OSCILLATION ACTIVE DAMPING CONTROL METHOD AND SYSTEM FOR GRID-TIED TYPE-4 WIND TURBINE GENERATOR

Abstract

The application relates to an oscillation active damping control method and system for grid-tied type-4 wind turbine generator. The method comprises: based on an interconnection model of multiple subsystems, constructing a stored energy function and a dissipated energy function of a current inner loop control subsystem, and interaction energy functions between the current inner loop control subsystem and other subsystems are constructed, then establishing an energy feedback model of Type-4 wind turbine generator; when the oscillation occurs, obtaining instantaneous angular frequency of the PLL, and then based on the energy feedback model, adjusting the current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem, to make the stored energy function decrease with time, so as to suppress the oscillation.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims priority to Chinese Patent Application No. 202110616024.X, filed on Jun. 2, 2021, which is hereby incorporated by reference in their entirety.

TECHNICAL FIELD

The application relates to the field of wind power generation and especially to an oscillation active damping control method and system for grid-tied type-4 wind turbine generator.

BACKGROUND

In recent years, the grid-tied type-4 wind turbine generator (direct-drive wind turbine generator) in large-scale wind power collection areas frequently results in sub/super-synchronous oscillation. The sub/super-synchronous oscillation components passing through multi-level power grids excites the torsional vibration of the turbine set shafting, which resulting in the accidents of generator tripping and sudden drop of HVDC (High Voltage Direct Current) power, threatening the safe operation of the power grid.

To solve this problem, two main methods based on the frequency domain analysis including parameter optimization and supplementary control branches are used to improve system damping characteristics with specific frequency range of the wind turbines, so as to suppress oscillation.

In the prior art, there are at least the following defects: the oscillation suppression method based on frequency domain can only suppress the oscillation within narrow frequency band, i.e., at a certain frequency point and its neighborhood. When the oscillation frequency changes, that is, it is not within the narrow frequency band, it is necessary to re-optimize the parameters or re-adjust the additional control branch to suppress the oscillation, which is difficult to meet the suppression requirements in the scenario of oscillation frequency varying. Thus, the existing oscillation suppression method is complex, the applicable frequency band is narrow, the applicability is poor, and the efficiency is low.

SUMMARY

Based on the above analysis, embodiments of present application aim to propose an oscillation active damping control method and system for grid-tied type-4 wind turbine generator, which is used to solve the problem that the prior art cannot effectively suppress the oscillation of the grid-tied type-4 wind turbine generator in the scenario of oscillation frequency varying.

On the one hand, an embodiment of present application provides an oscillation active damping control method for grid-tied type-4 wind turbine generator the grid-tied type-4 wind turbine generator consists of multiple subsystems, wherein the method comprising:

based on an interconnection model of the multiple subsystems, constructing a stored energy function and a dissipated energy function of a current inner loop control subsystem, and a first interaction energy function between the current inner loop control subsystem and a DC voltage control subsystem, a second interaction energy function between the current inner loop control subsystem and a phase lock loop subsystem, and a third interaction energy function between the current inner loop control subsystem and a reactive power outer loop control subsystem, then, establishing an energy feedback model of the grid-tied type-4 wind turbine generator;

when the oscillation of the grid-tied type-4 wind turbine generator occurs, obtaining instantaneous angular frequency of a phase lock loop in the phase lock loop subsystem; based on the energy feedback model, adjusting a current reference value of a q-axis current inner loop generated by the reactive power outer loop control subsystem, to make the stored energy function value decrease with time, so as to suppress the oscillation.

Further, constructing energy functions based on the interconnection model of the multiple subsystems, a small signal model of d-axis of the current inner loop control subsystem and a small signal model of q-axis of the current inner loop control subsystem;

the expression of the stored energy function is:

$V_{s\_\mathrm{acc}}=\frac{1}{2}{L}_1\left(\Delta I_{1d}^2+\Delta I_{1d}^2\right)+\frac{1}{2}{C}_i\left(\Delta U_{\mathrm{id}}^2+\Delta U_{\mathrm{iq}}^2\right),$

where V_{s_acc} is the stored energy function, L₁ is filter inductance in the current inner loop control subsystem, ΔI_1d is d-axis perturbated current in the filter inductance L₁, ΔI_1q is q-axis perturbated current in the filter inductance L₁,

$C_i=\frac{1}{k_i\times k_{\mathrm{pwm}}},$

k_i is an integral gain of a PI controller in the current inner loop control subsystem, k_pwm is gain of a converter, ΔU_1d is perturbated state variable in an integral link of a d-axis PI controller in the current inner loop control subsystem, ΔU_1q is perturbated state variable in an integral link of a q-axis PI controller in the current inner loop control subsystem;

the expression of the dissipation energy function is:

V _{d_acc} =−R _p∫Δ_1d²dt−R_p∫ΔI_1q²dt,

where V_{d_acc} is the dissipation energy function, R_p=k_p×k_pwm, k_p is proportional gain of the PI controller in the current inner loop control subsystem, k_pwm is the gain of the converter;

the expression of the first interaction energy function is:

V _{t_dc} =∫ΔI _dref ΔU _id dt+∫ΔI _dref ΔI _1d dt,

where V_{t_dc} is the first interaction energy function, ΔI_dref is perturbated current reference value of the d-axis current inner loop generated by the DC voltage control subsystem;

the expression of the second interaction energy function is:

V _{t_pll} =−k ₁(∫Δθ_pllΔU_iddt+R_p∫ΔI_1ddt)+k₂(∫Δθ_pllΔU_iqdt+R_p∫Δθ_pllΔI_1qdt)+k₃∫Δθ_pllΔI_1ddt+k₄∫Δθ_pllΔ_1qdt,

where V_{t_pll} is the second interaction energy function, Δθ_pll is perturbated phase of the phase lock loop in the phase lock loop subsystem, k₁, k₂, k₃ and k₄ are constants whose expressions are shown below:

$\{\begin{array}{c}{k}_1=-I_{1q0}\\ k_2=I_{1d0}\\ k_3=-U_{pwmq0}+U_{pccq0}+{\omega }_0{L}_1{I}_{1d0}\\ k_4=U_{pwmd0}-U_{pccd0}+{\omega }_0{L}_1{I}_{1q0}\end{array},$

where I_1d0 is a steady-state value of d-axis current in the filter inductance L₁, I_1q0 is a steady-state value of q-axis current in the filter inductance L₁, U_pwmd0 is a steady-state value of the d-axis component of converter port voltage in the current inner loop control subsystem, U_pwmq0 is a steady-state value of the q-axis component of the converter port voltage of the current inner loop control subsystem, U_pccq0 is a steady-state value of d-axis voltage at a PCC (Point of Common Coupling), U_pccd0 is a steady-state value of q-axis voltage at the PCC, and ° ° is rated angular frequency of the phase lock loop (PLL);

the expression of the third interaction energy function is:

V _{t_q} =∫ΔI _qref ΔU _iq dt+∫R _p ΔI _qref ΔI _1q dt,

where V_{t_q} is the third interaction energy function, N V is perturbated current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem.

Further, the small signal model of d-axis of the current inner loop control subsystem is:

$\{\begin{array}{c}{C}_i\frac{d\Delta U_{\mathrm{id}}}{\mathrm{dt}}=\Delta I_{1d}+F_{Cd}\\ L_1\frac{d\Delta I_{1d}}{dt}=-R_p\Delta I_{1d}+\Delta U_{\mathrm{id}}+F_{Ld}\\ F_{Cd}=\Delta I_{\mathrm{dref}}-k_1\Delta {\theta }_{\mathrm{pll}}\\ F_{Ld}=k_3\Delta \theta +\left(\Delta I_{\mathrm{dref}}-k_1\Delta {\theta }_{\mathrm{pll}}\right)\end{array},$

the small signal model of the q-axis of the current inner loop control subsystem is given as:

$\{\begin{array}{c}{C}_i\frac{d\Delta U_{\mathrm{iq}}}{dt}=-\Delta I_{1q}+F_{Cq}\\ L_1\frac{d\Delta I_{1q}}{dt}=-R_p\Delta I_{1q}+\Delta U_{\mathrm{iq}}+F_{Lq}\\ F_{Cq}=\Delta I_{\mathrm{qref}}-k_2\Delta {\theta }_{\mathrm{pll}}\\ F_{Lq}=k_4\Delta {\theta }_{\mathrm{pll}}+R_p\left(\Delta I_{\mathrm{qref}}+k_2\Delta {\theta }_{\mathrm{pll}}\right)\end{array}.$

Further, the energy feedback model is given as:

V _{s_acc} =V _{d_acc} +V _{t_dc} +V _{t_pll} +V _{t_q}.

Further, according to the following compensation function, adjusting the current reference value of the q-axis the current inner loop generated by the reactive power outer loop control subsystem:

dI _qref =−I _dref×∫ω_plldt,

where dI_qref is adjusted current reference value of the q-axis current inner loop, I_dref is instantaneous current reference value of the d-axis current inner loop generated by the DC voltage control subsystem, ω_pll is deviation of the angular frequency measured by the phase lock loop, that is, the deviation between the instantaneous angular frequency measured by the phase lock loop and the rated angular frequency.

Further, the deviation of the angular frequency measured by the phase lock loop and the perturbated phase of PLL in the phase lock loop subsystem satisfy the following relationship:

Δθ_pll=∫ω_plldt,

where Δθ_pll is the perturbated phase of PLL in the phase lock loop subsystem, ω_pll is the deviation of the angular frequency measured by the phase lock loop.

based on the above relationship and the compensation function, the reduction of the third interaction energy function value generated by adjusted current reference value of the q-axis current inner loop which is generated by the reactive power outer loop control subsystem is offset by the increment of the second interaction energy function value that induced by the perturbated phase of the phase lock loop when the oscillation occurs, to make the stored energy function decrease with time, so as to suppress the oscillation.

Further, the oscillation comprises oscillation in sub-synchronous frequency band, from 2.5 Hz to 50 Hz, or oscillation in super-synchronous frequency band, from 50 Hz to 100 Hz.

On the other hand, another embodiment of present application provides an oscillation active damping control system for grid-tied type-4 wind turbine generator, comprising:

a model building module is configured to base on an interconnection model of multiple subsystems, construct a stored energy function, and a dissipated energy function of the current inner loop control subsystem in the multiple subsystems, and a first interaction energy function between the current inner loop control subsystem and a DC voltage control subsystem, a second interaction energy function between the current inner loop control subsystem and a phase-locked loop subsystem, and a third interaction energy function between the current inner loop control subsystem and a reactive power outer loop control subsystem, then, establish an energy feedback model;

an adjustment module is configured to obtain the instantaneous angular frequency of a phase lock loop in the phase lock loop subsystem when the oscillation of the grid-tied type-4 wind turbine generator occurs, based on the energy feedback model, adjust a current reference value of a q-axis current inner loop generated by the reactive power outer loop control subsystem, to make a value of the stored energy function decrease with time, so as to suppress the oscillation.

Further, the adjustment module is configured to adjust a current reference value of a q-axis current inner loop based on the following compensation function.

dI _qref =−I _dref×∫ω_plldt,

where dI_qref is adjusted current reference value of the q-axis current inner loop, I_dref is instantaneous current reference value of a d-axis current inner loop generated by the DC voltage control subsystem, ω_pll is deviation of angular frequency measured by the phase lock loop, that is, the deviation between the instantaneous angular frequency measured by the phase lock loop and rated angular frequency.

Further, the oscillation active damping control system comprises a monitoring module is configured to measure the angular frequency of the phase lock loop in real time.

Compared with prior arts, embodiment of present application has the follow contributions:

In the proposed oscillation active damping control method and system for grid-tied type-4 wind turbine generator, based on the stored energy and the dissipated energy of the current inner loop control subsystem, and the interaction energy between current inner loop control subsystem and other subsystems, an energy feedback model is constructed. By monitoring the angular frequency of phase lock loop in real time, adjusting the current reference value of the q-axis current inner loop to make the stored energy decrease over time, so as to suppress oscillation. By adjusting the variation rate of stored energy to negative, this method ensures the oscillation amplitude of wind turbine generator show an attenuation trend, no matter how the oscillation frequency changes in the sub-synchronous frequency band and super-synchronous frequency band, so as to effectively suppress oscillation.

In the proposed method oscillation active damping control method and system for grid-tied type-4 wind turbine generator, when the oscillation frequency of the wind turbine generator changes, the angular frequency of the phase lock loop also changes correspondingly, thus, by monitoring the angular frequency of the phase lock loop in real time, the instantaneous deviation between the instantaneous angular frequency measured by phase lock loop and rated angular frequency is obtained. And the current reference value of the q-axis current inner loop is adjusted by this deviation, so the oscillation of the wind turbine generator can be suppressed automatically, this method is not only easy to implement but also is effective.

In the application, the above technical schemes can also be combined with each other to realize more preferred combination schemes. Other features and advantages of the application will be described in the following instructions, and some advantages may become apparent from the instructions or be understood by implementing the application. The object and other advantages of the application can be realized and obtained through the contents specially pointed out in the instructions and the accompanying drawings.

BRIEF DESCRIPTION OF DRAWINGS

The attached figures are only for the purpose of illustrating specific embodiments, and are not considered to limit the present application. In the whole figures, the same reference symbols indicate the same components.

FIG. 1 is a flowchart of the oscillation active damping control method for grid-tied type-4 wind turbine generator according to an embodiment of the present application;

FIG. 2 is a schematic structure diagram of the grid-tied type-4 wind turbine generator according to an embodiment of the present application;

FIG. 3 is a schematic diagram of the interconnection model between different sub-systems of the grid-tied type-4 wind turbine generator according to an embodiment of the present application;

FIG. 4 is a schematic diagram of the energy feedback model constructed according to an embodiment of the present application;

FIG. 5 is a schematic structure diagram of the oscillation active damping control method for grid-tied type-4 wind turbine generator according to an embodiment of the present application;

FIG. 6 is a schematic diagram of the oscillation active damping control system for grid-tied type-4 wind turbine generator according to an embodiment of the present application;

FIG. 7 shows wave recordings of active power obtained before and after the oscillation active damping control for oscillation with frequency of around 34 Hz in sub-synchronous frequency band according to an embodiment of the present application;

FIG. 8 shows wave recordings of angular frequency of phase lock loop obtained before and after the oscillation active damping control for oscillation with frequency of around 34 Hz in sub-synchronous frequency band according to an embodiment of the present application;

FIG. 9 shows wave recordings of damping link dI_qref obtained before and after the oscillation active damping control for oscillation with frequency of around 34 Hz in sub-synchronous frequency band according to an embodiment of the present application;

FIG. 10 shows the results of FFT analysis of angular frequency of the phase lock loop obtained before and after the oscillation active damping control for oscillation with frequency of around 34 Hz in sub-synchronous frequency band according to an embodiment of the present application;

FIG. 11 is a schematic diagram of variation trajectory of total feedback energy obtained before and after the oscillation active damping control for oscillation with frequency of around 34 Hz in sub-synchronous frequency band according to an embodiment of the present application;

FIG. 12 shows wave recordings of active power obtained before and after the oscillation active damping control for oscillation with frequency of around 60 Hz in super-synchronous frequency band according to an embodiment of the present application;

FIG. 13 shows wave recordings of angular frequency of the phase lock loop obtained before and after the oscillation active damping control for oscillation with frequency of around 60 Hz in super-synchronous frequency band according to an embodiment of the present application;

FIG. 14 shows wave recordings of damping link dI_qref obtained before and after the oscillation active damping control for oscillation with frequency of around 60 Hz in super-synchronous frequency band according to an embodiment of the present application;

FIG. 15 shows the results of FFT analysis of angular frequency of the phase lock loop obtained before and after the oscillation active damping control for oscillation with frequency of around 60 Hz in super-synchronous frequency band according to an embodiment of the present application;

FIG. 16 is a schematic diagram of variation trajectory of total feedback energy obtained before and after the oscillation active damping control for oscillation with frequency of around 60 Hz in super-synchronous frequency band according to an embodiment of the present application;

wherein: 100—model building module, 120—adjustment module, 130—monitoring module.

DESCRIPTION OF EMBODIMENTS

The preferred embodiment of the present disclosure is described in detail with reference to the attached drawings. The accompanying drawings form a part of the present application and are used together with embodiments of the present disclosure to explain the principles of the disclosure and are not intended to limit the scope of the disclosure.

One embodiment of present application proposes an oscillation active damping control method for grid-tied type-4 wind turbine generator. As shown in FIG. 1, the method includes the following two steps.

Since the permanent magnet synchronous generator (PMSG) and machine-side converter (MSC) in type-4 wind turbine generator have little influence on the sub/super-synchronous oscillation, the sub/super-synchronous oscillation is mainly related to the grid-side converter (GSC) in type-4 wind turbine generator. Where the grid-tied type-4 wind turbine generator contains a current inner loop control subsystem, a DC voltage control subsystem, a PLL subsystem, a reactive power outer control loop subsystem and a power grid subsystem.

S110: based on an interconnection model of the multiple subsystems, constructing a stored energy function and a dissipated energy function of the current inner loop control subsystem, and a first interaction energy function between the current inner loop control subsystem and a DC voltage control subsystem, a second interaction energy function between the current inner loop control subsystem and a phase lock loop subsystem and a third interaction energy function between the current inner loop control subsystem and a reactive power outer loop control subsystem, then, establishing an energy feedback model of the grid-tied type-4 wind turbine generator;

S120: when the oscillation of the grid-tied type-4 wind turbine generator occurs, obtaining the instantaneous angular frequency of a phase lock loop in the phase lock loop subsystem. And based on the energy feedback model, adjusting the current reference value of q-axis of the current inner loop generated by the reactive power outer loop control subsystem, to make the value of the stored energy function decrease with time, so as to suppress the oscillation.

The beneficial effects of this embodiment are: in the sub-synchronous band or super-synchronous band, no matter how the oscillation frequency of the wind turbine generator varies, it can ensure that the corresponding oscillation amplitude shows an attenuation trend by adjusting the stored energy function to decrease with time, so as to suppress the oscillation. The principle based on the above embodiment is explained:

The each previously mentioned subsystem in the interconnection model can all be unified as the following general form:

$\{\begin{array}{c}C\frac{d\Delta U}{\mathrm{dt}}=-K_L\Delta I+F_C\\ L\frac{d\Delta I}{dt}=-K_R\Delta I+K_C\Delta U+F_L\end{array},$

Where C is equivalent capacitance in the general form of subsystem, L is equivalent inductance in the general form of subsystem, K_R is equivalent resistance in the general form of subsystem, ΔU and ΔI are an equivalent voltage variable and equivalent current variable in the general form of subsystem, respectively. K_C is the magnification of the equivalent voltage variable ΔU. K_L is the magnification of the equivalent current variable ΔI. F_C represents the disturbance to the equivalent voltage variable ΔU of this subsystem from other subsystems in all interaction links acting on this subsystem. F_L represents the disturbance to the equivalent current variable ΔI of this subsystem from other subsystems in all interaction links acting on this subsystem.

First, divide the first equation by the second equation in formula (1). Then, according to the first integration principle, conduct cross multiplication and then do integration on both sides of obtained equation, so that the following results is obtained:

$\frac{1}{2}·L·K_L·\Delta I^2+\frac{1}{2}·C·K_C·\Delta U^2+K_L{K}_R\int \Delta I^2\mathrm{dt}-K_L\int F_L\Delta I\mathrm{dt}-K_C\int F_C\Delta U\mathrm{dt}=\mathrm{Cons},$

where Cons is a constant term.

The general form of energy function for each subsystem is:

V=V _s −V _d −V _t,

$\begin{array}{cc}\{\begin{array}{c}{V}_s=\frac{1}{2}C·K_C·\Delta U^2+\frac{1}{2}L·K_L·\Delta I^2\\ V_d=-K_L{K}_R\int \Delta I^2\mathrm{dt}\\ V_t=K_C\int F_C\Delta \mathrm{Udt}+K_L\int F_L\Delta \mathrm{Idt}\end{array}& \left(2\right)\end{array}$

where V_s is the general form of the stored energy. V_d is general form of the dissipation energy. V_t is the general form of interaction energy between the subsystems, where the 1st term in V_t is the general form of interaction energy via interaction link F_C, and the 2st term is the general form of interaction energy via interaction link F_L.

Calculate the partial derivative of the energy function V of the subsystem with respect to time t.

$\stackrel{.}{V}=\frac{\partial V}{\partial \Delta U}·\frac{d\Delta U}{dt}+\frac{\partial V}{\partial \Delta I}·\frac{d\Delta I}{dt}+\frac{\partial V}{\partial t}=\left(K_C{F}_C·\Delta U-K_L{K}_C\Delta U·\Delta I\right)+\left(K_L{F}_L·\Delta I+K_L{K}_C·\Delta U\Delta I-K_L{K}_R\Delta I^2\right)+\left(K_L{K}_R·\Delta I^2-K_C{F}_C·\Delta U-K_L{F}_L·\Delta I\right)=0,$

It can be seen that, the derivative of energy function V with respect to time t is 0, which proves that the system conforms to the energy conservation. According to the energy conservation, the stored energy is equal to the sum of the dissipation energy and the interaction energy, i.e.,

V _s =V _d +V _t.

When the wind turbine generator is disturbed causing sub/super-synchronous oscillation with mode α+jω_c, the corresponding oscillation component in the equivalent voltage variable ΔU and the equivalent current variable ΔI can be expressed as:

ΔU=A_Ue^αt cos(ω_ct+θ_U)

ΔI=A₁e^αt cos(ω_ct+θ₁)

Where α is a damping factor of the wind turbine generator corresponding to the oscillation mode, ω_c is an oscillation frequency corresponding to the oscillation mode, A_U is an amplitude of voltage component of the corresponding subsystem, θ_U is an initial phase of voltage component, A₁ is an amplitude of current component of the corresponding subsystem, θ₁ is an initial phase of current component.

When the wind turbine generator is disturbed causing sub/super-synchronous oscillation with mode α+jω_c, the corresponding oscillation component in disturbance F_C can be expressed as:

F _C =A _FC e ^αt cos(ω_ct+θ_FC),

When the wind turbine generator is disturbed causing sub/super-synchronous oscillation with mode α+jω_c, the corresponding oscillation component in disturbance F_L can be expressed as:

F _L =A _FL e ^αt cos(ω_ct+θ_FL),

Where A_FC is the amplitude of the disturbance to the equivalent voltage variable ΔU in interaction links, θ_FC is the initial phase of the disturbance to the equivalent voltage variable ΔU in interaction links, Δ_FL is the amplitude of the disturbance to the equivalent current variable ΔI in interaction links, θ_FL is the initial phase of the disturbance to the equivalent current variable ΔI in interaction links.

Apply the above expressions (ΔU, ΔI, F_C, F_L) to the stored energy function V_S, and calculate the partial derivative with respect to time t, the derivative of V_S to t, i.e., {dot over (V)}_s(t) is given as:

{dot over (V)} _s(t)={dot over (V)}_{s_ac}(t)+{dot over (V)}_{s_dc}(t),

where {dot over (V)}_{s_ac}(t) and {dot over (V)}_{s_dc}(t) are periodic and non-periodic components of the derivative of the stored energy function respectively. And the expressions are given as:

$$\begin{gathered} {\stackrel{.}{V}}_{s\_\mathrm{ac}}\left(t\right)=C·K_C·A_U^2\left(\frac{\alpha \cos \left(2{\omega }_{c}t+2{\theta }_U\right)-{\omega }_c·\sin \left(2{\omega }_{c}t+2{\theta }_U\right)}{2}\right)e^{2\alpha t}+L·K_L·A_I^2\left(\frac{\alpha ·\cos \left(2{\omega }_{c}t+2{\theta }_I\right)-{\omega }_c·\sin \left(2{\omega }_{c}t+2{\theta }_I\right)}{2}\right)e^{2\alpha t},\text{ \\ V.s⁢\_⁢dc(t)=λs⁢e2⁢α⁢t=α⁡(C·KC·AU2+L·KL·AI22)⁢e2⁢α⁢t,} \end{gathered}$$

Since term

$\left(\frac{C·K_C·A_U^2+L·K_L·A_I^2}{2}\right)$

in {dot over (V)}_{s_dc}(t) is constantly positive, the analytical relationship between the real part α of sub/super-synchronous oscillation mode a-jo, and the change rate of the stored energy function V_s is established, and it is proved that no matter how the oscillation frequency ω_c changes, the oscillation divergence and convergence processes correspond to the increasing and decreasing processes of the stored energy function V_s respectively.

Specifically, if the non-periodic component of the stored energy function V_s increases progressively, it means λ_S>0 and the real part of eigenvalue (feature value) α>0 in the {dot over (V)}_{s_dc}(t), i.e., oscillation of the wind turbine generator diverges. If the non-periodical component decreases progressively, it means λ_S<0 and the real part of eigenvalue α<0 in {dot over (V)}_{s_dc}(t), i.e., the oscillation of the wind turbine generator converges. If the non-periodical component neither increases nor decreases, it means λ_S=0 and the real part of eigenvalue α=0 in {dot over (V)}_{s_dc}(t), i.e., the oscillation of the wind turbine generator has a constant amplitude.

Based on the above analysis, it is worth emphasizing that the variation trend of the stored energy of the wind turbine generator correlates strictly positively with the variation trend of oscillation amplitude. Moreover, the variation trend of the stored energy can be controlled by the dissipation energy and the interaction energy. Therefore, by controlling the stored energy of the current inner control loop subsystem of the grid-side converter to show a decreasing trend, the oscillation of the wind turbine generator can be suppressed. Thus, the damping control strategy is inspired. Specifically, by making the stored energy decrease progressively, the real part α of the corresponding sub/super-synchronous oscillation mode can be kept negative. In this case, no matter how the oscillation frequency ω_c varies, the amplitude A₁e^α1 of sub/super-synchronous oscillation component A₁e^αt cos(ω_ct+φ) of the output current of GSC (grid-side converter) always exhibits an attenuation trend. In other words, the WTG (wind turbine generator) does not inject continuously diverging oscillation component in current to the power grid, thus the interaction between the wind turbine generator and power grid is blocked, and the active damping control in a broad frequency band can be realized.

The structure of grid-tied type-4 wind turbine generator is shown in FIG. 2. It consists of generator-side components which include a permanent magnetic synchronous generator (PMSG) and a machine-side converter (MSC), and grid-side components which include a grid-side converter (GSC), DC bus, an AC filter inductance L₁, an equivalent resistance R₁, a filter capacitance C_t, and an equivalent grid inductance L₂ and a resistance R₂.

Since the permanent magnet synchronous generator (PMSG) and machine-side converter (MSC) in type-4 wind turbine generator have little influence on the sub/super-synchronous oscillation, the sub/super-synchronous oscillation is mainly related to the grid-side converter (GSC) in type-4 wind turbine generator. Accordingly, the interconnection model of the multiple subsystems based on the structure of the grid-tied type-4 wind turbine generator is shown in FIG. 3. The current inner control loop subsystem of the GSC includes an inner loop PI controller, the L₁ and the R₁. The DC voltage control loop subsystem includes a DC voltage PI controller and the DC bus. The PLL subsystem includes a PI controller and a voltage controlled oscillator. The reactive power outer loop control subsystem has no dynamic process in unit power factor control since I_qref is 0. And the equivalent power grid subsystem includes the C_t, L₂ and R₂. In FIG. 3, each arrow represents the interaction link between the corresponding two subsystems, and the symbol on the arrow represents the variable in the expression of the interaction link. P is the active power output of GSC. I₁ is the current output of GSC, that is the current in the inductance L₁. I_dref is the current reference value of the d-axis current inner loop generated by the DC voltage control subsystem. I_qref is the current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem. U_pcc is the voltage at PCC (point of common coupling). θ_pll is the phase of U_pcc measured by the PLL.

In step 110, constructing the energy functions based on the interconnection model of multiple subsystems, a small signal model of d-axis of the current inner loop control subsystem and a small signal model of q-axis of the current inner loop control subsystem.

Specifically, according to the mathematical model of the subsystem, taking into account voltage feedforward decoupling item of the power grid and cross decoupling terms of dq axes, the small signal model of d-axis of the current inner loop control subsystem is given as:

$\{\begin{array}{c}{C}_i\frac{d\Delta U_{\mathrm{id}}}{dt}=-\Delta I_{1d}+F_{\mathrm{Cd}}\\ L_1\frac{d\Delta I_{1d}}{dt}=-R_p\Delta I_{1d}+\Delta U_{\mathrm{id}}+F_{\mathrm{Ld}}\\ F_{\mathrm{Cd}}=\Delta I_{\mathrm{dref}}-k_1\Delta {\theta }_{\mathrm{pll}}\\ F_{\mathrm{Ld}}=k_3\mathrm{\Delta \theta }+\left(\Delta I_{\mathrm{dref}}-k_1\Delta {\theta }_{\mathrm{pll}}\right),\end{array}$

The small signal model of q-axis of the current inner loop control subsystem is given as:

$\{\begin{array}{c}{C}_i\frac{d\Delta U_{\mathrm{iq}}}{\mathrm{dt}}=-\Delta I_{1q}+F_{Cq}\\ L_1\frac{d\Delta I_{1q}}{\mathrm{dt}}=-R_p\Delta I_{1q}+\Delta U_{\mathrm{iq}}+F_{Lq}\\ F_{Cq}=\Delta I_{\mathrm{qref}}-k_2\Delta {\theta }_{\mathrm{pll}}\\ F_{Lq}=k_4\Delta {\theta }_{\mathrm{pll}}+R_p\left(\Delta I_{\mathrm{qref}}+k_2\Delta {\theta }_{\mathrm{pll}}\right).\end{array}$

Specifically, the obtained small signal models of d-axis and q-axis of the current inner loop control subsystem are compared with the general form of the mathematical model of the subsystem, and the detailed expression of state variables and interaction links of the small signal models of the d-axis and the q-axis of the current inner loop control subsystem are substituted into the general expression of the energy function to construct the detailed energy functions.

Specifically, the expression of the stored energy function is given as:

$V_{s\_\mathrm{acc}}=\frac{1}{2}{L}_1\left(\Delta I_{1d}^2+\Delta I_{1q}^2\right)+\frac{1}{2}{C}_i\left(\Delta U_{\mathrm{id}}^2+\Delta U_{\mathrm{iq}}^2\right),$

Where V_{s_acc} is the stored energy function, L₁ is the filter inductance in the current inner loop control subsystem, ΔI_1d is the d-axis perturbated current in filter inductance L₁, ΔI_1q is the q-axis perturbated current in the filter inductance L₁,

$C_i=\frac{1}{k_i\times k_{\mathrm{pwm}}},$

k_i is the integral gain of the PI controller in the current inner loop control subsystem, k_pwm is gain of a converter, ΔU_1d is perturbated state variable of integral link of d-axis PI controller in the current inner loop control subsystem, ΔU_1q is perturbated state variable of integral link of q-axis PI controller in the current inner loop control subsystem.

The expression of the dissipation energy function is given as:

V _{d_acc} =−R _p ∫ΔI _1d ² dt−R _p ∫ΔI _1q ² dt,

Where V_{d_acc} is the dissipation energy function, R_p=k_p×k_pwm, k_p is proportional gain of the PI controller in the current inner loop control subsystem, k_pwm is the gain of the converter.

The expression of the first interaction energy function is given as:

V _{t_dc} =∫ΔI _dref ΔU _id dt+∫ΔI _dref ΔI _1d dt,

Where V_{t_dc} is the first interaction energy function, ΔI_dref is perturbated current reference value of the d-axis current inner loop generated by the DC voltage control subsystem.

The expression of the second interaction energy function is given as:

V _{t_pll} =−k ₁(∫Δθ_pllΔU_iddt+R_p∫Δθ_1ddt)+k₂(∫Δθ_pllΔU_iqdt+R_p∫Δθ_pllΔI_1qdt)+k₃∫Δθ_pllΔI_1ddt+k₄∫Δθ_pllΔ_1qdt,

Where V_{t_pll} is the second interaction energy function, Δθ_pll is perturbated phase measured by the phase lock loop in the phase lock loop subsystem, k₁, k₂ k₃, and k₄ are constants whose expressions are shown below.

$\{\begin{array}{c}{k}_1=-I_{1q0}\\ k_2=I_{1d0}\\ k_3=-U_{pwmq0}+U_{pccq0}+{\omega }_0{L}_1{I}_{1d0}\\ k_4=U_{pwmd0}-U_{pccd0}+{\omega }_0{L}_1{I}_{1q0},\end{array}$

Where I_1d0 is a steady-state value of d-axis current in the filter inductance L₁, I_1q0 is a steady-state value of q-axis current in the filter inductance L₁, U_pwmd0 is a steady-state value of the d-axis component of converter port voltage in the current inner loop control subsystem, U_pwmq0 is a steady-state value of the q-axis component of the converter port voltage in the current inner loop control subsystem, U_pccd0 is a steady-state value of d-axis voltage at PCC (point of common coupling), U_pccq0 is a steady-state value of q-axis voltage at PCC (point of common coupling), and ω₀ is rated angular frequency of the phase lock loop (PLL), ω₀=100π. The converter means the GSC in this embodiment.

The expression of the third interaction energy function is given as:

V _{t_q} =∫ΔI _qref ΔU _iq dt+∫R _p ΔI _qref ΔI _1q dt,

where V_{t_q} is the third interaction energy function, ΔI_qref is perturbated current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem.

The energy feedback model is given as follow based on the above energy functions.

V _{s_acc} =V _{d_acc} +V _{t_dc} +V _{t_pll} +V _{t_q}.

Specifically, according to the summation relationship of different energies in expression of the energy conservation, the energy feedback model is constructed by taking the stored energy of the current inner loop control subsystem as the controlled object, and taking the internal dissipation energy of the current inner loop control subsystem and the interaction energy between the current inner loop control subsystem and the DC voltage control subsystem, the phase lock loop subsystem, the reactive outer loop control subsystem as the feedback channels. Accordingly, there are totally four feedback channels as shown in FIG. 4: 1) The active power energy feedback channel formed by the first interaction energy between the DC voltage control subsystem and the current inner loop control subsystem (i.e., {circle around (1)} in FIG. 4); 2) the PLL energy feedback channel formed by the second interaction energy between the PLL subsystem and the current inner loop control subsystem (i.e., {circle around (2)} in FIG. 4); 3) the reactive power energy feedback channel formed by the third interaction energy between the reactive power outer loop control subsystem and the current inner loop control subsystem (i.e., {circle around (3)} in FIG. 4); 4) The dissipation energy feedback channel formed by the dissipation energy of the current inner loop control subsystem itself (i.e., {circle around (4)} in FIG. 4).

The energy characteristics of each feedback channel are analyzed below respectively:

For the first interaction energy function V_{t_dc} corresponding to {circle around (1)} the active power energy feedback channel, according to the small signal model of d-axis of the current inner loop control subsystem, the dynamic relationship between ΔI_1d, ΔI_dref and ΔU_id are:

$\Delta U_{\mathrm{id}}=\frac{-\Delta I_{1d}+\Delta I_{\mathrm{dref}}}{C_{i}s}\Delta I_{\mathrm{dref}}=\frac{L_1{C}_i{s}^2+R_p{C}_{i}s+1}{C_{i}s+1}\Delta I_{1d}$

Applying them to the expression of the first interaction energy function V_{t_dc} corresponding to the active power energy feedback channel yields:

$V_{t\_\mathrm{dc}}=\int \frac{\Delta I_{\mathrm{dref}}}{C_{i}s}\Delta I_{\mathrm{dref}}\mathrm{dt}+\int \left(R_p+L_{1}s+\frac{1}{C_{i}s}\right)\frac{C_{i}s-1}{C_{i}s+1}\Delta I_{1d}^2\mathrm{dt}$

For the 1^st term in the above equation, because there is the integral part

$\frac{1}{C_{i}s},$

phase difference of integration variable is 90°. Thus energy in the 1^st term is 0 and not affect the stability of the wind turbine generator, it can be negligible. But for the 2^nd term in the above equation, the sum V_sum of the 2^nd term in the above equation and the 1^st term (i.e., −R_p∫ΔI_1d²dt) in the dissipation energy function V_{d_acc} is calculated:

$V_{\mathrm{sum}}=\int \underset{G\left(s\right)}{\underbrace{\left(\left(R_p+L_{1}s+\frac{1}{C_{i}s}\right)\frac{C_{i}s-1}{C_{i}s+1}-R_p\right)}}\Delta I_{1d}^2\mathrm{dt}$

whether energy V_sum above increasing or decreasing with time depends on the phase value of transfer function G(s). When the phase of G(s) falls in the range of [−90°, 90°], V_sum increases with time; when the phase of G(s) falls in the range of [90°, 270°], V_sum decreases with time. And, by applying jω to G(s) and judging the sign (positive or negative) of the real part of G(jω) (i.e., Re(G(jω)), the phase of G(s) can be obtained.

$\mathrm{Re}\left(G\left(j\omega \right)\right)=-\frac{2\left(\left(R_p-1\right)+{\omega }^2{L}_1{C}_i\right)}{{\omega }^2{C}_i^2+1},$

In order to guarantee the response speed of the current inner loop, usually R_p>1. Thus, Re(G(jω))<0, the phase of G(s) remains in [90°, 270° ], and V_sum decreases with time. In other words, the first interaction energy function V_{t_dc} corresponding to the active power energy channel can be offset by d-axis dissipation energy −R_p∫ΔI_1d²dt of the current inner loop control subsystem, and thus it has no adverse effect on the stability of wind turbine generator.

For the second interaction energy function V_{t_pll} corresponding to {circle around (2)} the PLL energy feedback channel, when the Type-4 WTG operates in unit power factor, the q-axis steady-state component I_1q0 of the current flowing through inductance L₁ is 0 in control coordinate, i.e., I_1q0=0. Thus in the expression of V_{t_pll}, constant term k₁ is 0. And the values of constant terms k₃ and k₄ can be analyzed according to the following dynamic equation of the inductance L₁:

$\{\begin{array}{c}{L}_1\frac{d{I}_{1d}}{\mathrm{dt}}={\omega }_0{L}_1{I}_{1q}+U_{\mathrm{pwmd}}-U_{\mathrm{pccd}}\\ L_1\frac{d{I}_{1q}}{\mathrm{dt}}=-{\omega }_0{L}_1{I}_{1q}+U_{\mathrm{pwmq}}-U_{\mathrm{pccq}}\end{array}$

where I_1d and I_1q are d-axis and q-axis current flowing through the inductance L₁; U_pwmq and U_pwmq are d-axis and q-axis voltage of the GSC; U_pccd and U_pccq are d-axis and q-axis voltage at the PCC.

The steady-state value of the differential term in the above equation is 0, thus:

$\{\begin{array}{c}{\omega }_0{L}_1{L}_{1q0}+U_{\mathrm{pwmd}0}-U_{\mathrm{pccd}0}=0\\ -{\omega }_0{L}_1{L}_{1q0}+U_{\mathrm{pwmq0}}-U_{\mathrm{pccq}0}=0\end{array}$

Comparing the above equation with the expressions of k₃ and k₄, it can be seen that k₃ and k₄ are both 0. Therefore, the expression of the PLL energy feedback channel V_{t_pll} can be written as:

V _{t_pll} =∫k ₂Δθ_pllΔU_1qdt+∫R_pk₂Δθ_pllΔI_1qdt

The constant term k₂ and state variables Δθ_pll, ΔU_iq and ΔI_1q are not 0, thus V_{t_pll} in oscillating process is no longer 0, and it may have adverse effect on the oscillation.

According to the above analysis, the q-axis component of dissipation energy V_{d_acc} (i.e., −R_p∫ΔI_1q²dt) decreases progressively with time. The sum of d-axis component of dissipation energy V_{d_acc} (i.e., −R_p∫ΔI_1d²dt) and the first interaction energy V_{t_dc} corresponding to the active power energy feedback channel decreases progressively with time. Thus, in order to make the stored energy V_{s_acc} decrease over time so as to suppress oscillation, the second interaction energy V_{t_pll} corresponding to the PLL energy feedback channel and the third interaction energy V_{t_q} corresponding to the reactive power energy feedback channel should be adjusted.

Since the third interaction energy V_{t_q} corresponding to the reactive power energy feedback channel (i.e., the third interaction energy function) is originally 0 when the wind turbine generator is running stably, it can be adjusted flexibly. Therefore, in order to suppress the influence of the second interaction energy V_{t_pll} of the PLL energy feedback channel, by constructing a compensation function in the current reference value I_dref of the q-axis current inner loop, a none-zero additional reactive power energy feedback channel, i.e., dV_{t_q}, can be generated to reversely compensate for the effect of PLL energy feedback channel V_{t_pll} on the stability. When the wind turbine generator is running stably, the third interaction energy V_{t_q} corresponding to the reactive power energy feedback channel is 0, thus when oscillation occurs, the third interaction energy v_{t_q} corresponding to the reactive power energy feedback channel is adjusted, variation rate dV_{t_q} is the adjusted third interaction energy corresponding to the reactive power energy feedback channel.

According to the above idea, in order for the third interaction energy dV_{t_q} generated by the additional reactive power energy feedback channel being offset by the second interaction energy V_{t_pll} generated by the PLL energy feedback channel, parameters such as PLL variable Δθ_pll and constant term k₂ in the expression of V_{t_pll} need be introduced to the compensation function. The following compensation function in the current reference value of the q-axis current inner loop is designed:

dI _qref =−I _dref×∫ω_plldt=−I_dref×Δθ_pll

where on is the deviation of PLL angular frequency from the rated angular frequency, the deviation is 0 when the wind turbine generator is in steady state. I_dref is the current reference value of the d-axis current inner loop at the present moment. The active damping control (that is, oscillation suppression) diagram containing the above compensation function is shown in FIG. 5, where the active damping control containing the compensation function is in the dotted box in the figure.

Specifically, by linearizing dI_qref and considering the steady state value Δθ_pll0 of perturbated phase Δθ_pll measured by the phase lock loop is zero, the linearized result is:

ΔdI_qref=I_dref0×Δθ_pll−Δθ_pll0×ΔI_dref=I_dref0×Δθ_pll,

Applying the ΔdI_qref to the position of ΔI_qref in the expression of V_{t_q}, the expression of the third interaction energy function dV_{t_q} corresponding to the additional reactive power energy feedback channel is:

dV _{t_q} =−I _dref0∫Δθ_pllΔU_iqdt−R_pI_dref0∫Δθ_pllΔI_1qdt

where I_dref0 is the current reference value I_dref of the d-axis current inner loop in the stable state.

The sum of dV_{t_q} and V_{t_pll} is:

V _{t_pll} +dV _{t_q}=(k₂−I_dref0)∫Δθ_pllΔU_iqdt+R_p(k₂−I_dref)∫Δθ_pllΔI_1qdt

where k₂=I_dref0, I_dref0=I_1d0, thus the sum of V_{t_pll} and dV_{t_q} is 0. It means the third interaction energy dV_{t_q} generated by the additional reactive power energy feedback channel can be offset by the second interaction energy generated by the PLL energy feedback channel. In detail, the third interaction energy V_{t_q} generated by the reactive power energy feedback channel is zero before adjusting, the third interaction energy dV_{t_q} is negative after adjusting. In other words, the reduction of the adjusted third interaction energy function value generated by the current reference value of the q-axis current inner loop which is generated by the reactive power outer loop control subsystem can be offset by the increment of the second interaction energy function value generated by the variation rate of perturbated phase of the phase lock loop when the oscillation occurs. On this basis, it can eliminate the adverse effect of the second interaction energy generated by the PLL energy feedback channel on the stored energy.

In summary, the features of the above different feedback energy terms are summarized as follows: 1) the q-axis component of dissipation energy V_{q_acc} (i.e., −R_p∫Δ_1q²dt) decreases progressively with time; 2) the sum of d-axis component of dissipation energy V_{d_acc} (i.e., −R_p∫ΔI_1d²dt) and the first interaction energy V_{t_dc} corresponding to the active power energy feedback channel decreases progressively with time; 3) the sum of the second interaction energy V_{t_pll} corresponding to the PLL energy feedback channel and the third interaction energy dV_{t_q} corresponding to the additional reactive power energy feedback channel is 0. It can be seen that, the total feedback energy decreases progressively with time, thus the stored energy V_{s_acc} decreases progressively with time, and the active damping control of oscillations in a broad frequency band can be realized, that is, the oscillation in a broad frequency band is suppressed.

Based on the above analysis, when the oscillation occurs, adjust the current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem according to the following compensation function.

dI _qref =−I _dref×ω_plldt

Where dI_qref is an adjusted current reference value of the q-axis current inner loop, I_dref is the instantaneous current reference value of the d-axis current inner loop generated by the DC voltage control subsystem, ω_pll is the deviation of the angular frequency measured by phase lock loop, i.e., the deviation between the instantaneous angular frequency measured by phase lock loop and the rated angular frequency.

The mentioned deviation of the angular frequency measured by the phase lock loop and mentioned perturbated phase measured by the phase lock loop in the phase lock loop subsystem satisfy the following relationship.

Δθ_pll=∫ω_plldt,

Where Δθ_pll is perturbated phase measured by the phase lock loop in the phase lock loop subsystem, ω_pll is the deviation of the angular frequency measured by the phase lock loop.

Based on the above relationship and compensation function, the reduction of the adjusted third interaction energy function value generated by the current reference value of the q-axis current inner loop which is generated by the reactive power outer loop control subsystem can be offset by the increment of the second interaction energy function value generated by variation rate of the perturbated phase in the phase lock loop when the oscillation occurs, to make the stored energy function value decrease with time, so as to suppress the oscillation.

The proposed active damping control method can suppress efficiently the oscillations in sub-synchronous frequency band (2.5-50 Hz) and sub-synchronous frequency band (50-100 Hz).

Another embodiment of present application proposes an oscillation active damping control system for grid-tied type-4 wind turbine generator. Since the system embodiments have the same working principle as the method embodiments above, the embodiment of the method above can be referred to for repetition and will not be described here.

Specifically, as shown in FIG. 6, the oscillation active damping control system comprising:

a model building module 110 is configured to base on an interconnection model of multiple subsystems in grid-tied type-4 wind turbine generator, construct a stored energy function and a dissipated energy function of a current inner loop control subsystem of the multiple subsystems, and a first interaction energy function between the current inner loop control subsystem and a DC voltage control subsystem, a second interaction energy function between the current inner loop control subsystem and a phase-locked loop subsystem, and a third interaction energy function between the current inner loop control subsystem and a reactive power outer loop control subsystem. On this basis, establish an energy feedback model of the grid-tied type-4 wind turbine generator;

an adjustment module 120 is configured to obtain the instantaneous angular frequency measured by a phase lock loop in the phase-locked loop subsystem when the oscillation of the grid-tied type-4 wind turbine generator occurs, then based on the energy feedback model, adjust the current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem, to make the stored energy function value decrease with time, so as to suppress the oscillation.

Further, the adjustment module is configured to adjust the current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem according the following compensation function.

dI _qref =−I _dref×ω_plldt,

Where dI_qref is adjusted current reference value of the q-axis current inner loop, I_dref is the instantaneous current reference value of the d-axis current inner loop generated by the DC voltage control subsystem, ω_pll is the deviation of the angular frequency measured by the phase lock loop, i.e., the deviation between the instantaneous angular frequency measured by the phase lock loop and the rated angular frequency.

In addition, the oscillation active damping control system comprises a monitoring module 130 configured to measure the angular frequency of the phase lock loop in real time.

The model building module is configured to construct energy functions based on an interconnection model of the multiple subsystems, a small signal model of d-axis of the current inner loop control subsystem and a small signal model of q-axis of the current inner loop control subsystem, wherein expression of the stored energy function is:

$V_{s\_\mathrm{acc}}=\frac{1}{2}{L}_1\left(\Delta I_{1d}^2+\Delta I_{1q}^2\right)+\frac{1}{2}{C}_i\left(\Delta U_{\mathrm{id}}^2+\Delta U_{\mathrm{iq}}^2\right),$

Where V_{s_acc} is the stored energy function, L₁ is filter inductance in the current inner loop control subsystem, Mid is d-axis perturbated current in the filter inductance L₁, ΔI_1q is q-axis perturbated current in the filter inductance L₁,

$C_i=\frac{1}{k_i\times k_{\mathrm{pwm}}},$

k_i is an integral gain of a PI controller in the current inner loop control subsystem, k_pwm is gain of a converter, ΔU_1d is perturbated state variable in an integral link of a d-axis PI controller in the current inner loop control subsystem, ΔU_1q is perturbated state variable in an integral link of a q-axis PI controller in the current inner loop control subsystem.

The expression of the dissipation energy function is:

V _{d_acc} =−R _p ∫ΔI _1d ² dt−R _p ∫ΔI _1q ² dt,

where V_{d_acc} is the dissipation energy function, R_p=k_p×k_pwm, k_p is the proportional gain of the PI controller in the current inner loop control subsystem, k_pwm is the gain of the converter.

The expression of the first interaction energy function is:

V _{t_dc} =∫ΔI _dref ΔU _id dt+∫∫ΔI _dref ΔI _1d dt

Where V_{t_dc} is the first interaction energy function, ΔI_dref is perturbated current reference value of the d-axis current inner loop generated by the DC voltage control subsystem.

The expression of the second interaction energy function is:

V _{t_pll} =−k ₁(∫Δθ_pllΔU_iddt+R_p∫Δθ_pllΔI_1ddt)+k₂(∫Δθ_pllΔU_iqdt+R_p∫Δθ_pllΔI_1qdt)+k₃∫Δθ_pllΔI_1ddt+k₄∫Δθ_pllΔI_1qdt

Where V_{t_pll} is the second interaction energy function, Δθ_pll is perturbated phase of the phase lock loop in the phase lock loop subsystem, k₁, k₂ k₃, and k₄ are constants whose expressions are shown below:

$\{\begin{array}{c}{k}_1=-I_{1q0}\\ k_2=I_{1d0}\\ k_3=-U_{\mathrm{pwmq}0}+U_{\mathrm{pccq}0}+{\omega }_0{L}_1{I}_{1d0}\\ k_4=U_{\mathrm{pwmd}0}-U_{\mathrm{pccd}0}+{\omega }_0{L}_1{I}_{1q0}\end{array}$

Where I_1d0 is a steady-state value of d-axis current in the filter inductance L₁, I_1q0 is a steady-state value of q-axis current in the filter inductance L₁, U_pwmd0 is a steady-state value of the d-axis component of converter port voltage in the current inner loop control subsystem, U_pwmq0 is a steady-state value of the q-axis component of the converter port voltage in the current inner loop control subsystem, U_pccd0 is a steady-state value of d-axis voltage at a point of common coupling, U_pccq0 is a steady-state value of q-axis voltage at the point of common coupling, and ω₀ is rated angular frequency of the phase lock loop.

The expression of the third interaction energy function is:

V _{t_q} =∫ΔI _qref ΔU _iq dt+∫R _p ΔI _qref ΔI _1q dt

Where V_{t_q} is the third interaction energy function, ΔI_qref is perturbated current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem.

Wherein the small signal model of the d-axis of the current inner loop control subsystem is configured to be:

$\{\begin{array}{c}{C}_i\frac{d{\Delta U}_{\mathrm{id}}}{\mathrm{dt}}=-\Delta I_{1d}+F_C\\ L_1\frac{d{\Delta I}_{1d}}{\mathrm{dt}}=-R_p\Delta I_{1d}+\Delta U_{\mathrm{id}}+F_{\mathrm{Ld}}\\ F_{\mathrm{Cd}}=\Delta I_{\mathrm{dref}}-k_1\Delta {\theta }_{\mathrm{pll}}\\ F_{\mathrm{Ld}}=k_3\mathrm{\Delta \theta }+\left(\Delta I_{\mathrm{dref}}-k_1{\mathrm{\Delta \theta }}_{\mathrm{pll}}\right)\end{array}.$

The small signal model of the q-axis of the current inner loop control subsystem is configured to be:

$\{\begin{array}{c}{C}_i\frac{d{\Delta U}_{\mathrm{iq}}}{\mathrm{dt}}=-\Delta I_{1q}+F_{\mathrm{Cq}}\\ L_1\frac{d{\Delta I}_{1q}}{\mathrm{dt}}=-R_p\Delta I_{1q}+\Delta U_{\mathrm{iq}}+F_{\mathrm{Lq}}{F}_{\mathrm{Cq}}=\Delta I_{\mathrm{qref}}-k_2\Delta {\theta }_{\mathrm{pll}}{F}_{\mathrm{Lq}}=k_4{\mathrm{\Delta \theta }}_{\mathrm{pll}}+R_p\left(\Delta I_{\mathrm{qref}}+k_2{\mathrm{\Delta \theta }}_{\mathrm{pll}}\right)\end{array}$

Wherein the energy feedback model is configured to be:

V _{s_acc} =V _{d_acc} +V _{t_dc} +V _{t_pll} +V _{t_q}.

Wherein the energy feedback model is configured that the deviation of the angular frequency measured by the phase lock loop and perturbated phase of the phase lock loop in the phase lock loop subsystem satisfy the following relationship:

Δθ_pll=∫ω_plldt

Where Δθ_pll is perturbated phase of the phase lock loop in the phase lock loop subsystem, ω_pll is the deviation of the angular frequency measured by the phase lock loop.

Based on the above relationship and the compensation function, the reduction of the third interaction energy function value generated by adjusted current reference value of the q-axis current inner loop which is generated by the reactive power outer loop control subsystem is offset by the increment of the second interaction energy function value that is induced by the perturbated phase of the phase lock loop when the oscillation occurs, to make the stored energy function value decreases with time, so as to suppress the oscillation.

Wherein the oscillation comprises oscillation in sub-synchronous frequency band, from 2.5 Hz to 50 Hz, or oscillation in super-synchronous frequency band, from 50 Hz to 100 Hz.

The beneficial effects of the oscillation active damping control method and system for grid-tied type-4 wind turbine generator proposed in the application are better illustrated by the following embodiment.

To verify the effectiveness and feasibility of the proposed method, the control program corresponding to the oscillation active damping control method for grid-tied type-4 wind turbine generator proposed in the application is embedded to the practical controller of a real 2.5 MW type-4 WTG (wind turbine generator). On this basis, a hardware-in-loop simulation platform is built by combining RTDS (real time digital simulator). By adjusting the short circuit ratio of power grid and converter control parameters, oscillations with frequency of around 34 Hz and around 60 Hz in dq coordinate system are excited respectively in the simulation system. And the effectiveness of the oscillation active damping control method for grid-tied type-4 wind turbine generator proposed in the application in suppressing oscillations within sub-synchronous frequency band (2.5 Hz-50 Hz)/super-synchronous frequency band (50 Hz-100 Hz) can be verified.

The First Embodiment: Effectiveness of the Oscillation in Sub-Synchronous Frequency Band

Set the short circuit ratio of the power grid to be 1.8, increase the integral link of the PLL, and increase the active power of WTG progressively to around 1240 kW. After sub-synchronous oscillation of around 34 Hz occurs, switch on the active damping control. In this case, the waveform recordings of active power, angular frequency of the PLL, the output of damping link dI_qref and the FFT analysis of PLL angular frequency are shown in FIG. 7-10. Meanwhile, collect the variation of state variables required for the calculation of feedback energy before and after the switching of damping control, and calculate the variation trajectory of energy using the trapezoidal integral algorithm. The variation trajectory of total feedback energy (that is, the stored energy) with time is shown in FIG. 11.

According to FIGS. 7 and 8, the active damping control link is switched when the peak values of sub-synchronous oscillation components in the waveforms of active power and PLL frequency reach 65 kW and 1 Hz respectively. After the active damping control is switched on, the oscillation in sub-synchronous frequency band attenuates quickly, and the oscillation amplitude drops below 10% of the original value within 0.2 s. After the waveform being stabilized, there is no longer any sub-synchronous oscillation component. And FIG. 11 shows that, the total feedback energy before the switching of active damping control exhibits positive feedback characteristics (i.e., increases progressively), thus the oscillation diverges. After the active damping control is switched on, the total feedback energy exhibits negative feedback characteristics (i.e., decreases progressively), thus the oscillation is suppressed and converges. The experimental results above are consistent with the theoretical analysis results based on energy, and verify that the active damping control method proposed in the application has a good suppression effectiveness on the oscillation in the sub-synchronous frequency band.

The Second Embodiment: Effectiveness of the Oscillation in Super-Synchronous Frequency Band

Set the short circuit ratio of the power grid to be 1.6, increase the bandwidths of the PLL and DC voltage outer control loop, and increase the active power of WTG progressively to around 550 kW. After super-synchronous oscillation of around 60 Hz occurs, switch on the active damping control. In this case, the wave recordings of active power, angular frequency of the PLL, the output of damping link dq, and the FFT analysis of PLL angular frequency are shown in FIG. 12-15. The variation trajectory of total feedback energy with time before and after the switching of the active damping control is shown in FIG. 16.

It can be seen from FIGS. 12 and 13 that, the active damping control link is switched when the peak values of super-synchronous oscillation components in the waveforms of active power and PLL frequency reach 60 kW and 0.5 Hz respectively. After the active damping control is switched on, the oscillation in super-synchronous frequency band attenuates quickly, and the oscillation amplitude drops below 10% of the original value within 0.1 is. After the waveform being stabilized, there is no longer any super-synchronous oscillation component. FIG. 16 shows that, the total feedback energy before the switching of active damping control exhibits positive feedback characteristics (i.e., increases progressively), thus the oscillation diverges. After the active damping control is switched on, the total feedback energy exhibits negative feedback characteristics (i.e., decreases progressively), thus the oscillation is suppressed and converges. The experimental results above are consistent with the theoretical analysis results based on energy, and verify that the active damping control method proposed in the application has a good suppression effectiveness on the oscillation in the super-synchronous frequency band.

Compared with prior arts, the proposed oscillation active damping control method and system for grid-tied type-4 wind turbine generator are proposed in an embodiment of the application. On one hand, based on the stored energy and dissipated energy of the current inner loop control subsystem, and the interaction energies between the current inner loop control subsystem and other subsystems, an energy feedback model is constructed. Adjusting the current reference value of the q-axis current inner loop by measuring angular frequency of the PLL in real time, makes the stored energy decrease progressively with time, so as to suppress the oscillation. By adjusting the variation rate of the stored energy being made negative, this method ensures the oscillation amplitude of the wind turbine generator show an attenuation trend, no matter how the oscillation frequency changes in the sub-synchronous frequency band and super-synchronous frequency band, so as to effectively suppress oscillation. On the other hand, the oscillation active damping control method and system for grid-tied type-4 wind turbine generator are proposed in an embodiment of the application, when the oscillation frequency of the wind turbine generator changes, the angular frequency measured by the phase lock loop changes correspondingly, by monitoring the angular frequency of the phase lock loop in real time, the instantaneous deviation between the instantaneous angular frequency measured by the phase lock loop and the rated angular frequency is obtained. And the current reference value of the q-axis current inner loop is adjusted by this deviation, so the oscillation of the wind turbine generator can be suppressed automatically, this method is not only easy to implement but also is effective.

It can be understood by those skilled in the art that the whole or part of the process of realizing the above embodiment method can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. The computer-readable storage medium is a disk, an optical disk, a read-only memory or a random storage memory, etc.

The above are only preferred specific embodiments of the present application, but the scope of protection of the present application is not limited to this, any person skilled in the art can easily think of changes or replacement changes within the technical scope disclosed by the present application should be covered within the protection scope of the present application.

Claims

1: An oscillation active damping control method for grid-tied type-4 wind turbine generator, the grid-tied type-4 wind turbine generator comprises multiple subsystems, wherein the method comprising: based on an interconnection model of the multiple subsystems, constructing a stored energy function and a dissipated energy function of a current inner loop control subsystem, and a first interaction energy function between the current inner loop control subsystem and a DC voltage control subsystem, a second interaction energy function between the current inner loop control subsystem and a phase lock loop subsystem, and a third interaction energy function between the current inner loop control subsystem and a reactive power outer loop control subsystem, then establishing an energy feedback model of the grid-tied type-4 wind turbine generator;when oscillation of the grid-tied type-4 wind turbine generator occurs, obtaining instantaneous angular frequency of a phase lock loop in the phase lock loop subsystem, based on the energy feedback model, adjusting a current reference value of a q-axis current inner loop generated by the reactive power outer loop control subsystem, to make the stored energy function value decrease with time, so as to suppress the oscillation.

2: The oscillation active damping control method for grid-tied type-4 wind turbine generator according to the claim 1, wherein constructing energy functions based on an interconnection model of the multiple subsystems, a small signal model of d-axis of the current inner loop control subsystem and a small signal model of q-axis of the current inner loop control subsystem; the expression of the stored energy function is:

3: The oscillation active damping control method for grid-tied type-4 wind turbine generator according to claim 2, wherein the small signal model of the d-axis of the current inner loop control subsystem is:

4: The oscillation active damping control method for grid-tied type-4 wind turbine generator according to claim 2, wherein the energy feedback model is: Vs_acc=Vd_acc+Vt_dc+Vt_pll+Vt_q.

5: The oscillation active damping control method for grid-tied type-4 wind turbine generator according to claim 1, wherein adjusting the current reference value of the q-axis current inner loop generated by the reactive power outer loop control subsystem according to the following compensation function: dIqref=−Idref×∫ωplldt, where dIqref is adjusted current reference value of the q-axis current inner loop, Idref is instantaneous current reference value of the d-axis current inner loop generated by the DC voltage control subsystem, ωpll is deviation of the angular frequency measured by the phase lock loop, that is, the deviation between the instantaneous angular frequency measured by phase lock loop and the rated angular frequency.

6: The oscillation active damping control method for grid-tied type-4 wind turbine generator according to claim 5, wherein the deviation of the angular frequency measured by the phase lock loop and perturbated phase of the phase lock loop in the phase lock loop subsystem satisfy the following relationship: Δθpll=∫ωplldt, where Δθpll is perturbated phase of the phase lock loop in the phase lock loop subsystem, ωpll is the deviation of the angular frequency measured by the phase lock loop;based on the above relationship and the compensation function, the reduction of the third interaction energy function value generated by adjusted current reference value of the q-axis current inner loop which is generated by the reactive power outer loop control subsystem is offset by the increment of the second interaction energy function value that is induced by the perturbated phase of the phase lock loop when the oscillation occurs, to make the stored energy function value decreases with time, so as to suppress the oscillation.

7: The oscillation active damping control method for grid-tied type-4 wind turbine generator according to claim 5, wherein the oscillation comprises oscillation in sub-synchronous frequency band, from 2.5 Hz to 50 Hz, or oscillation in super-synchronous frequency band, from 50 Hz to 100 Hz.

8: An oscillation active damping control system for grid-tied type-4 wind turbine generator, comprising: a model building module is configured to base on an interconnection model of multiple subsystems for the grid-tied type-4 wind turbine generator, construct a stored energy function and a dissipated energy function of the current inner loop control subsystem in the multiple subsystems, and a first interaction energy function between the current inner loop control subsystem and a DC voltage control subsystem, a second interaction energy function between the current inner loop control subsystem and a phase-locked loop subsystem, and a third interaction energy function between the current inner loop control subsystem and a reactive power outer loop control subsystem, then establish an energy feedback model;an adjustment module is configured to obtain the instantaneous angular frequency of a phase lock loop in the phase-locked loop subsystem when the oscillation of the grid-tied type-4 wind turbine generator occurs, based on the energy feedback model, adjust a current reference value of a q-axis current inner loop generated by the reactive power outer loop control subsystem, to make the value of the stored energy function decrease with time, so as to suppress the oscillation.

9: The oscillation active damping control system for grid-tied type-4 wind turbine generator according to claim 8, wherein the adjustment module is configured to adjust a current reference value of a q-axis current inner loop generated by the reactive power outer loop control subsystem based on the following compensation function: dIqref=−Idref×∫ωplldt, where dIqref is adjusted current reference value of the q-axis current inner loop, Idref is instantaneous current reference value of a d-axis current inner loop generated by the DC voltage control subsystem, ωpll is deviation of angular frequency measured by the phase lock loop, that is, the deviation between the instantaneous angular frequency measured by the phase lock loop and rated angular frequency.

10: The oscillation active damping control system for grid-tied type-4 wind turbine generator according to claim 9, wherein the oscillation active damping control system comprises a monitoring module is configured to measure the angular frequency of the phase lock loop in real time.

11: The oscillation active damping control system for grid-tied type-4 wind turbine generator according to claim 8, the model building module is configured to construct energy functions based on an interconnection model of the multiple subsystems, a small signal model of d-axis of the current inner loop control subsystem and a small signal model of q-axis of the current inner loop control subsystem, wherein expression of the stored energy function is:

12: The oscillation active damping control system for grid-tied type-4 wind turbine generator according to claim 11, wherein the small signal model of the d-axis of the current inner loop control subsystem is configured to be:

13: The oscillation active damping control system for grid-tied type-4 wind turbine generator according to claim 11, wherein the energy feedback model is configured to be: Vs_acc=Vd_acc+Vt_dc+Vt_pll+Vt_q.

14: The oscillation active damping control system for grid-tied type-4 wind turbine generator according to claim 9, wherein the energy feedback model is configured that the deviation of the angular frequency measured by the phase lock loop and perturbated phase of the phase lock loop in the phase lock loop subsystem satisfy the following relationship: Δθpll=∫ωplldt, where Δθpll is perturbated phase of the phase lock loop in the phase lock loop subsystem, ωpll is the deviation of the angular frequency measured by the phase lock loop;based on the above relationship and the compensation function, the reduction of the third interaction energy function value generated by adjusted current reference value of the q-axis current inner loop which is generated by the reactive power outer loop control subsystem is offset by the increment of the second interaction energy function value that is induced by the perturbated phase of the phase lock loop when the oscillation occurs, to make the stored energy function value decreases with time, so as to suppress the oscillation.

15: The oscillation active damping control method for grid-tied type-4 wind turbine generator according to claim 14, wherein the oscillation comprises oscillation in sub-synchronous frequency band, from 2.5 Hz to 50 Hz, or oscillation in super-synchronous frequency band, from 50 Hz to 100 Hz.