Subsynchronous oscillation suppression and parameter optimization method and system based on network construction type energy storage

Abstract

The invention relates to a subsynchronous oscillation suppression and parameter optimization method and system based on network construction type energy storage, and the method comprises the steps: carrying out the analysis of a system oscillation mode through building a small-signal state space model of an optical storage grid-connected system comprising a network-following type photovoltaic inverter and a network construction type energy storage converter, and recognizing a subsynchronous oscillation mode; on the basis, a damping path reconstruction method is introduced, and separation and quantitative characterization are carried out on interaction damping between the grid-forming type energy storage converter and the grid-following type photovoltaic inverter. And further, equivalent positive damping provided by the interaction damping in a subsynchronous frequency range is taken as a parameter optimization target, and optimal configuration is carried out on virtual synchronous machine control parameters of the network-forming type energy storage converter on the premise of meeting a system stability constraint condition. The system is used for realizing the method, and the optimized control parameters are applied to the network-forming type energy storage converter control system so as to suppress the subsynchronous oscillation of the optical storage grid-connected system.

Description

Technical Field

[0001] This invention relates to the fields of power system stability control and power electronics technology, specifically to a method and system for suppressing and optimizing subsynchronous oscillations based on grid-type energy storage, and a system for tuning the virtual synchronous machine control parameters of the energy storage converter. Background Technology

[0002] With the development of the "high-efficiency and high-value" trend in the power system, the proportion of new energy units such as wind and solar power in the power system is constantly increasing. Solar energy is gradually evolving from a supplementary energy source to a substitute energy source, and photovoltaic power generation technology has been widely used both domestically and internationally. As a high proportion of new energy is connected to the grid, the overall equivalent impedance level of the system decreases. Under certain operating conditions and specific control parameter configurations, new energy grid-connected systems are prone to exhibiting weak grid characteristics, which can lead to dynamic interaction problems between the inverter and the grid impedance. This may result in stability risks such as subsynchronous oscillations, posing new challenges to the safe and stable operation of traditional grid-connected new energy units.

[0003] To enhance the absorption capacity of new energy sources and the stability of power grid operation, the National Energy Administration has issued several policies in recent years to encourage the configuration of energy storage devices in new energy power generation systems. Against this backdrop, grid-based energy storage technology, due to its independence from phase-locked loops, ability to simulate the external characteristics of synchronous generators, and active participation in grid voltage and frequency regulation, is considered to have good stability support capabilities and stabilization potential under weak grid conditions. However, existing applications of grid-based energy storage primarily focus on improving the overall stability of the system. A clear and directly applicable analytical method for determining the specific damping path, mechanism, and quantitative relationship between grid-connected solar-energy storage systems and subsynchronous oscillations is still lacking.

[0004] Currently, research on oscillation problems in renewable energy grid-connected systems mainly employs impedance analysis and eigenvalue analysis. Impedance analysis typically establishes a small-signal frequency-domain impedance model of the power electronic devices and uses the Nyquist criterion or its extensions to assess system stability. Eigenvalue analysis linearizes the system near a specific steady-state operating point and obtains the frequency and damping characteristics of oscillation modes by analyzing the system's eigenvalues ​​and their corresponding eigenvectors. However, while these methods can be used for stability assessment, they struggle to reveal the subsynchronous interaction paths between different grid-connected devices from the perspective of engineering parameter tuning, and they also find it difficult to quantitatively characterize the specific damping contribution of grid-connected energy storage in suppressing photovoltaic subsynchronous oscillations. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a method for suppressing and optimizing subsynchronous oscillations based on grid-connected energy storage. This method can accurately quantify the damping effect path of grid-connected energy storage on grid-connected photovoltaic oscillations, optimize energy storage control parameters, and effectively suppress subsynchronous oscillations. The technical solution is as follows: A method for suppressing and optimizing subsynchronous oscillations based on grid-type energy storage is provided, including the following steps: (1) Establish a small-signal state-space model of a photovoltaic-storage grid-connected system that includes grid-connected photovoltaic inverters and grid-connected energy storage converters; (2) Perform eigenvalue analysis on the small-signal state-space model to identify subsynchronous oscillation modes in the system whose oscillation frequency is within the subsynchronous frequency range; (3) Based on the damping path analysis method, the subsynchronous oscillation mode is damped and reconstructed to separate the interaction damping path between the grid-type energy storage converter and the grid-connected photovoltaic inverter, and obtain the dynamic transmission relationship corresponding to the interaction damping path. (4) Taking the magnitude of the positive damping provided by the interaction damping path in the subsynchronous frequency range as the optimization objective, under the premise of satisfying the system stability constraints, the Bayesian optimization algorithm is used to optimize and solve the virtual synchronous machine control parameters of the grid-type energy storage converter to obtain a set of optimal control parameters; (5) The optimal control parameters are configured into the control system of the grid-type energy storage converter to change the dynamic response characteristics of the grid-type energy storage converter to the power disturbance of the common coupling point, thereby providing equivalent positive damping to the photovoltaic system within the subsynchronous frequency range to suppress the subsynchronous oscillation of the photovoltaic-energy storage grid-connected system.

[0006] Further, in step (1), the photovoltaic system is a grid-connected photovoltaic system, and the photovoltaic inverter adopts a control structure of DC voltage outer loop and current inner loop; the energy storage system is a grid-connected energy storage system, and the grid-connected energy storage converter adopts a virtual synchronous machine control structure; the photovoltaic-energy storage grid-connected system adopts an AC side grid-connected mode.

[0007] Furthermore, the establishment of the small-signal state-space model in step (1) includes the following sub-steps: (1) Establish nonlinear differential equation models for grid-connected photovoltaic inverters, grid-connected energy storage converters and their corresponding control systems, filters and power grid lines respectively; (2) At the preset steady-state operating point, the nonlinear differential equation model is linearized to obtain the small-signal model of each component; (3) Based on the electrical connection relationship and control signal interaction relationship between the components, the small signal models are interconnected and intermediate algebraic variables are eliminated to form the global small signal state space model of the photovoltaic-storage grid-connected system.

[0008] Furthermore, the identification of subsynchronous oscillation modes in step (2) includes: The state matrix in the small-signal state-space model is decomposed into eigenvalues. The oscillation frequency of the oscillation mode is determined based on the imaginary part of the eigenvalues, and the stability of the oscillation mode is determined based on the real part of the eigenvalues ​​or the corresponding damping ratio. The oscillation mode whose oscillation frequency is within the subsynchronous frequency range is identified as the subsynchronous oscillation mode.

[0009] Furthermore, the damping path mentioned in step (3) is the signal path formed by the transmission and interaction of disturbances between different devices during the closed-loop dynamic process of the system, and the damping reconstruction includes: Taking the subsynchronous oscillation mode dominated by the DC-side energy storage element of the photovoltaic inverter as the analysis object, the total system damping is decomposed into: (1) The first type of damping path formed by the internal dynamic links of the photovoltaic inverter itself; (2) The second type of damping path formed by the interaction between the photovoltaic inverter and the AC power grid; (3) The third type of damping path formed by the interaction between the photovoltaic inverter and the grid-type energy storage converter; The third type of damping path is used to characterize the damping contribution of the grid-type energy storage converter to the subsynchronous oscillation mode.

[0010] Furthermore, the optimization using the Bayesian optimization algorithm described in step (4) includes: (1) Construct an objective function to characterize the equivalent damping provided by the third type of damping path in the subsynchronous frequency range, and set the range of values ​​for the virtual synchronizing machine control parameters and system stability constraints; (2) Select multiple initial parameter sample points within the design space of the control parameters, calculate the corresponding objective function values, and form an initial training dataset; (3) Construct a Gaussian process regression surrogate model based on the initial training dataset; (4) Based on the agent model and the preset acquisition function, iteratively evaluate the control parameters of the virtual synchronizer and update the agent model; (5) Under the premise of satisfying the system stability constraints, select the control parameter with the optimal objective function value from all evaluated virtual synchronous machine control parameters as the optimal control parameter.

[0011] Furthermore, the virtual synchronous machine control parameters include the moment of inertia parameter J and the damping coefficient parameter D. The objective function is used to characterize the influence relationship between the moment of inertia parameter J and the damping coefficient parameter D on the equivalent damping magnitude of the third type of damping path, and the parameter optimization process must satisfy the preset parameter boundary constraints.

[0012] Furthermore, before evaluating the control parameters of the virtual synchronizer, the stability of other dominant oscillation modes, excluding the subsynchronous oscillation mode, is checked to ensure that the damping ratio of the other dominant oscillation modes is higher than the preset safety threshold, so as to avoid introducing new low-damped oscillation modes during the parameter optimization process.

[0013] On the other hand, a parameter optimization system for a grid-type energy storage converter is provided, including: The model building module is used to build a small-signal state-space model of the photovoltaic-storage grid-connected system; An oscillation identification module is used to identify the subsynchronous oscillation mode in the system based on the small-signal state-space model. The mechanism analysis module is used to reconstruct the damping path of the subsynchronous oscillation mode and output the interaction damping contribution information corresponding to the grid-type energy storage converter. The parameter optimization module is used to solve the optimal virtual synchronous machine control parameters of the grid-type energy storage converter based on the interaction damping contribution information and using a Bayesian optimization algorithm. The parameter configuration module is used to write the optimal virtual synchronous machine control parameters into the control system of the grid-type energy storage converter to suppress the subsynchronous oscillation of the photovoltaic-storage grid-connected system.

[0014] Furthermore, the parameter configuration module is configured as follows: Before the photovoltaic-storage grid-connected system is put into operation, the optimal virtual synchronous machine control parameters are written into the control system of the grid-type energy storage converter through offline parameter tuning. Alternatively, during the operation of the photovoltaic-storage grid-connected system, under the premise of meeting the system stability constraints, the virtual synchronous machine control parameters of the grid-connected energy storage converter can be adjusted through online parameter updates.

[0015] The beneficial effects of the technical solution provided by the embodiments of the present invention are as follows: This invention provides a method for suppressing and optimizing subsynchronous oscillations based on grid-connected energy storage. By constructing a small-signal state-space model of a photovoltaic-energy storage grid-connected system and identifying subsynchronous oscillation modes in the system, a damping path reconstruction analysis method is introduced to separate and quantify the damping of the interaction between the grid-connected energy storage converter and the grid-connected photovoltaic inverter. This clarifies the damping path and degree of participation of grid-connected energy storage in suppressing photovoltaic subsynchronous oscillations, providing a clear physical basis for subsequent control parameter adjustment.

[0016] Furthermore, this invention uses the equivalent positive damping provided by the interactive damping within the subsynchronous frequency range as the parameter optimization target. Under the premise of satisfying the overall system stability constraints, it performs targeted optimization of the virtual synchronous machine control parameters of the grid-type energy storage converter, so that the control parameter tuning process no longer relies on experience adjustment or trial adjustment, which is conducive to improving the pertinence, rationality and engineering feasibility of parameter tuning.

[0017] Furthermore, this invention combines mechanistic analysis methods with parameter optimization processes, thereby improving the ability to suppress subsynchronous oscillations in photovoltaic-storage grid-connected systems without altering the photovoltaic system's control structure or relying on additional hardware configurations. This helps enhance the application effect and engineering promotion value of grid-connected energy storage in the stable control of new energy grid connection under weak grid conditions. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 The overall flowchart of the subsynchronous oscillation suppression and parameter optimization method based on grid-type energy storage provided in the embodiments of the present invention is shown below. Figure 2 This is a schematic diagram of the structure of the grid-connected photovoltaic-storage system according to an embodiment of the present invention; Figure 3 This is a block diagram of the closed-loop transfer function of the grid-connected photovoltaic-storage system according to an embodiment of the present invention; Figure 4 This is a schematic diagram of the damping path of the grid-connected photovoltaic-storage system according to an embodiment of the present invention; Figure 5 The flowchart of the Bayesian optimization algorithm described in the embodiment of the present invention is shown below; Figure 6 The DC capacitor voltage U before and after the change of control parameters of the energy storage inverter described in this embodiment of the invention. dc Schematic diagram of time-domain simulation. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0021] This embodiment provides a method for suppressing subsynchronous oscillations and optimizing parameters based on grid-type energy storage.

[0022] This method takes the grid-type energy storage converter as the regulation object, models and analyzes the dynamic characteristics of the photovoltaic-energy storage grid-connected system, identifies the subsynchronous oscillation mode in the system, and optimizes the virtual synchronous machine control parameters of the grid-type energy storage converter so that the grid-type energy storage can provide equivalent positive damping to the photovoltaic system within the subsynchronous frequency range, thereby suppressing the subsynchronous oscillation of the photovoltaic-energy storage grid-connected system.

[0023] Figure 1 The overall implementation process of this method is as follows: (1) Based on Figure 2 Based on the system topology shown, a mathematical model describing the dynamic characteristics of the system is established.

[0024] In this embodiment, the photovoltaic system adopts a grid-following control method, and the photovoltaic inverter adopts a control structure that combines a DC voltage outer loop and a current inner loop; the energy storage system adopts a grid-connected control method, and the grid-connected energy storage converter adopts a virtual synchronous machine control structure. The photovoltaic-energy storage system is connected to the grid through AC side grid connection.

[0025] An electromagnetic transient simulation model of the photovoltaic-storage system was built in PSCAD / EMTDC. In MATLAB / Simulink, the system was subdivided into different modules. Near the system's steady-state operating point, the nonlinear differential equations were linearized, and connections were established using a linkage matrix to obtain the system's small-signal state-space model. The photovoltaic DC-side voltage reference value was stepped to 1.05U. dcref0 The accuracy of the small-signal model is verified by comparing the simulation results of the time-domain simulation model and the small-signal model. (2) In MATLAB / Simulink, the standard global state-space equations of the system are formed, as shown in equation (1): (1) Where x is the system state variable vector, u is the input vector, y is the output vector, A is the system state matrix, B is the input matrix, C is the output matrix, and D is the direct transmission matrix.

[0026] Based on the established small-signal state-space model of the photovoltaic-storage grid-connected system, eigenvalue analysis is performed on the system state matrix to obtain the oscillation mode characteristics of the system near the operating point. This allows for the identification of subsynchronous oscillation modes whose oscillation frequencies fall within the subsynchronous frequency range. The obtained eigenvalues ​​are then used to further analyze the oscillations. real part and the virtual part Calculate the oscillation frequency of each mode. Damping ratio As shown in equations (2) and (3): (2) (3) Each eigenvalue describes an oscillation mode of the system, and the imaginary part determines the oscillation frequency of the mode. The corresponding damping is relative to how quickly the quantifiable mode decays.

[0027] Based on the obtained eigenvalue information, oscillation modes with oscillation frequencies within the subsynchronous frequency range are selected, and participation factor analysis is performed on these oscillation modes to determine the state variables and control links that play a dominant role in these subsynchronous oscillation modes, providing a basis for subsequent damping path reconstruction analysis.

[0028] (3) The dynamic equation of the photovoltaic DC capacitor is: (4) (5) (6) in, This represents the steady-state value of the photovoltaic DC bus voltage. , These represent the input power disturbance value and the output power disturbance value, respectively. This represents the DC bus voltage disturbance value. Indicates from arrive The open-loop transfer function.

[0029] In frequency Next, substitute Substitute equation (5) into equation (4), and then... dc (s) expands to the real part virtual part By adding together, equation (4) can be transformed into a second-order equation: (7) According to the Heffron-Phillips model of the synchronous machine rotor, as shown in equation (8), it can be seen that the dynamic equation (7) of the photovoltaic DC capacitor has the same form as equation (8).

[0030] (8) Where J represents the equivalent rotational inertia of the system, and D represents the equivalent damping coefficient of the system, in the second-order equation The equivalent damping coefficient D of the corresponding component is used as an indicator to judge the stability of the system. When D < 0, the system is unstable, and in this synchronous oscillation mode, it exhibits negative damping, which manifests as oscillation decay.

[0031] Based on the subsynchronous oscillation mode dominated by the photovoltaic DC capacitor link, to analyze the damping effect of different devices on this oscillation mode, the grid-type photovoltaic-storage system is divided into three parts: the photovoltaic system, the energy storage system, and the AC grid. A closed-loop transfer function block diagram of the system is then established, as follows: Figure 3 As shown.

[0032] (4) The closed-loop transfer function block diagram of the entire system is derived, with the DC capacitor voltage disturbance value of the grid-connected photovoltaic inverter as an example. As an input, the inverter power disturbance value As an output, the three damping paths within the system are separated. ,like Figure 4 As shown above, the damping coefficient corresponding to the DC capacitor subsynchronous mode is... The total system damping can be expressed as a linear superposition of the damping of three paths, thus obtaining the damping of each path. The first type of damping path is composed of the internal dynamic elements of the photovoltaic inverter itself, G. a (s) reflects the damping effect within the photovoltaic inverter itself. The second type of damping path is formed by the interaction between the photovoltaic inverter and the AC grid. G b (s) reflects the damping of the interaction between the photovoltaic inverter and the AC grid. The third type of damping path is formed by the interaction between the photovoltaic inverter and the grid-connected energy storage converter. G c (s) reflects the damping of the interaction between the photovoltaic inverter and the grid-connected energy storage converter, and its corresponding damping coefficient can be set as K. dc3 Its expression is: (9) (5) In order to guide the control parameters of the grid-type energy storage converter to be adjusted in the direction of improving the subsynchronous oscillation damping capability, the equivalent positive damping provided by the third type of damping path in the subsynchronous frequency range is used as the optimization target. Under the premise of ensuring the overall stability of the system, the corresponding objective function and constraint conditions are constructed as shown in Equations (10) and (11): (10) (11) Where f is the objective function, equation (11) is the control variable constraint, and G c (s) reflects the damping of the interaction between the photovoltaic inverter and the grid-connected energy storage converter. This indicates the range of subsynchronous oscillation frequencies of the system.

[0033] At the same time, it is necessary to first verify whether the damping ratios of other dominant oscillation modes are all higher than the preset safety threshold.

[0034] After determining the objective function and constraints, the following steps are taken: Figure 5 The Bayesian optimization algorithm shown uses the control parameters of the energy storage converter, namely the moment of inertia J and the damping coefficient D, as optimization design variables, forming a two-dimensional vector [J, D]. Ten initial points are uniformly selected within the design space using the Latin hypercube sampling method to form an initial sample set. For the i-th initial point, [J] i D i Substitute the small-signal model of the photovoltaic-storage system into the eigenvalue analysis method, update the damping coefficient in the third type of damping path of the system, calculate the objective function value at the point according to the objective function, and verify it based on the constraint conditions. If the constraint conditions are violated, a very low penalty value is assigned to obtain the initial dataset. Based on the initial dataset mentioned above, Gaussian process regression is used as a surrogate model for the objective function. The overall shape of the objective function is learned by analyzing the relationship between existing data points. The maximum number of evaluations is set to 100. The expected improvement function is used as the acquisition function in this paper. Based on the surrogate model and the acquisition function, the next candidate point of the control parameter to be evaluated is determined. This point is substituted into the objective function and constraints for calculation and evaluation. Based on the evaluation structure, the dataset and surrogate model are updated. Through the above parameter evaluation and model update process, under the premise of satisfying the system stability constraints, the parameter combination with the optimal objective function value is selected from all evaluated virtual synchronous machine control parameters as the optimal control parameters of the grid-type energy storage converter [J1, D1].

[0035] (6) To verify the vibration suppression effect of parameter optimization, the obtained optimal virtual synchronous machine control parameters were configured into the control system of the grid-connected energy storage converter through offline parameter tuning. Based on the electromagnetic transient simulation model established by PSCAD / EMTDC, the subsynchronous oscillation response of the photovoltaic-storage grid-connected system before and after parameter optimization was compared and analyzed. Subsynchronous oscillation was excited at t=3s to obtain the comparison results of the photovoltaic-storage system under two operating conditions. Figure 6 The DC capacitor voltage U is defined by the initial control parameters [J0, D0] of the energy storage inverter. dc The time-domain simulation diagram and the DC capacitor voltage U when the control parameters [J1, D1] are optimized using an energy storage inverter. dc The time-domain simulation diagram shows that after 3 seconds, the optical storage system will experience subsynchronous oscillation. After using optimized parameters, the amplitude of the subsynchronous oscillation is reduced. This demonstrates that applying the optimized parameters to the system can suppress the subsynchronous oscillation, proving the effectiveness of the method of this invention.

[0036] In summary, this invention provides a method for suppressing and optimizing subsynchronous oscillations based on grid-connected energy storage. It reveals the damping path and participation degree of grid-connected energy storage in grid-connected photovoltaic oscillations. By taking the maximization of positive damping provided by the interaction damping path as the optimization objective, a Bayesian optimization algorithm is used to perform targeted optimization of energy storage control parameters, effectively improving the efficiency and accuracy of parameter optimization.

[0037] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. These modifications or substitutions do not cause the essence of the corresponding technical solutions to depart from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for suppressing and optimizing subsynchronous oscillations based on grid-type energy storage, characterized in that, Includes the following steps: (1) Establish a small-signal state-space model of a photovoltaic-storage grid-connected system that includes grid-connected photovoltaic inverters and grid-connected energy storage converters; (2) Perform eigenvalue analysis on the small-signal state-space model to identify subsynchronous oscillation modes in the system whose oscillation frequency is within the subsynchronous frequency range; (3) Based on the damping path analysis method, the subsynchronous oscillation mode is damped and reconstructed to separate the interaction damping path between the grid-type energy storage converter and the grid-connected photovoltaic inverter, and obtain the dynamic transmission relationship corresponding to the interaction damping path. (4) Taking the magnitude of the positive damping provided by the interaction damping path in the subsynchronous frequency range as the optimization objective, under the premise of satisfying the system stability constraints, the Bayesian optimization algorithm is used to optimize and solve the virtual synchronous machine control parameters of the grid-type energy storage converter to obtain a set of optimal control parameters; (5) The optimal control parameters are configured into the control system of the grid-type energy storage converter to change the dynamic response characteristics of the grid-type energy storage converter to the power disturbance of the common coupling point, thereby providing equivalent positive damping to the photovoltaic system within the subsynchronous frequency range to suppress the subsynchronous oscillation of the photovoltaic-energy storage grid-connected system.

2. The method according to claim 1, characterized in that, In step (1), the photovoltaic system is a grid-connected photovoltaic system, and the photovoltaic inverter adopts a control structure of DC voltage outer loop and current inner loop; the energy storage system is a grid-connected energy storage system, and the grid-connected energy storage converter adopts a virtual synchronous machine control structure; the photovoltaic-storage grid-connected system adopts an AC side grid-connected mode.

3. The method according to claim 1, characterized in that, The establishment of the small-signal state-space model in step (1) includes the following sub-steps: (1) Establish nonlinear differential equation models for grid-connected photovoltaic inverters, grid-connected energy storage converters and their corresponding control systems, filters and power grid lines respectively; (2) At the preset steady-state operating point, the nonlinear differential equation model is linearized to obtain the small-signal model of each component; (3) Based on the electrical connection relationship and control signal interaction relationship between the components, the small signal models are interconnected and intermediate algebraic variables are eliminated to form the global small signal state space model of the photovoltaic-storage grid-connected system.

4. The method according to claim 1, characterized in that, The identification of subsynchronous oscillation modes in step (2) includes: The state matrix in the small-signal state-space model is decomposed into eigenvalues. The oscillation frequency of the oscillation mode is determined based on the imaginary part of the eigenvalues, and the stability of the oscillation mode is determined based on the real part of the eigenvalues ​​or the corresponding damping ratio. The oscillation mode whose oscillation frequency is within the subsynchronous frequency range is identified as the subsynchronous oscillation mode.

5. The method according to claim 1, characterized in that, The damping path mentioned in step (3) is the signal path formed by the transmission and interaction of disturbances between different devices during the closed-loop dynamic process of the system. The damping reconstruction includes: Taking the subsynchronous oscillation mode dominated by the DC-side energy storage element of the photovoltaic inverter as the analysis object, the total system damping is decomposed into: (1) The first type of damping path formed by the internal dynamic links of the photovoltaic inverter itself; (2) The second type of damping path formed by the interaction between the photovoltaic inverter and the AC power grid; (3) The third type of damping path formed by the interaction between the photovoltaic inverter and the grid-type energy storage converter; The third type of damping path is used to characterize the damping contribution of the grid-type energy storage converter to the subsynchronous oscillation mode.

6. The method according to claim 1, characterized in that, The optimization using the Bayesian optimization algorithm in step (4) includes: (1) Construct an objective function to characterize the equivalent damping provided by the third type of damping path in the subsynchronous frequency range, and set the range of values ​​for the virtual synchronizing machine control parameters and system stability constraints; (2) Select multiple initial parameter sample points within the design space of the control parameters, calculate the corresponding objective function values, and form an initial training dataset; (3) Construct a Gaussian process regression surrogate model based on the initial training dataset; (4) Based on the agent model and the preset acquisition function, iteratively evaluate the control parameters of the virtual synchronizer and update the agent model; (5) Under the premise of satisfying the system stability constraints, select the control parameter with the optimal objective function value from all evaluated virtual synchronous machine control parameters as the optimal control parameter.

7. The method according to claim 6, characterized in that, The virtual synchronous machine control parameters include rotational inertia parameter J and damping coefficient parameter D. The objective function is used to characterize the influence relationship between the rotational inertia parameter J and damping coefficient parameter D on the equivalent damping magnitude of the third type of damping path, and the parameter optimization process must meet the preset parameter boundary constraints.

8. The method according to claim 6, characterized in that, Before evaluating the control parameters of the virtual synchronizer, the stability of the other dominant oscillation modes, excluding the subsynchronous oscillation mode, is checked to ensure that the damping ratio of the other dominant oscillation modes is higher than the preset safety threshold, so as to avoid introducing new low-damped oscillation modes during the parameter optimization process.

9. A parameter optimization system for a grid-type energy storage converter, characterized in that, For performing the method according to any one of claims 1 to 8, comprising: The model building module is used to build a small-signal state-space model of the photovoltaic-storage grid-connected system; An oscillation identification module is used to identify the subsynchronous oscillation mode in the system based on the small-signal state-space model. The mechanism analysis module is used to reconstruct the damping path of the subsynchronous oscillation mode and output the interaction damping contribution information corresponding to the grid-type energy storage converter. The parameter optimization module is used to solve the optimal virtual synchronous machine control parameters of the grid-type energy storage converter based on the interaction damping contribution information and using a Bayesian optimization algorithm. The parameter configuration module is used to write the optimal virtual synchronous machine control parameters into the control system of the grid-type energy storage converter to suppress the subsynchronous oscillation of the photovoltaic-storage grid-connected system.

10. The system according to claim 9, characterized in that, The parameter configuration module is configured as follows: Before the photovoltaic-storage grid-connected system is put into operation, the optimal virtual synchronous machine control parameters are written into the control system of the grid-type energy storage converter through offline parameter tuning. Alternatively, during the operation of the photovoltaic-storage grid-connected system, under the premise of meeting the system stability constraints, the virtual synchronous machine control parameters of the grid-connected energy storage converter can be adjusted through online parameter updates.

Images