Virtual synchronous generator control-based energy storage converter grid connection method

Abstract

The invention relates to an energy storage converter grid-connected method based on virtual synchronous generator control. Compared with the prior art, the defects that in an existing energy storage converter grid-connected control technology, the response capacity to power grid disturbance is weak, the power control precision and stability are insufficient, and the robustness of a controller is poor are overcome. The method comprises the following steps: designing a self-adaptive rule of virtual inertia and a damping coefficient in an active loop of a VSG controller; self-adaptive adjustment is carried out; designing a VSG power decoupling controller based on hyper-local model cooperative control; smooth power following and effective decoupling control of active power and reactive power. According to the method, the inertia and damping characteristics of a synchronous generator are achieved in the operation process of the energy storage grid-connected converter, when grid-connected active power or grid-connected frequency is disturbed, frequency fluctuation can be restrained through self-adaptive adjustment of the virtual inertia and the damping coefficient, and the good power decoupling control capacity is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy grid-connected power generation, and specifically to a grid-connected method for an energy storage converter based on virtual synchronous generator control. Background Art

[0002] In recent years, China has made remarkable progress in distributed renewable energy fields such as wind power and photovoltaic power. The grid connection penetration rate of these energies has gradually increased, bringing the characteristics of volatility and intermittency of renewable energy power generation, which seriously threatens the safe and stable operation of the power system. Especially in the case of large load changes and significant fluctuations in renewable energy power generation, the frequency and voltage stability of the power grid face huge challenges.

[0003] Energy storage grid-connected operation interacts with the power grid through power electronic devices. Large-scale renewable energy power generation and energy storage grid-connected operation have gradually transformed the power grid into a low-inertia and underdamped system dominated by power electronic converters. This transformation poses challenges to traditional power system control methods, especially the ability to regulate frequency and power is significantly reduced. Therefore, the virtual synchronous generator (VSG) control, which simulates the characteristics of synchronous generators, increases inertia and damping, and improves the frequency stability of the power grid, has emerged.

[0004] Currently, existing energy storage converter grid-connected systems based on VSG control, by simulating the mechanical equations and droop characteristics of synchronous generators, can not only provide support for active power, reactive power and frequency of the power grid, but also effectively solve the problems of insufficient inertia and damping of power electronic converters. These systems have suppressed the fluctuations of the power grid frequency to a certain extent and improved the stability of the power grid operation. However, the VSG control with fixed virtual inertia and damping coefficients cannot achieve fast and accurate frequency recovery when the power grid frequency deviates from the nominal value, resulting in system response lag. Secondly, the line impedance in medium and low voltage distribution networks often shows resistive-inductive characteristics, which will cause coupling between active power and reactive power during grid-connected operation of VSG control, affecting the dynamic and steady-state control performance of VSG control, and further affecting the stability and safe and reliable operation of the power system.

[0005] Considering the influence of virtual inertia and damping coefficients on the dynamic performance of the system, it is urgent to design an adaptive control law related to the output angular frequency deviation and angular frequency change rate of VSG control, which can have a more flexible and efficient response during grid-connected operation. At the same time, the influence of the actual connection impedance characteristics may lead to coupling between active power and reactive power control. Therefore, it is also necessary to jointly break through the power decoupling technology of VSG control to achieve the continuous and safe and stable operation of the system.

[0006] In view of the adverse effects of fixed virtual inertia and damping coefficient design and power coupling problems on the grid connection of energy storage converters, it is urgent to develop a new method for the grid connection of energy storage converters based on virtual synchronous generator (VSG) control. When the grid-connected active power or grid-connected frequency is disturbed, this new method can not only dynamically adjust the virtual inertia and damping coefficient to adapt to the real-time changes of the power grid, but also effectively reduce the coupling effect between the VSG active power and reactive power, optimize the frequency and active power regulation ability of the power system, thereby providing solid technical support for the high-proportion grid connection of renewable energy and promoting the safe, stable and sustainable development of the power system. Summary of the Invention

[0007] The purpose of the present invention is to solve the defects of weak response ability to grid disturbances, insufficient power control accuracy and stability, and poor controller robustness in the existing grid connection control technology of energy storage converters, and to provide a grid connection method for energy storage converters based on virtual synchronous generator control to solve the above problems.

[0008] In order to achieve the above purpose, the technical solution of the present invention is as follows:

[0009] A grid connection method for an energy storage converter based on virtual synchronous generator control, comprising the following steps:

[0010] Design an adaptive rule for virtual inertia and damping coefficient in the active loop of the VSG controller;

[0011] Perform adaptive adjustment: Based on the adaptive rule of virtual inertia and damping coefficient, adaptively adjust the virtual inertia and damping coefficient;

[0012] Design a VSG power decoupling controller based on super-local model cooperative control;

[0013] Power smooth following and effective decoupling control of active power and reactive power.

[0014] The designed adaptive rule for virtual inertia and damping coefficient in the active loop of the VSG controller includes the following steps:

[0015] Set the adaptive rule for virtual inertia and damping coefficient in the active loop of the VSG controller as:

[0016]

[0017] where w, w n , w d , are respectively the angular frequency of the energy storage converter grid connection, the rated angular frequency of the energy storage converter grid connection, the frequency deviation of the energy storage converter grid connection, the angular frequency change rate during the energy storage converter grid connection process, ΔP is the power deviation at the grid connection, P refis the active power reference value of the virtual synchronous generator, P e is the actual active power of the virtual synchronous generator; D is the VSG damping coefficient, and J is the VSG virtual inertia;

[0018] It is set that when there is a disturbance in the active power, the angular frequency ω undergoes damped oscillation above and below ω n When ω > ω

[0019] When ω > ω n at this time,

[0020] When is greater than 0, the virtual inertia J is increased to suppress the growth of the angular frequency ω,

[0021] When is less than 0, the virtual inertia J is decreased to promote ω to recover to ω n ;

[0022] When ω < ω n at this time,

[0023] When is less than 0, the virtual inertia J is increased to suppress the reverse growth of the angular frequency ω,

[0024] When is less than 0, the virtual inertia J is decreased to promote ω to recover to ω n ;

[0025] When ω > ω n at this time,

[0026] When is greater than 0, the damping coefficient D is increased to accelerate the decay of the angular frequency ω oscillation,

[0027] When is less than 0, the damping coefficient D is decreased to avoid excessive energy dissipation;

[0028] When ω < ω n at this time,

[0029] When is less than 0, the damping coefficient D is increased to accelerate the decay of the reverse oscillation of the angular frequency ω,

[0030] When is less than 0, the damping coefficient D is decreased to avoid excessive energy dissipation.

[0031] The said adaptive adjustment includes the following steps:

[0032] Perform adaptive adjustment on the VSG virtual inertia J and the damping coefficient D, and its expression is:

[0033]

[0034] Where: J 0 represents the initial value of the virtual inertia, k 1 represents the adaptive adjustment coefficient of the virtual inertia, D 0 is the initial value of the damping coefficient, k 2 represents the adaptive adjustment coefficient of the damping coefficient, J is the VSG virtual inertia, D is the VSG damping coefficient, w d and |w d | are the angular frequency deviation of the energy storage converter connected to the grid and the absolute value of the angular frequency deviation of the energy storage converter connected to the grid, is the change rate of the angular frequency deviation during the grid connection process of the energy storage converter, and e is the exponential function;

[0035] Set the rotor motion equation in the active power loop of the VSG controller as:

[0036]

[0037] Substitute Equation (2) into the rotor motion equation of (3), then there is:

[0038]

[0039] In the formula, w n , w are respectively the rated angular frequency of the energy storage converter connected to the grid and the angular frequency of the energy storage converter connected to the grid, P ref is the active power reference value of the VSG, P e is the actual power of the VSG; D is the VSG damping coefficient;

[0040] Let the difference between the active power reference value and the actual value be △P, which is the power deviation at the grid connection; w d is the angular frequency deviation of the energy storage converter connected to the grid;

[0041] After derivation, it is obtained that:

[0042]

[0043] is the derivative of w d , solve Then there is:

[0044]

[0045] When the system is in a steady state, the angular frequency deviation w d of the energy storage converter connected to the grid = 0, the denominator of Equation (6) is zero, and the equation is invalid. Therefore, its denominator is rationalized as:

[0046]

[0047] Equation (7) represents the dynamic relationship between the power deviation ΔP at the grid connection point and the rate of change of the angular frequency deviation during the grid connection process of the energy storage converter when the active power is disturbed. between them;

[0048] When the active power is disturbed, the virtual inertia and damping coefficient of the VSG in the active power loop of the VSG controller are adjusted to reduce the angular frequency fluctuation.

[0049] The VSG power decoupling controller based on super-local model cooperative control includes the following steps:

[0050] Due to the non-purely inductive line impedance, the transmitted active power is controlled not only by the power angle δ but also by the voltage amplitude E, that is, there is a strong coupling between the active power and the reactive power. At the steady-state operating point δ 0 , E 0 The small-signal model near is expressed as:

[0051]

[0052] From Equation (8), it can be seen that

[0053] If δ 0 ≠0, that is, k′ pe ≠0, k′ qδ ≠0, with the phase of the grid voltage as the reference phase, δ 0 , E 0 being the stable operating points of the power angle and voltage amplitude, Z g being the line impedance between the grid connection point and the grid, U g being the grid voltage amplitude, θ being the line impedance angle, ΔP e being the active power deviation, ΔQ e being the reactive power deviation, sin and cos being the sine and cosine trigonometric functions, Δδ being the power angle deviation, ΔE being the voltage amplitude deviation, k pδ being the active power coefficient of the power angle deviation, k qδ the reactive power coefficient of the power angle deviation, k pe being the active power coefficient of the voltage amplitude deviation, k qe the reactive power coefficient of the voltage amplitude deviation;

[0054] Even if the line impedance is purely inductive, that is, θ = 90° and Z g = X g , there is still coupling between the active power and the reactive power;

[0055] Extract k pe and k qδ from Equation (8), and divide the two to get:

[0056]

[0057] The E in Equation (9) 0 is the stable operating point of the voltage amplitude, and X g is the line impedance from the grid connection point to the power grid, which is inductive and equal to the power grid voltage amplitude under steady state. In VSG control, the influence of the power angle on reactive power is much greater than that of the voltage amplitude, that is, the coupling effect of active power change on reactive power is greater, while the influence of reactive power change on active power is negligible;

[0058] First measure the reactive power output by the VSG, and then perform low-pass filtering, so that the bandwidth of the VSG power control loop is much smaller than the bandwidth of its voltage and current control loops, then there is:

[0059]

[0060] In the formula: Q em is the reactive power after passing through the low-pass filter, ω c is the cut-off frequency of the low-pass filter, s is the Laplace operator, and Q e is the actual reactive power;

[0061] The small-signal model in the time domain of Equation (10) is:

[0062]

[0063] In the formula: ΔQ em is the reactive power deviation caused by the change in active power, is the derivative of the reactive power deviation;

[0064] Based on Equation (8) and Equation (11), taking the voltage amplitude fluctuation amount when there is reactive power deviation as the compensation voltage, the dynamic equation of the reactive power deviation of the VSG power decoupling controller is established as:

[0065]

[0066] In the formula, E comp is the voltage compensation amount output by the VSG power decoupling controller,

[0067] The power decoupling controller is to achieve decoupling between the active power and reactive power of the VSG, that is, when the active power command suddenly changes, the reactive power deviation remains 0. For this reason, the macro variable ψ e is designed as:

[0068]

[0069] In the formula, is the expected reactive power deviation, λ 1 is the proportional coefficient, and λ 2 is the integral coefficient, and ∫e is the integral of e;

[0070] Design the dynamic evolution equation of cooperative control, and there is

[0071]

[0072] In the formula, T e is the reactive power deviation convergence time parameter, is the derivative of the macro variable ψ in Equation (14) e ;

[0073] The cooperative control law E output by the VSG power decoupling controller obtained through derivation comp is:

[0074]

[0075] Combine the cooperative control law E output by the VSG power decoupling controller comp with the voltage compensation and the VSG reactive power control loop to achieve power decoupling;

[0076] For a system with single input and single output, it is expressed by an algebraic differential equation as:

[0077]

[0078] In the formula, t is the time variable, u and y are the input and output of the system respectively, E is a differentiable function, y (n) is the nth derivative of the system output, u (b) is the bth derivative of the system input, y (a) is the bth derivative of the system output,

[0079] Equation (16) is written in the form of a superlocal model, and there is:

[0080] y (n) (t) = F(t) + αu(t) (17)

[0081] In the formula, y (n) (t) is the nth derivative of the system output at time t, n represents the system order, usually taken as 1, α is the proportionality factor of the system input, the continuously updated variable F(t) represents the known part, unknown part and various possible disturbances of the system, and the estimation of the variable F(t) is determined by using the input and output data of the system, u(t) is the input of the system at time t;

[0082] Based on the dynamic equation of the reactive power deviation of the VSG power decoupling controller, in order to get rid of its sensitive dependence on the line impedance and the change of the power angle, establish a superlocal model of the VSG reactive power deviation, and there is:

[0083]

[0084] Wherein, a n is the proportional parameter, Δa is the error voltage coefficient, F est represents the known and unknown parts, E comp is the voltage compensation value output by the VSG power decoupling controller, is the derivative of the reactive power deviation, ΔQ em is the reactive power deviation caused by the change in active power, ω c is the cut-off frequency of the low-pass filter, Z g is the line impedance between the grid connection point and the power grid, U g is the amplitude of the grid voltage, sin and cos are sine and cosine trigonometric functions, Δδ is the power angle deviation, δ 0 、E 0 are the stable operating points of the power angle and voltage amplitude, θ is the line impedance angle;

[0085] F est Online estimation: Based on the VSG reactive power deviation hyperlocal model, online estimation of F est is carried out.

[0086] The online estimation of the said F est includes the following steps:

[0087] Assume that F est is a constant within a relatively short period of time, represents its estimated value in the frequency domain. The frequency domain expression of Equation (18) is:

[0088]

[0089] Wherein, ΔQ e0 represents the initial value of the reactive power deviation, s is the Laplace transform operator, a n is the proportional parameter, E comp is the voltage compensation value output by the VSG power decoupling controller, ΔQ em is the reactive power deviation caused by the change in active power;

[0090] Differentiate both sides of the equation with respect to s simultaneously to eliminate the influence of the initial value, and we get:

[0091]

[0092] Wherein, is the derivative of ΔQ em in the frequency domain, is the derivative of E comp in the frequency domain;

[0093] In order to avoid the noise amplification caused by the noise existing in the measurement of reactive power due to differentiation, multiply both sides of the equation by s-2 , we get:

[0094]

[0095] Estimate based on the algebraic identification method, transform Equation (21) into the time domain form, and obtain the estimated value of F within the sampling interval [0, T F as: est The estimated value of F is:

[0096]

[0097] In the formula, is the estimated value of the known and unknown parts F est at time t, T s is the control period, T F = n F T s is the sliding window length, t is the time variable, E comp (t) and ΔQ em (t) are the voltage compensation amount and reactive power deviation at time t respectively, and T F is the sliding window length;

[0098] Substitute the estimated by Equation (22) into Equation (18), establish a super-local model of the VSG reactive power deviation, and express it as:

[0099]

[0100] In the formula, is the derivative of the reactive power deviation, T F is the sliding window length, E comp is the voltage compensation value output by the VSG power decoupling controller, E comp (t) and ΔQ em (t) are the voltage compensation amount and reactive power deviation at time t respectively, and t is the time variable;

[0101] Based on the designed cooperative control VSG power decoupling controller with a reactive power deviation super-local model, the generated cooperative control law is:

[0102]

[0103] Among them, is the estimation based on the algebraic identification method in step 54), λ 1 is the proportionality coefficient, λ 2 is the integral coefficient, ∫e is the integral of e, T e is the reactive power deviation convergence time parameter.

[0104] The steps of the power smoothing following and the effective decoupling control of active power and reactive power are as follows: In the grid-connected mode, the active power is calculated by using the obtained current and voltage at the grid connection point, which serves as the input of the active loop of the VSG controller. After being adjusted by the adaptive rules of the virtual inertia and damping coefficient, the active power is smoothly tracked; based on the reactive power calculated at the grid connection point, which serves as the input of the VSG power decoupling controller based on the cooperative control of the super-local model, a macro variable is designed with the control objective of keeping the reactive power deviation zero to maintain the stability of the reactive power and achieve robust control that is insensitive to the line impedance parameters.

[0105] Beneficial effects

[0106] A grid-connected method for an energy storage converter based on virtual synchronous generator control according to the present invention provides, compared with the prior art, an energy storage grid-connected converter with the inertia and damping characteristics of a synchronous generator during operation. When disturbances occur in the grid-connected active power or grid-connected frequency, the frequency fluctuations can be suppressed by the adaptive adjustment of the virtual inertia and damping coefficient, enhancing the stability of the grid-connected operation of the energy storage converter;

[0107] The present invention proposes a VSG power decoupling controller based on the cooperative control of the super-local model. A manifold is designed with the control objective of zero reactive power deviation, enabling the system to obtain a more obvious power decoupling effect and achieve robust control that is insensitive to the line impedance parameters;

[0108] The present invention builds an energy storage converter grid-connected simulation model based on Simulink software, and verifies the supporting effect of the improved VSG control on the system frequency stability through system simulation research. Description of the drawings

[0109] Figure 1 It is the flowchart of the method of the present invention;

[0110] Figure 2 It is the equivalent circuit diagram of the virtual synchronous generator;

[0111] Figure 3 It is the VSG active loop control diagram based on the adaptive dynamic virtual inertia and damping coefficient;

[0112] Figure 4 It is the VSG power decoupling reactive loop control diagram based on the cooperative control of the super-local model;

[0113] Figure 5 It is the dynamic schematic diagram of the grid-connected angular frequency;

[0114] Figure 6 It is the schematic diagram of the adaptive change of the virtual inertia;

[0115] Figure 7 It is the schematic diagram of the adaptive change of the damping coefficient;

[0116] Figure 8 It is a schematic diagram for comparing the decoupling performance of active and reactive power under the line impedance Rg = 1Ω, Lg = 2mH;

[0117] Figure 9 Schematic diagram for comparing the active and reactive power decoupling performance when the line impedance Rg = 3Ω, Lg = 2mH. DETAILED DESCRIPTION

[0118] In order to have a further understanding and recognition of the structural features and the effects achieved by the present invention, a preferred embodiment and accompanying drawings are used for detailed description as follows:

[0119] like Figure 1 As shown, the energy storage converter grid-connected method based on virtual synchronous generator control described in the present invention comprises the following steps:

[0120] In the first step, the adaptive rules of virtual inertia and damping coefficient in the active loop of VSG controller are designed.

[0121] For high-ratio grid-connected systems, when a large power disturbance occurs, the system will have a large frequency deviation due to its low inertia, causing the grid-connected system to be disconnected from the main grid. Figure 2 It is the equivalent circuit diagram of grid-connected energy storage converter based on VSG control.

[0122] (1) The adaptive rule for the virtual inertia and damping coefficient in the active loop of the VSG controller is set as follows:

[0123]

[0124] In the formula, w, w n 、w d , are the grid-connected angular frequency of the energy storage converter, the rated angular frequency of the energy storage converter, the frequency deviation of the energy storage converter, and the angular frequency change rate during the grid-connected process of the energy storage converter. ΔP is the power deviation at the grid connection point, P ref is the active power reference value of the virtual synchronous generator, P e is the actual active power of the virtual synchronous generator, D is the damping coefficient of VSG, and J is the virtual inertia of VSG.

[0125] (2) When the active power is disturbed, the angular frequency w is set to w n Damped oscillation occurs up and down,

[0126] A1) When w>w n hour,

[0127] when Greater than 0, the growth of angular frequency ω is suppressed by increasing the virtual inertia J.

[0128] When is less than 0, the virtual inertia J is decreased to prompt ω to recover to ω n ;

[0129] A2) When ω < ω n at that time,

[0130] When is less than 0, the reverse growth of angular frequency ω is suppressed by increasing the virtual inertia J.

[0131] When is less than 0, the virtual inertia J is decreased to prompt ω to recover to ω n ;

[0132] A3) When ω > ω n at that time,

[0133] When is greater than 0, the damping coefficient D is increased to accelerate the decay of the oscillation of angular frequency ω.

[0134] When is less than 0, the damping coefficient D is decreased to avoid excessive energy dissipation;

[0135] A4) When ω < ω n at that time,

[0136] When is less than 0, the damping coefficient D is increased to accelerate the decay of the reverse oscillation of angular frequency ω.

[0137] When is less than 0, the damping coefficient D is decreased to avoid excessive energy dissipation.

[0138] The second step is to perform adaptive adjustment: Based on the adaptive rules of virtual inertia and damping coefficient, the virtual inertia and damping coefficient are adaptively adjusted.

[0139] Flexibly variable virtual inertia and damping coefficient are set in the active power loop of the VSG controller to make timely adjustments to the power disturbances occurring in the power grid and strengthen the grid connection stability. Figure 3 It is the control diagram of the VSG active power loop based on adaptive dynamic virtual inertia and damping.

[0140] (1) The virtual inertia J and damping coefficient D of the VSG are adaptively adjusted, and their expressions are:

[0141]

[0142] Among them: J 0 represents the initial value of the virtual inertia, k1 Denotes the adaptive regulation coefficient of the virtual inertia, D 0 Is the initial value of the damping coefficient, k 2 Denotes the adaptive regulation coefficient of the damping coefficient, J is the virtual inertia of the VSG, D is the damping coefficient of the VSG, w d And |w d | Are the angular frequency deviation of the energy storage converter grid connection and the absolute value of the angular frequency deviation of the energy storage converter grid connection, Is the change rate of the angular frequency deviation during the grid connection process of the energy storage converter, and e is the exponential function.

[0143] (2) Set the rotor motion equation in the active power loop of the VSG controller as:

[0144]

[0145] B1) Substitute Equation (2) into (3) the rotor motion equation, then there is:

[0146]

[0147] In the formula, w n , w are respectively the rated angular frequency of the energy storage converter grid connection and the angular frequency of the energy storage converter grid connection, P ref Is the active power reference value of the VSG, P e Is the actual power output by the VSG.

[0148] B2) Let the difference between the active power reference value and the actual value be △P, which is the power deviation at the grid connection; w d Is the frequency deviation of the energy storage converter grid connection;

[0149] After derivation, it is obtained:

[0150]

[0151] B3) Is the derivative of w d , Solve Then there is:

[0152]

[0153] B4) When the system is in a steady state, the angular frequency deviation w of the energy storage converter grid connection d = 0, the denominator of Equation (6) is zero, and the equation is invalid. Therefore, its denominator is rationalized as:

[0154]

[0155] Equation (7) is the dynamic relationship between the active power disturbance ΔP and the change rate of the angular frequency deviation during the grid connection process of the energy storage converter Between;

[0156] When there is a disturbance in the active power, the adaptive dynamic virtual inertia and damping coefficient in the active loop of the VSG controller are adjusted to reduce the angular frequency fluctuation.

[0157] Step 3: Design a VSG power decoupling controller based on super-local model cooperative control.

[0158] When the energy storage converter based on VSG control is connected to the grid, due to the line impedance in the medium and low voltage distribution network often presenting an inductive resistance characteristic, it will cause mutual coupling between the active power and the reactive power during grid-connected operation, affecting the dynamic and steady-state control performance of the VSG control. Design a VSG power decoupling controller based on super-local model cooperative control, aiming to achieve robust control insensitive to the line impedance parameters.

[0159] (1) Since the line impedance is not purely inductive, the transmitted active power is controlled not only by the power angle δ but also by the voltage amplitude E, that is, there is strong coupling between the powers. At the steady-state operating point δ 0 、E 0 The small-signal model near is:

[0160]

[0161] As known from Equation (8),

[0162] If δ 0 ≠0, that is, k′ pe ≠0, k′ qδ ≠0, as Figure 2 As shown in the equivalent circuit diagram of the energy storage converter based on VSG control connected to the grid, the phase of the grid voltage is the reference phase, δ 0 、E 0 are the stable operating points of the power angle and voltage amplitude, Z g is the line impedance between the grid connection point and the grid, U g is the grid voltage amplitude, θ is the line impedance angle, ΔP e is the active power deviation, ΔQ e is the reactive power deviation, sin and cos are sine and cosine trigonometric functions, Δδ is the power angle deviation, ΔE is the voltage amplitude deviation, k pδ is the active power coefficient of the power angle deviation, k qδ reactive power coefficient of the power angle deviation, k pe is the active power coefficient of the voltage amplitude deviation, k qe reactive power coefficient of the voltage amplitude deviation;

[0163] (2) Although the line impedance is not purely inductive, that is, θ = 90° and Z g = X g , there is still coupling between the active and reactive powers;

[0164] Extract \(k\) in Equation (8) pe and \(k\) qδ , and divide the two to get:

[0165]

[0166] \(E\) in Equation (9) 0 is the stable operating point of the voltage amplitude, \(X\) g is the line impedance from the point of common coupling to the power grid, which is inductive and equal to the power grid voltage amplitude under steady state. In VSG control, the influence of the power angle on the reactive power is much greater than that of the voltage amplitude, that is, the coupling influence of the active power change on the reactive power is greater, while the influence of the reactive power change on the active power is negligible.

[0167] (3) Measure the reactive power output by the VSG first, and then perform low-pass filtering, so that the bandwidth of the VSG power control loop is much smaller than the bandwidth of its voltage and current control loops, then there is:

[0168]

[0169] In the formula: \(Q\) em is the reactive power after passing through the low-pass filter, \(\omega\) c is the cut-off frequency of the low-pass filter, \(s\) is the Laplace operator, \(Q\) e is the actual reactive power;

[0170] The small-signal model in the time domain of Equation (10) is:

[0171]

[0172] In the formula: \(\Delta Q\) em is the reactive power deviation caused by the change of the active power, is the derivative of the reactive power deviation;

[0173] Based on Equation (8) and Equation (11), taking the voltage amplitude fluctuation when there is a reactive power deviation as the compensation voltage, the dynamic equation of the reactive power deviation of the VSG power decoupling controller is established as:

[0174]

[0175] In the formula, \(E\) comp is the voltage compensation amount output by the VSG power decoupling controller,

[0176] The power decoupling controller is to achieve the power decoupling of the VSG, that is, when the active power command suddenly changes, the reactive power deviation remains 0. For this reason, the macro variable \(\psi\) e is designed as:

[0177] \(\psi\)e = λ 1 e + λ 2 ∫e (13)

[0178] Wherein, is the desired reactive power deviation, λ 1 is the proportionality coefficient, λ 2 is the integral coefficient, and ∫e is the integral of e.

[0179] (4) Design the dynamic evolution equation of coordinated control, and there is

[0180]

[0181] Wherein, T e is the reactive power deviation convergence time parameter, is the derivative of the macro variable ψ in formula (14) e with respect to;

[0182] The coordinated control law E output by the VSG power decoupling controller obtained through derivation comp is:

[0183]

[0184] Combine the coordinated control law E output by the VSG power decoupling controller comp with the voltage compensation and the VSG reactive power control loop to achieve power decoupling;

[0185] However, the generated coordinated control law is based on the known line impedance value and the system being in a steady state. However, the actual line impedance is difficult to estimate, and the operating conditions of the microgrid based on VSG control may change at any time. Therefore, it is urgent to study a model-free scheme to get rid of the control law that depends on system parameters, Figure 4 which is the VSG power decoupling reactive power loop control diagram based on hyperlocal model coordinated control.

[0186] (5) For a controller system with a single control input and a single output, it is expressed by an algebraic differential equation as:

[0187]

[0188] Wherein, t is the time variable, u and y are the input and output of the controller system respectively, E is a differentiable function, y (n) is the nth derivative of the output of the controller system, u (b) is the bth derivative of the input of the controller system, y (a) is the bth derivative of the output of the controller system,

[0189] Formula (16) can be written in the form of a hyperlocal model, and there is:

[0190] y(n) y(t)=F(t)+αu(t) (17)

[0191] Wherein, y (n) (t) is the nth derivative of the output of the controller system at time t, n represents the system order, usually taken as 1, α is the proportionality factor of the system input, the continuously updated variable F(t) represents the known part, unknown part and various possible disturbances of the system, and the estimation of the variable F(t) is determined by using the input and output data of the system, and u(t) is the input of the controller system at time t.

[0192] (6) Based on the dynamic equation of the reactive power deviation of the VSG power decoupling controller, in order to get rid of its sensitive dependence on the line impedance and the change of the power angle, a super-local model of the VSG reactive power deviation is established, and there is:[[]]

[0193]

[0194] Wherein, a n is the proportionality parameter, Δa is the error voltage coefficient, F est represents the known and unknown parts, E comp is the voltage compensation value output by the VSG power decoupling controller, is the derivative of the reactive power deviation, ΔQ em is the reactive power deviation caused by the active power change, ω c is the cut-off frequency of the low-pass filter, Z g is the line impedance between the grid connection point and the power grid, U g is the amplitude of the grid voltage, sin and cos are sine and cosine trigonometric functions, Δδ is the power angle deviation, δ 0 、E 0 is the stable operating point of the voltage amplitude of the sum of the power angles, θ is the line impedance angle.

[0195] (7) Online estimation of F est : Based on the super-local model of the VSG reactive power deviation, online estimation of F est is carried out.

[0196] C1) Assume that F est is a constant within a relatively short time, represents its estimated value in the frequency domain, and the frequency domain expression of Equation (18) is:[[]]

[0197]

[0198] Wherein, ΔQ e0 represents the initial output, s is the Laplace transform operator, a n is the proportionality parameter, E compis the voltage compensation value output by the VSG power decoupling controller, ΔQ em is the reactive power deviation caused by the change in active power;

[0199] C2) Differentiate both sides of the equation with respect to s to eliminate the influence of the initial output, and we get:

[0200]

[0201] In the formula, is ΔQ em Take the derivative in the frequency domain, is E comp Take the derivative in the frequency domain;

[0202] C3) In order to avoid the noise amplification caused by the noise existing in the reactive power measurement during differentiation, multiply both sides of the equation by s -2 , and we get:

[0203]

[0204] C4) Estimate based on the algebraic identification method. Convert Equation (21) into the time domain form, and obtain the estimated value of F F within the sampling interval [0, T est as:

[0205]

[0206] In the formula, is the estimated value of the known and unknown parts F est at time t, T s is the control period, T F = n F T s is the sliding window length, t is the time variable, E comp (t) and ΔQ em (t) are the voltage compensation amount and reactive power deviation at time t respectively, T F is the sliding window length;

[0207] C5) Substitute the estimated obtained from Equation (21) into Equation (18) to establish a super-local model of the VSG reactive power deviation, and it is expressed as:

[0208]

[0209] In the formula, is the derivative of the reactive power deviation, T F is the sliding window length, E comp is the voltage compensation value output by the VSG power decoupling controller, E comp (t) and ΔQem (t) is the voltage compensation amount and reactive power deviation at time t respectively, and t is the time variable;

[0210] C6) Based on the designed cooperative control VSG power decoupling controller based on the reactive power deviation hyperlocal model, the output cooperative control law is:

[0211]

[0212] Wherein, is estimated by the method based on algebraic identification in step 54), λ 1 is the proportional coefficient, λ 2 is the integral coefficient, ∫e is the integral of e, T e is the reactive power deviation convergence time parameter.

[0213] Use the Lyapunov stability theorem to prove the stability of the grid-connected system of the energy storage converter based on VSG control. Define the Lyapunov function as:

[0214]

[0215] In the formula, V is the Lyapunov function variable, and ψ e is the macro variable defined in formula (13).

[0216] The derivative of the Lyapunov function is:

[0217]

[0218] It can be seen from formula (26) that as long as T e > 0, it can be ensured that the designed macro variable ψ e can be stabilized to 0, and ψ e is a linear combination of the proportional integral form of the reactive power deviation, which proves that the proposed VSG power decoupling controller based on hyperlocal model cooperative control is globally stable.

[0219] Fourth step, power smoothing following and active and reactive power effective decoupling control: In the grid-connected mode, use the obtained current and voltage at the grid connection point to calculate the active power, which is used as the input of the active loop of the VSG controller. After adjustment by the adaptive rules of the virtual inertia and damping coefficient, the active power is smoothly tracked; based on the reactive power calculated at the grid connection point, which is used as the input of the VSG power decoupling controller based on hyperlocal model cooperative control, a macro variable function is designed with the goal of keeping the reactive power deviation zero to keep the reactive power stable and achieve robust control that is insensitive to the line impedance parameters.

[0220] To verify the effectiveness of the proposed grid - connection method for energy storage converters based on VSG control, a Simulink simulation model was built using Matlab software.

[0221] First, the initial active power reference value was set to 2 kW and suddenly changed to 5 kW at t = 1.5 s. The dynamic grid - connection angular frequency is as Figure 5 shown, and the virtual inertia and damping coefficient change adaptively when the angular frequency changes dynamically, as Figure 6 and Figure 7 shown.

[0222] Then, the initial active power reference value was set to 0 kW and the reactive power was 0 var. At time t = 2 s, the active power suddenly changed to 10 kW and the reactive power remained 0 var. To compare the performance of the power decoupling controller, the root mean square error (RMSE) of the reactive power deviation was used in this invention to measure the effects of different control strategies. The dynamic comparison of the decoupling performance of the active and reactive powers output by the VSG under the line impedance Rg = 1Ω, Lg = 2mH is as Figure 8 shown, and the dynamic comparison of the decoupling performance of the active and reactive powers output by the VSG under the line impedance Rg = 3Ω, Lg = 2mH is as Figure 9 shown.

[0223] In Figure 5 , the upper line is the grid - connection angular frequency w(VSG) of the energy storage converter with traditional VSG grid - connection control, and the lower line is the grid - connection angular frequency w(Improved VSG) of the energy storage converter using the method designed in this invention; in Figure 6 , the upper line is the self - adjustable VSG virtual inertia self - J, D that can be adaptively adjusted dynamically designed in this invention, and the lower line is the traditional fixed VSG virtual inertia, constant - J, D; in Figure 7 , the upper line is the self - adjustable VSG damping coefficient, self - J, D that can be adaptively adjusted dynamically designed in this invention, and the lower line is the traditional fixed VSG damping coefficient, constant - J, D; in Figure 8 , from left to right are: under the working conditions of line impedance R g = 1Ω, L g = 2mH, power decoupling of the VSG is achieved without decoupling and based on the cooperative control of the super - local model; in Figure 9 , under the line impedance R g = 3Ω, L gUnder the condition of 2mH, from left to right are: without decoupling and coordinated control based on the hyperlocal model to achieve VSG power decoupling. When the sudden change of active power causes the fluctuation of the power grid frequency, the grid-connected method of the energy storage converter based on virtual synchronous generator control designed by the present invention can achieve the adaptive adjustment of virtual inertia and damping coefficient. Compared with the traditional VSG control, it has the technical advantages of smaller frequency fluctuation and faster frequency recovery speed.

[0224] By Figure 8 and Figure 9 Comparing the RMSE of reactive power in, the VSG power decoupling controller based on the coordinated control of the hyperlocal model of the present invention has a smaller RMSE and enjoys better power decoupling performance compared with the traditional VSG grid-connected control.

[0225] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and the principles described in the specification are only the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection required by the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for connecting an energy storage converter to the grid based on virtual synchronous generator control, characterized in that: The following steps are involved: 11) Design the adaptive rules of virtual inertia and damping coefficient in the active loop of VSG controller; 12) Adaptive adjustment: Adaptive adjustment is performed on the virtual inertia and the damping coefficient based on the adaptive rules of the virtual inertia and the damping coefficient; 13) Design a VSG power decoupling controller based on hyperlocal model cooperative control; 14) Power smooth following and effective decoupling control of active power and reactive power.

2. A method for connecting an energy storage converter to the grid based on virtual synchronous generator control according to claim 1, characterized in that: The designed adaptive rule of virtual inertia and damping coefficient in the active loop of the VSG controller includes the following steps: 21) The adaptive rules for the virtual inertia and damping coefficient in the active loop of the VSG controller are set as follows: In the formula, w, w n 、w d , are the grid-connected angular frequency of the energy storage converter, the rated angular frequency of the energy storage converter, the frequency deviation of the energy storage converter, and the angular frequency change rate during the grid-connected process of the energy storage converter. ΔP is the power deviation at the grid connection point, P ref is the active power reference value of the virtual synchronous generator, P e is the actual active power of the virtual synchronous generator; D is the VSG damping coefficient, and J is the VSG virtual inertia; 22) When the active power is disturbed, the angular frequency w is set to w n Damped oscillation occurs up and down, 221) When w>w n hour, when is greater than 0, and the virtual inertia J is increased to suppress the growth of the angular frequency w. when Less than 0, the virtual inertia J is reduced to restore w to w n ; 222) When w <w n hour, when Less than 0, the virtual inertia J is increased to suppress the reverse growth of the angular frequency w. when Less than 0, the virtual inertia J is reduced to restore w to w n ; 223) When w>w n hour, when is greater than 0, the damping coefficient D is increased to accelerate the attenuation of the oscillation of the angular frequency w. when Less than 0, the damping coefficient D is used to reduce it to avoid excessive energy dissipation; 224) When w <w n hour, when Less than 0, the damping coefficient D is increased to accelerate the attenuation of the reverse oscillation of the angular frequency w. when Less than 0, the damping coefficient D is reduced to avoid excessive energy dissipation.

3. A method for connecting an energy storage converter to the grid based on virtual synchronous generator control according to claim 1, characterized in that: The adaptive adjustment comprises the following steps: 31) Adaptively adjust the VSG virtual inertia J and damping coefficient D, and the expression is: Where: J0 represents the initial value of virtual inertia, k1 represents the adaptive adjustment coefficient of virtual inertia, D0 is the initial value of damping coefficient, k2 represents the adaptive adjustment coefficient of damping coefficient, J is VSG virtual inertia, D is VSG damping coefficient, w d and |w d | is the angular frequency deviation of the energy storage converter connected to the grid and the absolute value of the angular frequency deviation of the energy storage converter connected to the grid, is the rate of change of angular frequency deviation during grid connection of energy storage converter, and e is an exponential function; 32) Set the rotor motion equation in the active loop of the VSG controller to: 321) Substituting equation (2) into the rotor motion equation (3), we have: In the formula, w n , w are the rated angular frequency of the energy storage converter connected to the grid and the angular frequency of the energy storage converter connected to the grid, P ref is the active power reference value of VSG, P e is the actual power of VSG, D is the VSG damping coefficient; 322) Let the difference between the active power reference value and the actual value be △P, which is the power deviation at the grid connection point; w d is the angular frequency deviation of the energy storage converter connected to the grid; It is deduced that: 323) w d The derivative of Then we have: 324) When the system is in steady state, the angular frequency deviation w of the energy storage converter connected to the grid d =0, the denominator of formula (6) is zero, and the equation is invalid, so its denominator is rationalized as follows: Formula (7) is the power deviation ΔP at the grid connection point and the angular frequency deviation change rate of the energy storage converter during the grid connection process when the active power is disturbed. The dynamic relationship between When the active power is disturbed, the VSG virtual inertia and damping coefficient in the active loop of the VSG controller are adjusted to reduce the angular frequency fluctuation.

4. The method for connecting an energy storage converter to the grid based on virtual synchronous generator control according to claim 1, characterized in that: The design of a VSG power decoupling controller based on hyperlocal model cooperative control includes the following steps: 41) Since the line impedance is not purely inductive, the transmitted active power is not only controlled by the power angle δ, but also by the voltage amplitude E, that is, there is a strong coupling between the active power and the reactive power. The small signal model near the steady-state operating point δ0, E0 is expressed as: From formula (8), we know that If δ0≠0, that is, k′ pe ≠0,k′ qδ ≠0, the phase of the grid voltage is the reference phase, δ0 and E0 are the stable operating points of the power angle and voltage amplitude, Z g is the line impedance between the grid connection point and the grid, U g is the grid voltage amplitude, θ is the line impedance angle, ΔP e is the active power deviation, ΔQ e is the reactive power deviation, sin and cos are the sine and cosine trigonometric functions, Δδ is the power angle deviation, ΔE is the voltage amplitude deviation, k pδ is the active power coefficient of the power angle deviation, k qδ Reactive power coefficient of power angle deviation, k pe is the active power coefficient of voltage amplitude deviation, k qe Reactive power factor of voltage amplitude deviation; 42) Even if the line impedance is purely inductive, that is, θ = 90° and Z g =X g , there is still coupling between active power and reactive power; Extract k in formula (8) pe and k qδ , dividing the two gives: E0 in formula (9) is the stable operating point of the voltage amplitude, X g The line impedance between the grid connection point and the grid is inductive, and is equal to the grid voltage amplitude in steady state. In VSG control, the influence of the power angle on reactive power is much greater than that of the voltage amplitude, that is, the coupling effect of active power change on reactive power is greater, while the effect of reactive power change on active power is negligible. 43) The reactive power output of VSG is first measured and then low-pass filtered, so that the bandwidth of VSG power control loop is much smaller than the bandwidth of its voltage and current control loop. Then: Where: Q em is the reactive power after passing through the low-pass filter, ω c is the cutoff frequency of the low-pass filter, s is the Laplace operator, Q e is the actual reactive power; The time domain small signal model of formula (10) is: Where: ΔQ em is the reactive power deviation caused by changes in active power, is the derivative of reactive power deviation; Based on equations (8) and (11), the voltage amplitude fluctuation when there is reactive power deviation is used as the compensation voltage, and the dynamic equation of reactive power deviation of VSG power decoupling controller is established as follows: In the formula, E comp is the voltage compensation output by the VSG power decoupling controller, The power decoupling controller is used to achieve the decoupling between the active power and reactive power of the VSG, that is, when the active power command changes suddenly, the reactive power deviation remains at 0. To this end, the macro variable ψ e Designed for: ψ e =λ1e+λ2∫e (13) In the formula, is the desired reactive power deviation, λ1 is the proportional coefficient, λ2 is the integral coefficient, and ∫e is the integral of e; 44) Design the dynamic evolution equation of cooperative control, and have Where, T e is the reactive deviation convergence time parameter, is the macro variable ψ in equation (14) e Derivation; The derived coordinated control law E of the VSG power decoupling controller output comp for: The coordinated control law E output by the VSG power decoupling controller comp As voltage compensation combined with VSG reactive control loop, power decoupling is achieved; 45) For a system with a single input and a single output, it can be expressed as an algebraic differential equation: Where t is the time variable, u and y are the input and output of the system respectively, E is a differentiable function, and y (n) is the nth-order derivative of the system output, u (b) is the b-order derivative of the system input, y (a) is the b-order derivative of the system output, Formula (16) is written in the form of a super-local model, and we have: y (n) (t)=F(t)+αu(t) (17) In the formula, y (n) (t) is the nth-order derivative of the system output at time t, n represents the system order, usually 1, α is the proportional factor of the system input, the continuously updated variable F(t) represents the known part, unknown part and various possible interferences of the system, the estimate of the variable F(t) is determined by using the input and output data of the system, u(t) is the input of the system at time t; 46) Based on the dynamic equation of reactive power deviation of VSG power decoupling controller, in order to get rid of its sensitive dependence on line impedance and power angle changes, a super-local model of VSG reactive power deviation is established, and: In the formula, a n is the proportional parameter, Δa is the error voltage coefficient, F est Indicates the known and unknown parts, E comp is the voltage compensation value output by the VSG power decoupling controller, is the derivative of reactive power deviation, ΔQ em is the reactive power deviation caused by the change of active power, ω c is the low-pass filter cutoff frequency, Z g is the line impedance between the grid connection point and the grid, U g is the grid voltage amplitude, sin and cos are the sine and cosine trigonometric functions, Δδ is the power angle deviation, δ0 and E0 are the stable operating points of the power angle and voltage amplitude, and θ is the line impedance angle; 47)F est Online estimation of F: Based on the VSG reactive power deviation super-local model, est Get an online estimate.

5. A method for connecting an energy storage converter to the grid based on virtual synchronous generator control according to claim 4, characterized in that: The F est The online estimation of includes the following steps: 51) Assume F est In a shorter period of time, it is a constant. Represents its estimated value in the frequency domain. The frequency domain expression of formula (18) is: Where ΔQ e0 represents the initial value of reactive power deviation, s is the Laplace transform operator, a n is the scale parameter, E comp is the voltage compensation value output by the VSG power decoupling controller, ΔQ em It is the reactive power deviation caused by the change of active power; 52) Differentiate both sides of the equation to eliminate the effect of the initial value and obtain: In the formula, = ΔQ em Taking the derivative in the frequency domain, For E comp Derivative in the frequency domain; 53) In order to avoid the noise amplification caused by differentiation when measuring reactive power, both sides of the equation are multiplied by s -2 ,have to: 54) Based on the algebraic identification method, the equation (21) is transformed into the time domain form. F ] to obtain F est The estimated value of is: In the formula, is the known and unknown parts F at time t est The estimated value of T s is the control period, T F =n F T s is the sliding window length, t is the time variable, E comp (t) and ΔQ em (t) are the voltage compensation and reactive power deviation at time t, T F is the sliding window length; 55) The estimated value of formula (22) Substituting into equation (18), the ultra-local model of VSG reactive power deviation is established and expressed as: In the formula, is the derivative of reactive power deviation, T F is the sliding window length, E comp is the voltage compensation value output by the VSG power decoupling controller, E comp (t) and ΔQ em (t) are the voltage compensation and reactive power deviation at time t, respectively, and t is the time variable; 56) Based on the designed cooperative control VSG power decoupling controller based on reactive deviation super-local model, the generated cooperative control law is: in, is estimated by the algebraic identification method in step 54), λ1 is the proportional coefficient, λ2 is the integral coefficient, ∫e is the integral of e, T e It is the reactive power deviation convergence time parameter.

6. The method for connecting an energy storage converter to the grid based on virtual synchronous generator control according to claim 1, characterized in that: The steps of power smooth following and effective decoupling control of active power and reactive power are as follows: in the grid-connected mode, the active power is calculated using the current and voltage at the grid-connected point, which is used as the input of the active loop of the VSG controller, and the active power smooth tracking is achieved through adjustment of the virtual inertia and damping coefficient adaptive rules; based on the reactive power calculated at the grid-connected point, which is used as the input of the VSG power decoupling controller based on the hyperlocal model collaborative control, the macro variable is designed with keeping the reactive deviation at zero as the control target, so as to keep the reactive power stable and achieve robust control that is insensitive to line impedance parameters.

Images