric regression model is: ; in: These represent the positive-order, negative-order, and zero-order dynamic components, respectively; Y m It consists of the d-axis components of the grid-connected converter port voltage in positive sequence, negative sequence, and zero sequence. u m,od and q-axis components u m,oq The observation vector formed; Φ m The regression matrix Φ is in positive, negative, and zero order. m T It is Φ m The transpose of, in which i m,od , i m,oq These represent the d-axis and q-axis components of the grid-connected converter port current in positive sequence, negative sequence, and zero sequence, respectively. oh It is the system's real-time angular frequency. yes i m,od Differential terms in positive, negative, and zero order yes i m,oq Differential terms in positive, negative, and zero order; Θ m It is the resistance value of the equivalent resistance described in positive sequence, negative sequence, and zero sequence.R m,line The inductance values ​​of the equivalent inductance in positive sequence, negative sequence, and zero sequence. L m,line and the d-axis voltage components of the equivalent voltage source in positive sequence, negative sequence, and zero sequence. u m,gd and q-axis voltage component u m,gq The parameter estimation vector is formed.

[0014] Further, step S4 includes: solving the parameter regression model using a recursive least squares method with an adaptive forgetting factor, and simultaneously outputting the resistance value of the equivalent resistance, the inductance value of the equivalent inductance, and the d-axis voltage component and q-axis voltage component of the equivalent voltage source in the positive, negative, and zero sequences.

[0015] Furthermore, step S4 also includes: to facilitate online parameter identification using the recursive least squares method, discretizing the parameter regression model, and setting Y... m =Φ m T Θ m Convert to the following discrete form: ; in, k Indicates the discrete sampling time sequence number. k =1,2,3,…;Y m ( k ) is the first k The observation vector at each sampling time, Φ m ( k ) is the first k The regression matrix at each sampling time, Θ m ( k ) is the first k The parameter estimation vector at each sampling time point; The formula for calculating the recursive least squares method with an adaptive forgetting factor is as follows: ; in, Indicates the first k The parameter estimation vector at each sampling time point, K m ( k ) is the gain vector, ε m ( k P is the prediction error vector. m ( k Let be the covariance matrix, and I be the identity matrix of the corresponding dimension; l m (k) is the adaptive forgetting factor, with a value range of 0 < l m (k) <1, β This is a stability adjustment coefficient, used to adjust the predicted stability index. S m (k) Incorporate gain calculation S m (k) is the predictive stability evaluation index, defined as follows: ; in: For predicting stability evaluation indicators; For parameter prediction error; This serves as the baseline value for error normalization. For the dynamic fluctuation of parameters; This serves as the baseline value for parameter normalization. , , Let be the weighting coefficients, and satisfy: ; To achieve dynamic prediction and online updating of parameters under positive, negative, and zero order, an adaptive forgetting factor is introduced. l m Design of (k), adaptive forgetting factor l m (k) Based on the predicted stability index The dynamic adjustment is calculated as follows: ; in, l m,min and l m,max These are the lower and upper limits of the forgetting factor, respectively, and their values ​​satisfy 0 < l m,min < l m,max ≤1; As a predictive stability evaluation index, it consists of the prediction error of each order component and the parameter fluctuation. c The adjustment coefficient is a positive real number, usually in the range of 0.1 to 5. It is adjusted according to the system power level, sampling period and dynamic response requirements to balance the stability under normal operating conditions and the dynamic tracking capability under fault conditions. Methods for dynamic prediction and updating of power grid parameters based on the order domain include: The port voltage and port current of the grid-connected converter are sampled in real time, and the initial parameters of the algorithm and the forgetting factor λ are set. Among them, the initial values ​​P of the covariance matrix under positive sequence, negative sequence and zero sequence are... m (0) Initialize as a diagonal matrix, with diagonal elements taking values ​​in the range of 10.2 ~10 6 This reflects the high degree of uncertainty in parameter estimation at the initial time; the initial values ​​Θ of the parameter estimation vectors in positive, negative, and zero sequences. m (0) Based on empirical values, either initialize it as a zero vector or as an initial value vector within the physically feasible range of each parameter to be identified, to ensure the stability and convergence of the parameter estimation process; in the first... k At each sampling time, based on the port voltage and port current obtained from the current sampling, observation vectors Y in positive sequence, negative sequence, and zero sequence are constructed. m ( k ) and regression matrix Φ m ( k And compare it with the parameter estimates Θ in the positive, negative and zero sequences of the previous sampling time. m ( k -1) and covariance matrix P m ( k Substituting -1) into the calculation formula of the recursive least squares method with adaptive forgetting factor, we obtain the current time step. k Parameter estimation results Θ in positive order, negative order and zero order m ( k The covariance matrix P in positive, negative, and zero order. m ( k ), where Θ m ( k This includes the resistance values ​​of the equivalent resistance under positive sequence, negative sequence, and zero sequence, the inductance values ​​of the equivalent inductance under positive sequence, negative sequence, and zero sequence, and the d-axis voltage components and q-axis voltage components of the equivalent voltage source under positive sequence, negative sequence, and zero sequence. Subsequently, the above sampling, construction, and recursive update process is repeated at the next sampling time to realize the online identification of the equivalent impedance parameters and the equivalent voltage source parameters under positive sequence, negative sequence, and zero sequence. The parameter identification is performed in positive, negative, and zero order. Real-time angular frequency and phase angle are used for regression model updates. The adaptive forgetting factor is dynamically adjusted according to changes in system operating conditions to improve robustness to unbalanced voltage and frequency fluctuations.

[0016] Furthermore, the method includes the following steps: S5, constructing an overall port impedance model of the grid-connected converter based on the resistance values ​​of the equivalent resistance under positive sequence, negative sequence, and zero sequence obtained from parameter identification, and the inductance values ​​of the equivalent inductance under positive sequence, negative sequence, and zero sequence; S6, calculating the short-circuit ratio at the grid connection point of the grid-connected converter based on the overall port impedance model; the short-circuit ratio calculation is mainly based on the positive sequence equivalent impedance and the positive sequence voltage amplitude, while the negative sequence and zero sequence parameters are used to assist in judging voltage imbalance or fault conditions; S7, judging the operating status of the external power grid based on the d-axis voltage components and q-axis voltage components of the equivalent voltage source output in step S4 under positive sequence, negative sequence, and zero sequence, including whether there is voltage drop, voltage fluctuation, power grid strength change, or voltage imbalance fault, and providing a basis for the converter to adopt corresponding control strategies or operating modes.

[0017] Furthermore, the overall port impedance model is as follows: ; in, R m,line , L m,line These are the equivalent resistance values ​​for positive sequence, negative sequence, and zero sequence obtained in step 4, and the equivalent inductance values ​​for positive sequence, negative sequence, and zero sequence, respectively. oh m (k) represents the real-time angular frequency of the system at the current moment in positive, negative, and zero order. j The imaginary unit; real-time angular frequency of the system oh m Under (k), the magnitude of the overall port impedance model is expressed as: ; The overall port impedance model is used to characterize the overall electrical characteristics of the external power grid visible to the grid-connected converter port, and serves as the basis for the short-circuit ratio evaluation in step S6. Step S6 calculates the short-circuit ratio index, specifically including: Calculate the short-circuit capacity at the grid connection point S sc : ; in, U g The effective value of the grid voltage at the grid connection point; Combined with the rated capacity of the grid-connected converter S n The short-circuit ratio (SCR) is obtained as follows: ; This enables a quantitative assessment of the strength and weakness characteristics of the external power grid.

[0018] In another aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, can implement the steps of the aforementioned method.

[0019] In another aspect, this invention proposes a dynamic identification system for grid sequence domain parameters based on converter ports, comprising: a port acquisition module for acquiring real-time port voltage and port current at the AC grid-connected port of the grid-connected converter, and converting the signals to positive, negative, and zero-sequence coordinate systems to improve identification accuracy under unbalanced voltage conditions; a dynamic synchronization reference system construction module for constructing a dynamic synchronization reference system based on the system's real-time frequency and phase information; a sequence domain dynamic decomposition module for performing sequence domain dynamic decomposition on the three-phase voltage and three-phase current signals; a parameter adaptive dynamic update module for dynamically adjusting the parameter update process based on the predicted stability evaluation results; an equivalent model construction module for constructing a Thevenin equivalent model of the external grid of the grid-connected converter, wherein the Thevenin equivalent model includes equivalent impedance and equivalent voltage sources, and the equivalent impedance includes equivalent resistance and equivalent inductance; and a parameter regression model construction module for constructing a system based on the... The Thevenin equivalent model constructs a parametric regression model between port voltage, port current, equivalent impedance, and equivalent voltage source. The parameters to be identified in the parametric regression model include the resistance value of the equivalent resistance, the inductance value of the equivalent inductance, and the voltage of the equivalent voltage source. A parameter identification module uses a recursive least squares algorithm with an adaptive forgetting factor to perform online parameter identification on the parametric regression model and outputs the latest estimated values ​​of each parameter to be identified for updating the regression model and online parameters in the next sampling period. An impedance calculation module constructs an overall port impedance model of the grid-connected converter's grid-connected port based on the equivalent resistance and equivalent inductance output by the parameter identification module. A state assessment module calculates the short-circuit ratio based on the overall port impedance model and determines the external power grid operating state based on the voltage components of the equivalent voltage source in positive, negative, and zero sequence.

[0020] The beneficial effects of this invention are mainly reflected in the following aspects: This invention acquires voltage and current signals at the grid-connected converter port and constructs parameter regression models based on the Thevenin equivalent model under positive-sequence, negative-sequence, and zero-sequence conditions. Combined with the system's real-time angular frequency and phase angle update regression model, a recursive least squares algorithm with an adaptive forgetting factor is used for online parameter identification. Relying only on single-end electrical quantities (port voltage, port current), it achieves synchronous and online identification of external grid equivalent parameters (equivalent impedance, equivalent voltage source dq components) under positive-sequence, negative-sequence, and zero-sequence conditions without needing to obtain grid topology, line parameters, or remote measurement information. Ultimately, this invention enables high-precision online identification of the equivalent impedance and equivalent voltage source dq components of the external grid under positive-sequence, negative-sequence, and zero-sequence conditions without relying on prior knowledge of grid topology, introducing additional hardware, or affecting normal system operation. The parameter identification results can quickly respond to changes in power grid operating conditions, and realize efficient, adaptive, and non-intrusive online identification of equivalent parameters of the external power grid. It has the advantages of strong real-time performance, low implementation cost, and good engineering adaptability.

[0021] Furthermore, based on the parameter identification results, the present invention constructs the overall port impedance presented to the grid-connected port of the grid-connected converter through step S5, and further calculates the short-circuit ratio at the grid connection point through step S6, providing a basis for the stability analysis, control parameter tuning and operation strategy selection of the converter.

[0022] Furthermore, based on the parameter identification results, the present invention uses the equivalent voltage source voltage and its variation characteristics output in step S4 in step S7 to determine the operating status and stability of the remote power grid, thereby improving the converter's ability to sense changes in the power grid operating environment.

[0023] The technical effects of the present invention are significantly different from those of the prior art. In particular, when dealing with unbalanced voltage conditions, power grid frequency fluctuations and changes in system operating conditions, it relies on positive, negative and zero sequence identification, real-time angular frequency / phase angle updates and adaptive forgetting factors to achieve online high-precision, dynamically adaptive port impedance and equivalent voltage source parameter identification capabilities that cannot be provided by existing methods.

[0024] Unlike conventional variable forgetting factor recursive least squares methods that adjust the forgetting factor solely based on identification error, the adaptive forgetting factor of this invention is adjusted under the constraints of positive-sequence, negative-sequence, and zero-sequence port equivalent models, combined with prediction error and dynamic fluctuation of parameters, to support the synchronous online identification of external equivalent impedance and equivalent voltage source parameters. Attached Figure Description

[0025] Figure 1 This is a flowchart of a grid parameter identification method based on grid-connected converter port measurements according to an embodiment of the present invention.

[0026] Figure 2 This is the main circuit topology of the grid-connected converter in an embodiment of the present invention.

[0027] Figure 3 This is a flowchart of the power grid parameter identification process according to an embodiment of the present invention.

[0028] Figure 4 This is the parameter identification result under the first power grid condition in the embodiment of the present invention.

[0029] Figure 5 This is the parameter identification result under the second power grid condition in the embodiment of the present invention.

[0030] Figure 6 This is a comparison of the results of the embodiments of the present invention with those of the traditional recursive least squares identification method based on the forgetting factor. Detailed Implementation

[0031] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but do not constitute a limitation on the scope of protection of the present invention. For those skilled in the art, any equivalent modifications or improvements made without departing from the concept of the present invention should fall within the scope of protection of the present invention.

[0032] This invention provides a dynamic identification method for grid sequence domain parameters based on converter ports. This method addresses engineering scenarios where the external grid structure and parameters are difficult to obtain and operating conditions are constantly changing during the actual operation of grid-connected converters. It constructs an external grid equivalent parameter identification mechanism that relies solely on converter port measurement information. This invention uses the converter grid-connected port as the sole observation interface. By online acquisition of port voltage and current signals in positive, negative, and zero sequence states, a Thevenin equivalent model is established, including positive, negative, and zero sequence equivalent impedances and equivalent voltage sources. The equivalent impedance includes the equivalent resistance R. m,line and equivalent inductance L m,line The equivalent voltage source at the system's real-time angular frequency ω m (k) and real-time phase angle θ m (k) represents the condition, and a parametric regression model is constructed within this model framework, introducing an adaptive forgetting factor λ. m (k), by predicting the stability index S m (k) Dynamic adjustment enables continuous parameter updates and rapid response to changes in grid operating conditions, allowing the parameter identification process to continuously track and adaptively update when grid operating conditions change. This provides real-time and accurate data for grid strength assessment, judging the operating status of remote grids (voltage drop, voltage fluctuation, strength change or imbalance fault), and optimizing converter control strategies.

[0033] Please refer to Figure 1 The grid parameter identification method based on grid-connected converter port measurement proposed in this embodiment of the invention mainly includes the following steps S1~S4: S1. Port Voltage and Current Signal Acquisition: Real-time acquisition of three-phase voltage and current signals at the AC grid-connected port of the grid-connected converter. The acquired signals are then converted to the d-axis and q-axis coordinate systems (synchronously rotating coordinate systems) corresponding to positive, negative, and zero sequences, respectively, to obtain the d-axis and q-axis components of the port voltage and the port current, thus overcoming the influence of unbalanced voltage on port impedance identification. During the conversion process, the system's real-time angular frequency ω is used. m (k) and real-time phase angle θ m (k) Update the dq coordinate reference to adapt to grid frequency fluctuations and changes in operating conditions; S2. Establishment of External Power Grid Equivalent Model: The external power grid of the grid-connected converter is equivalent to a Thevenin equivalent model, including equivalent impedance and equivalent voltage sources in positive-sequence, negative-sequence, and zero-sequence coordinate systems. The equivalent impedance includes equivalent resistance. R m,line and equivalent inductance L m,line The equivalent voltage source is represented by dq components and updated under the real-time angular frequency and phase angle conditions of the system to ensure that the model can accurately reflect the dynamic characteristics of the power grid. S3. Parametric Regression Model Construction: Based on the Thevenin equivalent model, a linear parametric regression model is constructed in the dq coordinate system corresponding to the positive, negative, and zero sequences, relating the port voltage, current, equivalent impedance, and the dq component of the equivalent voltage source. The regression model includes an observation vector, a regression matrix, and a parameter estimation vector. The observation vector represents the dq component of the port voltage, the regression matrix represents the dq component of the port current and its differential terms, and the parameter estimation vector includes the equivalent resistance. R m,line Equivalent inductance L m,line and the dq component of the equivalent voltage source u m,gd , u m,gq This model ensures independent identification of the three sequences and can be directly used for parameter updates in recursive least squares algorithms. S4. Online parameter identification: Employing an adaptive forgetting factor. l m The recursive least squares algorithm (k) is used to perform online parameter identification for the regression model, including the forgetting factor. l m(k) It can be dynamically adjusted according to changes in port current or power, enabling the identification process to adapt to changes in system operating conditions. The algorithm independently outputs the latest estimated values ​​of each parameter to be identified, including equivalent resistance, in positive, negative, and zero-sequence coordinate systems. R m,line Equivalent inductance L m,line and the dq component of the equivalent voltage source u m,gd , u m,gq This provides a foundation for updating the regression model and online parameters in the next sampling period, enabling synchronous, online, and high-precision updates of port impedance and equivalent voltage source parameters.

[0034] In a preferred embodiment of the present invention, after completing the above parameter identification, based on the parameter identification results, further online identification of port impedance can be performed, as well as short-circuit ratio evaluation, equivalent voltage source d and q component prediction, and external power grid status judgment based on the identification results. Based on this, as... Figure 1 As shown, the method of this embodiment of the invention may further include the following steps S5 to S7: S5. Based on the resistance values ​​of the equivalent resistances in the positive sequence, negative sequence, and zero sequence obtained from parameter identification, and the inductance values ​​of the equivalent inductances in the positive sequence, negative sequence, and zero sequence, construct the overall port impedance model presented to the outside of the grid-connected converter's grid-connected port. S6. Calculate the short-circuit ratio at the grid connection point of the grid-connected converter based on the overall port impedance model. In a preferred embodiment, the short-circuit ratio calculation is mainly based on the positive sequence equivalent impedance and the positive sequence voltage amplitude, while the negative sequence and zero sequence parameters are used to assist in judging voltage imbalance or fault conditions. S7. Based on the d-axis voltage components and q-axis voltage components of the equivalent voltage source output in step S4 under positive sequence, negative sequence and zero sequence, the operating status of the external power grid is judged, including whether there is voltage drop, voltage fluctuation, power grid strength change or voltage imbalance fault, and the basis is provided for the converter to take corresponding control strategies or operating modes.

[0035] In step S2, the external power grid of the converter is equivalently represented by a Thevenin equivalent model that includes equivalent impedance and equivalent voltage sources, where the equivalent impedance includes equivalent resistance and equivalent inductance. Specifically, the main circuit topology of the grid-connected converter is as follows: Figure 2 As shown, where: V dc This refers to the DC side voltage of the grid-connected converter. C v For the DC-side filter capacitor of the grid-connected converter, L f For filter inductance, C f For filter capacitors,u o , i o These are the three-phase output voltage and three-phase output current of the grid-connected converter, respectively. R line , L line These are the equivalent resistance and equivalent inductance of the external power grid line impedance (i.e., the equivalent impedance), respectively. u g The voltage at the grid connection point is denoted as . This equivalent method can characterize the comprehensive electrical effect of the external power grid on the grid-connected converter under single-port observation conditions, transforming the complex problem of power grid parameter identification into a standard linear system parameter estimation problem. This facilitates the subsequent construction of parameter regression models and online identification, laying the foundation for the application of efficient recursive algorithms.

[0036] According to such Figure 2 Based on the circuit structure shown and Kirchhoff's voltage law equations, the expression for the three-phase stationary coordinate system can be obtained as follows: (1); After coordinate transformation, the expression in the dq coordinate system is obtained as follows: (2); In equation (2), These represent the positive-order, negative-order, and zero-order dynamic components, respectively. u m,od and u m,oq These represent the d-axis and q-axis components of the three-phase output voltage (i.e., port voltage) of the grid-connected converter in positive sequence, negative sequence, and zero sequence. i m,od and i m,oq These represent the d-axis and q-axis components of the three-phase output current (i.e., port current) of the grid-connected converter in positive sequence, negative sequence, and zero sequence. u m,gd and u m,gq The d-axis and q-axis voltage components of the equivalent voltage source in positive sequence, negative sequence, and zero sequence. Oh, oh m The system angular frequencies in positive sequence, negative sequence, and zero sequence are given. s is the complex frequency domain variable in the Laplace transform.

[0037] In step S3, based on the Thevenin equivalent model from step S2, a parametric regression model is constructed in the dq coordinate system, relating port voltage, port current, equivalent impedance, and equivalent voltage source. This model includes the equivalent impedance parameters (i.e., the resistance values ​​of the equivalent resistance) under positive, negative, and zero sequence conditions. R s,lineInductance value of equivalent inductance L s,line The d-axis voltage components and q-axis voltage components of the equivalent voltage source in positive sequence, negative sequence, and zero sequence are uniformly used as parameters to be identified.

[0038] The specific process of constructing the parametric regression model includes: rearranging the port voltage equations (2) in the dq coordinate system to obtain: (3); By expressing the d- and q-axis components of the port voltage as a linear combination of the d- and q-axis components of the port current, their differential terms, and the parameters to be identified, a parametric regression model is obtained, the mathematical expression of which is: (4); This parametric regression model includes observation vectors, regression matrices, and parameter estimation vectors in positive, negative, and zero order. Wherein, Y... m It consists of the d-axis components of the grid-connected converter port voltage in positive sequence, negative sequence, and zero sequence. u m,od and q-axis components u m,oq The observation vector formed; Φ m The regression matrix Φ is in positive, negative, and zero order. m T It is Φ m The transpose of, in which i m,od , i m,oq These represent the d-axis and q-axis components of the grid-connected converter port current in positive sequence, negative sequence, and zero sequence, respectively. oh It is the system's real-time angular frequency. yes i m,od Differential terms in positive, negative, and zero order yes i m,oq Differential terms in positive, negative, and zero order; Θ m It is the resistance value of the equivalent resistance described under positive sequence, negative sequence, and zero sequence. R m,line The inductance values ​​of the equivalent inductance under positive sequence, negative sequence, and zero sequence. L m,line and the d-axis voltage components of the equivalent voltage source in positive sequence, negative sequence, and zero sequence. u m,gd and q-axis voltage component u m,gq The parameter estimation vector is formed.

[0039] For the parametric regression model shown in equation (4), regression algorithms such as the least squares algorithm can be used to solve for Y.m =Φ m T Θ m Because Y m and Φ m The value of Θ changes continuously in the time domain, and the data is continuously updated through the algorithm, prompting Θ to... m The parameters approximate the true values. In practical applications, when the impedance undergoes a step change, the data at the two steady-state operating points before and after the impedance change do not satisfy the same estimation equation, which leads to fitting errors. Therefore, this embodiment of the invention uses a recursive least squares method with an adaptive forgetting factor to solve equation (4) for online real-time parameter identification.

[0040] In step S4, to facilitate online parameter identification using the recursive least squares method, the continuous parametric regression model of equation (4) needs to be discretized. Therefore, Y... m =Φ m T Θ m Converted to discrete-time form, it can be expressed as: (5); in, k Indicates the discrete sampling time sequence number. k =1,2,3,…;Y m ( k ) is the first k The observation vector at each sampling time, Φ m ( k ) is the first k The regression matrix at each sampling time, Θ m ( k ) is the first k The parameter estimation vector at each sampling time point; The formula for calculating the recursive least squares method with an adaptive forgetting factor is as follows: ; in, Indicates the first k The parameter estimation vector at each sampling time point, K m ( k ) is the gain vector, ε m ( k P is the prediction error vector. m ( k Let be the covariance matrix, and I be the identity matrix of the corresponding dimension; l m (k) is the adaptive forgetting factor, with a value range of 0 < l m (k) <1, when lm When (k) is greater than 1, the weight of historical data increases with time, making it impossible to attenuate the weight of older data; when l m When the value of (k) is less than or equal to 0, the weight may be non-positive or degenerate to zero, causing the algorithm to lose the meaning of least squares or become unstable. β This is a stability adjustment coefficient, used to adjust the predicted stability index. S m (k) Incorporate gain calculation S m (k) is the predictive stability evaluation index, defined as follows: in: For predicting stability evaluation indicators; For parameter prediction error; This serves as the baseline value for error normalization. For the dynamic fluctuation of parameters; This serves as the baseline value for parameter normalization. , , Let be the weighting coefficients, and satisfy: To achieve dynamic prediction and online updating of parameters in positive, negative, and zero orders, an adaptive forgetting factor is introduced. l m Design of (k), adaptive forgetting factor l m (k) Based on the predicted stability index The dynamic adjustment is calculated as follows: in, l m,min and l m,max These are the lower and upper limits of the forgetting factor, respectively, and their values ​​satisfy 0 < l m,min < l m,max ≤1; As a predictive stability evaluation index, it consists of the prediction error of each order component and the parameter fluctuation. c This is the adjustment coefficient, which takes the value of a positive real number, usually in the range of 0.1 to 5. It is adjusted according to the system power level, sampling period and dynamic response requirements to balance the stability under normal operating conditions and the dynamic tracking capability under fault conditions.

[0041] refer to Figure 3 The calculation process of recursive least squares based on the forgetting factor includes: The port voltage and port current of the grid-connected converter are sampled in real time, and the initial parameters of the algorithm and the forgetting factor λ are set. Among them, the initial values ​​P of the covariance matrix under positive sequence, negative sequence and zero sequence are... m (0) Initialize as a diagonal matrix, with its diagonal elements taking large positive numbers, preferably in the range of 10. 2 ~10 6 This reflects the high degree of uncertainty in parameter estimation at the initial time; the initial values ​​Θ of the parameter estimation vectors in positive, negative, and zero sequences. m (0) Based on empirical values ​​(if there is a general understanding of the values ​​of the parameters to be identified, such as the d-axis component of the low-voltage distribution network voltage being around 300V, the initial value can be set to 300V to start identification, or it can be set to start identification from zero), or initialized as a zero vector, or initialized as an initial value vector within the physically feasible range of each parameter to be identified, to ensure the stability and convergence of the parameter estimation process; in the first k At each sampling time, based on the port voltage and port current obtained from the current sampling, an observation vector Y is constructed for positive sequence, negative sequence, and zero sequence. m ( k ) and regression matrix Φ m ( k And compare it with the parameter estimates Θ in the positive, negative and zero sequences of the previous sampling time. m ( k -1) and covariance matrix P m ( k Substituting -1) into the calculation formula of the recursive least squares method with adaptive forgetting factor, we obtain the current time step. k Parameter estimation results Θ in positive order, negative order and zero order m ( k The covariance matrix P in positive, negative, and zero order. m ( k ), where Θ m ( k This includes the resistance values ​​of the equivalent resistance under positive sequence, negative sequence, and zero sequence, the inductance values ​​of the equivalent inductance under positive sequence, negative sequence, and zero sequence, and the d-axis voltage components and q-axis voltage components of the equivalent voltage source under positive sequence, negative sequence, and zero sequence. Subsequently, the above sampling, construction, and recursive update process is repeated at the next sampling time to realize the online identification of the equivalent impedance parameters and the equivalent voltage source parameters under positive sequence, negative sequence, and zero sequence.

[0042] By employing a recursive least squares method with an adaptive forgetting factor to perform online parameter identification on the parametric regression model constructed in step S3, the equivalent resistance values ​​under positive, negative, and zero sequence conditions can be output simultaneously without injecting disturbance signals or acquiring voltage / current signals at the other end of the line. Rm,line The inductance values ​​of the equivalent inductance in positive sequence, negative sequence, and zero sequence. L m,line And the d-axis voltage components and q-axis voltage components of the equivalent voltage source in positive sequence, negative sequence and zero sequence.

[0043] Among the parameter identification methods described above, the Adaptive Forgetting Factor RLS (AFF-RLS) method used in this invention is not the only available parameter identification algorithm. However, it differs from conventional recursive least squares methods or other online identification methods mainly in that it can dynamically adjust the forgetting factor according to changes in port current or power. l s (k) enables rapid response and adaptive updating of parameter estimation during sudden changes in grid operating conditions or frequency fluctuations, and is particularly suitable for the synchronous identification of port impedance and equivalent voltage source dq components in positive-sequence, negative-sequence, and zero-sequence configurations. Those skilled in the art can also employ other parameter identification algorithms with online recursive update capabilities and apply them to the unified parameter identification under the constraints of the aforementioned grid-connected converter port equivalent model, so as to achieve the synchronous identification of overall port impedance parameters and external equivalent voltage source parameters using only port voltage and port current measurements.

[0044] In step S5, based on the equivalent resistance value output in step S4... R m,line and equivalent inductance value L m,line Construct an overall port impedance model of the grid-connected converter's external ports: (7); in, Z line It represents the overall port impedance presented to the outside world by the grid-connected port of the grid-connected converter under positive sequence, negative sequence and zero sequence (i.e. the equivalent line impedance of the power grid seen by the grid-connected port of the grid-connected converter under positive sequence, negative sequence and zero sequence). j It is the imaginary unit.

[0045] real-time angular frequency of the system oh m Under (k), the magnitude of the overall port impedance model described above can be expressed as: (8); The overall port impedance model described above is used to characterize the overall electrical characteristics of the external power grid visible to the grid-connected converter port, and serves as the basis for the short-circuit ratio evaluation in step S6.

[0046] Step S6 involves calculating the short-circuit ratio, specifically including: First, calculate the short-circuit capacity at the grid connection point.S sc : (9); in, U g The effective value of the grid voltage at the grid connection point; Then, combined with the rated capacity of the grid-connected converter S n The short-circuit ratio (SCR) is obtained as follows: (10); This enables a quantitative assessment of the strength and weakness characteristics of the external power grid.

[0047] In step S7, the components of the external power grid equivalent voltage source output in step S4 in the dq coordinate system corresponding to the positive sequence, negative sequence, and zero sequence are calculated. u m,gd , u m,gq This involves assessing the operating status of the external power grid, including the presence of voltage dips, voltage fluctuations, or changes in grid strength, and providing a basis for the converter to adopt corresponding control strategies or operating modes. Specifically, for positive-sequence, negative-sequence, or zero-sequence voltages, the equivalent voltage amplitude can be calculated. (11); Then, the voltage amplitudes in positive sequence, negative sequence, or zero sequence are | u m,g | Compare with its corresponding rated reference value to determine whether a voltage drop has occurred; determine whether there is voltage fluctuation based on the fluctuation range of each sequence voltage amplitude; analyze whether the strength of the external power grid has changed in conjunction with the overall port impedance model or the change in short-circuit ratio.

[0048] Using this method, step S7 can more accurately determine the external power grid status under three-sequence independent conditions, thereby improving the converter's response capability and control strategy adaptability under unbalanced operating conditions and power grid frequency fluctuations.

[0049] The foregoing embodiments of the present invention provide a method for dynamic identification of grid sequence domain parameters based on converter ports. This method overcomes the dependence of existing technologies on grid topology information, multi-point measurements or additional disturbance signals. It can identify the equivalent impedance parameters and equivalent voltage source dq components of the external grid in positive sequence, negative sequence and zero sequence by using only the voltage and current measurement information of the grid-connected converter's own ports online, synchronously and adaptively.

[0050] Based on the converter port equivalent model and parameter regression method, this invention can obtain the positive-sequence, negative-sequence, and zero-sequence port impedance parameters at the grid connection point, as well as the corresponding d-axis and q-axis components of the equivalent voltage source, without injecting additional disturbance signals or adding hardware measurement devices. On this basis, by analyzing the port impedance parameters, online evaluation of grid strength indicators such as the short-circuit ratio at the grid connection point can be achieved; by real-time monitoring of the dq components of the equivalent voltage source, it is possible to determine whether there are voltage drops, voltage fluctuations, or other changes in operating status in the remote grid, thus providing real-time and accurate basis for converter stability analysis, control parameter tuning, and operating strategy selection.

[0051] Furthermore, the recursive least squares algorithm with an adaptive forgetting factor used in this invention can dynamically adjust the forgetting factor λ according to changes in port current or power. s (k) ensures rapid response and adaptive updating of parameter identification under operating conditions such as grid frequency fluctuations or load changes; at the same time, the positive sequence, negative sequence and zero sequence identification strategies can effectively eliminate the impact of unbalanced voltage on identification accuracy.

[0052] Therefore, the method of the present invention has good compatibility with the existing grid-connected converter control and measurement system, which can reduce the complexity of engineering implementation, improve the operational safety and adaptability of the converter in complex and dynamic power grid environments, and provide reliable support for power grid strength assessment and converter control strategy optimization.

[0053] Another embodiment of the present invention provides a computer-readable storage medium storing a computer program thereon. When executed by a processor, this computer program can implement the steps of the power grid parameter identification method of the aforementioned embodiments. Based on this understanding, the technical solution of the aforementioned method of the present invention can be embodied in the form of a software product. This software product can be stored in a non-volatile storage medium (e.g., CD-ROM, USB flash drive, portable hard drive, etc.) and includes several instructions to cause a computer device (e.g., personal computer, server, or network device, etc.) to execute the steps of the method in various embodiments of the present invention.

[0054] Another embodiment of the present invention provides a dynamic identification system for grid sequence domain parameters based on converter ports, comprising: a port acquisition module for acquiring real-time port voltage and port current at the AC grid-connected port of the grid-connected converter, and converting the signals to positive, negative, and zero-sequence coordinate systems to improve identification accuracy under unbalanced voltage conditions; a dynamic synchronization reference system construction module for constructing a dynamic synchronization reference system based on the real-time frequency and phase information of the system; a sequence domain dynamic decomposition module for performing sequence domain dynamic decomposition on the three-phase voltage signal and the three-phase current signal; and a parameter adaptive dynamic update module for dynamically adjusting the parameter update process based on the predicted stability evaluation results. The equivalent model construction module is used to construct a Thevenin equivalent model of the external power grid for the grid-connected converter. The Thevenin equivalent model includes equivalent impedance and an equivalent voltage source. The equivalent impedance includes equivalent resistance and equivalent inductance. The parameter regression model construction module is used to construct a parameter regression model based on the Thevenin equivalent model, relating port voltage, port current, equivalent impedance, and equivalent voltage source. The parameters to be identified in the parameter regression model include the resistance value of the equivalent resistance, the inductance value of the equivalent inductance, and the voltage of the equivalent voltage source. The parameter identification module is used to employ an adaptive... A recursive least squares algorithm for the forgetting factor is used to identify the parameters of the parameter regression model online and output the latest estimated values ​​of each parameter to be identified for updating the regression model and online parameters in the next sampling period. An impedance calculation module is used to construct an overall port impedance model of the grid-connected converter's grid-connected port based on the equivalent resistance and equivalent inductance output by the parameter identification module. A state assessment module is used to calculate the short-circuit ratio based on the overall port impedance model and to determine the external power grid operating status based on the voltage components of the equivalent voltage source in positive sequence, negative sequence, and zero sequence.

[0055] Furthermore, the power grid parameter identification system may also include: an impedance calculation module for constructing the overall port impedance presented to the outside of the grid-connected converter's grid-connected port; and an evaluation module for calculating the short-circuit ratio index based on the overall port impedance and judging the external power grid operating status based on the external equivalent voltage source parameters.

[0056] Specifically, the port signal acquisition module is configured to perform the aforementioned step S1, the equivalent model construction module is configured to perform the aforementioned step S2, the parameter regression model construction module is configured to perform the aforementioned step S3, the parameter identification module is configured to perform the aforementioned step S4, the impedance calculation module is configured to perform the aforementioned step S5, and the evaluation module is configured to perform the aforementioned steps S6 and S7. That is, the specific composition and function of each module correspond to the content of the aforementioned steps and will not be repeated here.

[0057] This invention enables online identification of positive-sequence, negative-sequence, and zero-sequence parameters of the power grid. Considering that the grid voltage is usually dominated by the positive-sequence component under normal grid-connected operation conditions, and that the positive-sequence impedance plays a leading role in grid-connected system stability analysis, grid strength assessment, and short-circuit ratio calculation, this embodiment mainly uses the positive-sequence parameter identification results as the illustrative object to describe the technical solution of this invention.

[0058] The following specific embodiment illustrates the effectiveness of the present invention.

[0059] The system is simulated in MATLAB / Simulink. The grid-connected converter is not limited to either grid-based or grid-connected control strategies. An LC filter is used as an example, connected to a voltage source grid after passing through the line impedance. The simulation covers the following three grid conditions: The first type is the positive-sequence theoretical grid resistance. R line and inductor L line The voltage varies from 2Ω and 8mH to 1Ω and 4mH respectively; the second type is the d-axis component of the theoretical grid voltage. u gd and q-axis components u gq The voltage drops were calculated as follows: 1) the voltage changed from 282V and -130V to 267V and -124.5V respectively (a 5% drop in grid voltage); 2) the voltage dropped in the C-phase of the grid by 40%; 3) the voltage dropped in the C-phase of the grid by 40%; 4) the voltage dropped in the C-phase of the grid by 40%; 5) the voltage dropped in the C-phase of the grid by 40%; 6) the voltage dropped in the C-phase of the grid by 40%; 7) the voltage dropped in the C-phase of the grid by 40%; 8) the voltage dropped in the C-phase of the grid by 40%; 9 ...10 R line and inductor L line The values ​​are 0.5Ω and 1mH, respectively.

[0060] The actual measurement results for the first type of power grid condition are as follows: Figure 4 As shown, (a) shows the resistance values ​​in the positive sequence port impedance. R line The estimated curve, (b) shows the inductance value in the positive sequence port impedance. L line The estimated curve, (c) shows the d-axis component of the grid voltage. u gd The estimated curve, (d) shows the q-axis component of the grid voltage. u gq Estimating the curve. From... Figure 4 It can be seen that before and after the change in line impedance, the identification results of the line impedance are basically consistent with the actual values, and the identification results of the grid voltage are also basically consistent with the actual values, indicating that the identification results have high accuracy.

[0061] The actual measurement results for the second type of power grid condition are as follows: Figure 5 As shown, similarly, its (a) figure shows the resistance values ​​in the positive sequence port impedance. R line The estimated curve, (b) shows the inductance value in the positive sequence port impedance. L line The estimated curve, (c) shows the d-axis component of the grid voltage. u gd The estimated curve, (d) shows the q-axis component of the grid voltage. u gq Estimating the curve. From... Figure 5 It can be seen that the identification results of the grid voltage can accurately track the grid voltage before and after the voltage step, and the resistance and inductance values ​​of the identified line impedance are basically consistent with the actual values ​​before and after the voltage step. The identification results of the line impedance are not affected by the voltage step, which proves the effectiveness of the proposed method.

[0062] The actual measurement results for the third type of power grid condition are as follows: Figure 6 As shown, similarly, its (a) figure shows the resistance values ​​in the positive sequence port impedance. R line The estimated curve, (b) shows the inductance value in the positive sequence port impedance. L line The estimated curve, (c) shows the d-axis component of the grid voltage. u gd The estimated curve, (d) shows the q-axis component of the grid voltage. u gq Estimating the curve. From... Figure 6 It can be seen that before and after the voltage of phase c of the power grid drops by 40%, the identification result of the recursive least squares parameter identification algorithm with adaptive forgetting factor can accurately track the power grid voltage. Moreover, the resistance and inductance values ​​of the identified line impedance are basically consistent with the actual values ​​before and after the voltage step, and the identification result of the line impedance is not affected by the voltage fault. However, the traditional recursive least squares parameter identification algorithm based on forgetting factor has serious deviations from the actual values ​​of the resistance and inductance values ​​of the identified line impedance and the identified power grid voltage value after the voltage fault occurs, which proves the effectiveness of the proposed method.

[0063] It is evident that the present invention achieves the following objectives: Specifically, the present invention aims to achieve the following technical objectives: 1. Under the condition of only collecting the voltage and current of the converter's grid-connected port, it is not necessary to obtain the grid topology, line parameters and remote measurement information, so as to realize the online identification of the overall port impedance of the converter. 2. Under the condition of power grid imbalance, by modeling positive sequence, negative sequence and zero sequence dynamic components, stable online prediction of power grid parameters and equivalent voltage source parameters can be achieved, thereby improving the accuracy and robustness of parameter identification under the complex condition of power grid imbalance. 3. By identifying the overall port impedance, the short-circuit ratio at the grid connection point is further calculated, providing a basis for the stability analysis, control parameter tuning, and operation strategy selection of the converter; 4. Based on the overall port impedance identification results, the external power grid is equivalent to the Thevenin equivalent model of the series port impedance of the voltage source, so as to predict the voltage of the external equivalent voltage source without increasing the number of measurement points; 5. The predicted equivalent voltage source voltage and its variation characteristics are used to determine the operating status and stability of the remote power grid, thereby improving the converter's ability to sense changes in the power grid operating environment.

[0064] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, several equivalent substitutions or obvious modifications can be made without departing from the concept of the present invention, and all such modifications, achieving the same performance or purpose, should be considered within the scope of protection of the present invention.

Claims

1. A method for dynamic identification of power grid sequence domain parameters based on converter ports, characterized in that, Includes the following steps: S1. Obtain the real-time port voltage and port current at the AC grid-connected port of the grid-connected converter, and convert them into positive sequence, negative sequence and zero sequence components to overcome the influence of unbalanced voltage on port impedance identification. S2. The external power grid of the grid-connected converter is equivalent to a Thevenin equivalent model that includes equivalent impedances under positive sequence, negative sequence and zero sequence and equivalent voltage sources under positive sequence, negative sequence and zero sequence. The equivalent impedances include equivalent resistance and equivalent inductance. The equivalent voltage sources are represented under the real-time angular frequency and phase angle conditions of the system to adapt to the frequency fluctuations of the power grid. S3. Based on the Thevenin equivalent model, construct a parameter regression model between port voltage, port current, equivalent impedance and equivalent voltage source under positive sequence, negative sequence and zero sequence. The resistance value of the equivalent resistor, the inductance value of the equivalent inductor and the dq component of the equivalent voltage source are used as parameters to be identified and identified independently under positive sequence, negative sequence and zero sequence coordinates to eliminate the influence of unbalanced voltage on the identification accuracy. S4. The recursive least squares parameter identification algorithm with an adaptive forgetting factor is used to identify the parameters of the regression model online, so that the identification process can adapt to changes in system operating conditions and output the latest estimated values ​​of each parameter to be identified, which are used for the regression model update and online parameter update in the next sampling period. The parameter identification is performed in positive, negative, and zero order. The real-time angular frequency and phase angle of the system are used to update the regression model. The adaptive forgetting factor is dynamically adjusted according to the changes in system operating conditions to improve robustness to unbalanced voltage and frequency fluctuations.

2. The method for dynamic identification of power grid sequence domain parameters as described in claim 1, characterized in that, Step S1 includes: real-time acquisition of three-phase voltage and three-phase current signals at the AC grid-connected port of the grid-connected converter and conversion to the dq coordinate system corresponding to positive sequence, negative sequence and zero sequence respectively, to obtain the d-axis and q-axis components of the port voltage in positive sequence, negative sequence and zero sequence, and the d-axis and q-axis components of the port current in positive sequence, negative sequence and zero sequence.

3. The method for dynamic identification of power grid sequence domain parameters as described in claim 2, characterized in that, Step S3 includes: constructing the parameter regression model in the dq coordinate system corresponding to the positive sequence, negative sequence and zero sequence, wherein the voltage of the equivalent voltage source in the parameters to be identified in the positive sequence, negative sequence and zero sequence includes the d-axis voltage component and the q-axis voltage component; Step S3, the specific steps for constructing the parameter regression model, include: according to the port voltage equation in the dq coordinate system, expressing the d and q axis components of the port voltage in the positive, negative, and zero sequence as linear combinations of the d and q axis components of the port current in the positive, negative, and zero sequence and their differential terms with the parameters to be identified in the positive, negative, and zero sequence, thus obtaining the parameter regression model; The parameter regression model includes observation vectors, regression matrices, and parameter estimation vectors in positive, negative, and zero sequences. The observation vectors contain the d-axis and q-axis components of the port voltage in positive, negative, and zero sequences. The regression matrix contains the d-axis and q-axis components of the port current in positive, negative, and zero sequences, as well as their differential terms. The parameter estimation vectors contain the resistance values ​​of the equivalent resistance, the inductance values ​​of the equivalent inductance, and the equivalent voltage source in positive, negative, and zero sequences. The model is updated online during parameter identification using an adaptive forgetting factor to adapt to changes in system operating conditions.

4. The method for dynamic identification of power grid sequence domain parameters as described in claim 3, characterized in that, The mathematical expression for the parametric regression model is: ； in: These represent the positive-order, negative-order, and zero-order dynamic components, respectively; Y m It consists of the d-axis components of the grid-connected converter port voltage in positive sequence, negative sequence, and zero sequence. u m,od and q-axis components u m,oq The observation vector formed; Φ m The regression matrix Φ is in positive, negative, and zero order. m T It is Φ m The transpose of, in which i m,od , i m,oq These represent the d-axis and q-axis components of the grid-connected converter port current in positive sequence, negative sequence, and zero sequence, respectively. ω It is the system's real-time angular frequency. yes i m,od Differential terms in positive, negative, and zero order yes i m,oq Differential terms in positive, negative, and zero order; Θ m It is the resistance value of the equivalent resistance described in positive sequence, negative sequence, and zero sequence. R m,line The inductance values ​​of the equivalent inductance in positive sequence, negative sequence, and zero sequence. L m,line and the d-axis voltage components of the equivalent voltage source in positive sequence, negative sequence, and zero sequence. u m,gd and q-axis voltage component u m,gq The parameter estimation vector is formed.

5. The method for dynamic identification of power grid sequence domain parameters as described in claim 4, characterized in that, Step S4 includes: solving the parameter regression model using a recursive least squares method with an adaptive forgetting factor, and simultaneously outputting the resistance value of the equivalent resistance, the inductance value of the equivalent inductance, and the d-axis voltage component and q-axis voltage component of the equivalent voltage source in the positive, negative, and zero sequences.

6. The method for dynamic identification of power grid sequence domain parameters as described in claim 5, characterized in that, Step S4 further includes: to facilitate online parameter identification using the recursive least squares method, discretizing the parameter regression model, and setting Y... m =Φ m T Θ m Convert to the following discrete form: ； in, k Indicates the discrete sampling time sequence number. k =1,2,3,…;Y m ( k ) is the first k The observation vector at each sampling time, Φ m ( k ) is the first k The regression matrix at each sampling time, Θ m ( k ) is the first k The parameter estimation vector at each sampling time point; The formula for calculating the recursive least squares method with an adaptive forgetting factor is as follows: ; in, Indicates the first k The parameter estimation vector at each sampling time point, K m ( k ) is the gain vector, ε m ( k P is the prediction error vector. m ( k Let be the covariance matrix, and I be the identity matrix of the corresponding dimension; λ m (k) is the adaptive forgetting factor, with a value range of 0 < λ m (k)<1, β This is a stability adjustment coefficient, used to adjust the predicted stability index. S m (k) Incorporate gain calculation S m (k) is the predictive stability evaluation index, defined as follows: ； in: For predicting stability evaluation indicators; For parameter prediction error; This serves as the baseline value for error normalization. For the dynamic fluctuation of parameters; This serves as the baseline value for parameter normalization. , , Let be the weighting coefficient, and satisfy: ； To achieve dynamic prediction and online updating of parameters under positive, negative, and zero order, an adaptive forgetting factor is introduced. λ m Design of (k), adaptive forgetting factor λ m (k) Based on the predicted stability index The dynamic adjustment is calculated as follows: ; in, λ m,min and λ m,max These are the lower and upper limits of the forgetting factor, respectively, and their values ​​satisfy 0 < λ m,min < λ m,max ≤1; As a predictive stability evaluation index, it consists of the prediction error of each order component and the parameter fluctuation. γ The adjustment coefficient is a positive real number, usually in the range of 0.1 to 5. It is adjusted according to the system power level, sampling period and dynamic response requirements to balance the stability under normal operating conditions and the dynamic tracking capability under fault conditions. Methods for dynamic prediction and updating of power grid parameters based on the order domain include: The port voltage and port current of the grid-connected converter are sampled in real time, and the initial parameters of the algorithm and the forgetting factor λ are set. Among them, the initial values ​​P of the covariance matrix under positive sequence, negative sequence and zero sequence are... m (0) Initialize as a diagonal matrix, with diagonal elements taking values ​​in the range of 10. 2 ~10 6 This reflects the high degree of uncertainty in parameter estimation at the initial time; the initial values ​​Θ of the parameter estimation vectors in positive, negative, and zero sequences. m (0) Based on empirical values, either initialize it as a zero vector or as an initial value vector within the physically feasible range of each parameter to be identified, to ensure the stability and convergence of the parameter estimation process; in the first... k At each sampling time, based on the port voltage and port current obtained from the current sampling, observation vectors Y are constructed in positive sequence, negative sequence, and zero sequence. m ( k ) and regression matrix Φ m ( k And compare it with the parameter estimates Θ in the positive, negative and zero sequences of the previous sampling time. m ( k -1) and covariance matrix P m ( k Substituting -1) into the calculation formula of the recursive least squares method with adaptive forgetting factor, we obtain the current time step. k Parameter estimation results Θ in positive order, negative order and zero order m ( k And the covariance matrix P in positive, negative, and zero order. m ( k ); where Θ m ( k This includes the resistance values ​​of the equivalent resistance under positive sequence, negative sequence, and zero sequence, the inductance values ​​of the equivalent inductance under positive sequence, negative sequence, and zero sequence, and the d-axis voltage components and q-axis voltage components of the equivalent voltage source under positive sequence, negative sequence, and zero sequence. Subsequently, the above sampling, construction, and recursive update process is repeated at the next sampling time to realize the online identification of the equivalent impedance parameters and the equivalent voltage source parameters under positive sequence, negative sequence, and zero sequence.

7. The method for dynamic identification of power grid sequence domain parameters as described in claim 1, characterized in that, It also includes the following steps: S5. Based on the resistance values ​​of the equivalent resistances in the positive sequence, negative sequence, and zero sequence obtained from parameter identification, and the inductance values ​​of the equivalent inductances in the positive sequence, negative sequence, and zero sequence, construct the overall port impedance model presented to the outside of the grid-connected converter's grid-connected port. S6. Calculate the short-circuit ratio at the grid connection point of the grid-connected converter based on the overall port impedance model. The short-circuit ratio calculation is mainly based on the positive sequence equivalent impedance and the positive sequence voltage amplitude. The negative sequence and zero sequence parameters are used to assist in judging the voltage imbalance or fault state. S7. Based on the d-axis voltage components and q-axis voltage components of the equivalent voltage source output in step S4 under positive sequence, negative sequence and zero sequence, the operating status of the external power grid is judged, including whether there is voltage drop, voltage fluctuation, power grid strength change or voltage imbalance fault, and the basis is provided for the converter to take corresponding control strategies or operating modes.

8. The method for dynamic identification of power grid sequence domain parameters as described in claim 7, characterized in that, The overall port impedance model is as follows: ； in, R m,line , L m,line These are the resistance values ​​of the equivalent resistance under positive sequence, negative sequence, and zero sequence conditions identified in step 4, and the inductance values ​​of the equivalent inductance under positive sequence, negative sequence, and zero sequence conditions, respectively. ω m (k) represents the real-time angular frequency of the system at the current moment in positive, negative, and zero order. j The imaginary unit; real-time angular frequency of the system ω m Under (k), the magnitude of the overall port impedance model is expressed as: ； The overall port impedance model is used to characterize the overall electrical characteristics of the external power grid visible to the grid-connected converter port, and serves as the basis for the short-circuit ratio evaluation in step S6. Step S6 calculates the short-circuit ratio index, specifically including: Calculate the short-circuit capacity at the grid connection point S sc : ； in, U g The effective value of the grid voltage at the grid connection point; Combined with the rated capacity of the grid-connected converter S n The short-circuit ratio (SCR) is obtained as follows: ； This enables a quantitative assessment of the strength and weakness characteristics of the external power grid.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program can implement the steps of the method according to any one of claims 1-8.

10. A dynamic identification system for power grid sequence domain parameters based on converter ports, characterized in that, include: The port acquisition module is used to acquire the real-time port voltage and port current at the AC grid-connected port of the grid-connected converter, and convert the signals to positive sequence, negative sequence, and zero sequence coordinate systems to improve the identification accuracy under unbalanced voltage conditions. The dynamic synchronization reference frame construction module is used to construct a dynamic synchronization reference frame based on the system's real-time frequency and real-time phase information. The sequence domain dynamic decomposition module is used to perform sequence domain dynamic decomposition on three-phase voltage signals and three-phase current signals. The parameter adaptive dynamic update module is used to dynamically adjust the parameter update process based on the prediction stability evaluation results; The equivalent model construction module is used to construct the Thevenin equivalent model of the external power grid of the grid-connected converter. The Thevenin equivalent model includes equivalent impedance and equivalent voltage source. The equivalent impedance includes equivalent resistance and equivalent inductance. The parameter regression model construction module is used to construct a parameter regression model between port voltage, port current, equivalent impedance and equivalent voltage source based on the Thevenin equivalent model; wherein, the parameters to be identified in the parameter regression model include: the resistance value of the equivalent resistor, the inductance value of the equivalent inductor and the voltage of the equivalent voltage source; The parameter identification module is used to perform online parameter identification on the parameter regression model using a recursive least squares algorithm with an adaptive forgetting factor, and output the latest estimated values ​​of each parameter to be identified for the regression model update and online parameter update in the next sampling period. The impedance calculation module is used to construct an overall port impedance model of the grid-connected converter's grid-connected port based on the equivalent resistance and equivalent inductance output by the parameter identification module. The status assessment module is used to calculate the short-circuit ratio index based on the overall port impedance model, and to determine the operating status of the external power grid based on the voltage components of the equivalent voltage source in positive sequence, negative sequence and zero sequence.

Images