Virtual synchronous generator damping enhancement control method of nested state feedback controller

Abstract

The invention relates to a virtual synchronous generator damping enhancement control method of a nested state feedback controller, which is characterized by comprising the following steps of: converting a grid side angular frequency reference value of a virtual synchronous generator VSG and adding a correction term to obtain a reconstructed VSG swing equation and a reconstructed VSG small signal model frequency domain form; the correction term comprises an output power feedback compensation term and an angular frequency feedback compensation term; designing double-layer state feedback control by using nested state feedback control, and selecting a state variable of a nested state feedback double-layer control ring; according to the nested state feedback inner and outer ring state variables and the frequency domain form of the reconstructed VSG small signal model, listing a state equation and obtaining a correction term expression, and further obtaining a VSG active ring control model based on a nested state feedback controller; calculating a transfer function of the VSG active loop control model based on the nested state feedback observer; and performing parameter design on the feedback coefficient according to the transfer function by using zero pole distribution and amplitude-phase characteristics.

Description

Technical Field

[0001] The present invention belongs to the technical field of distributed energy grid-type inverter control, and in particular relates to an improved virtual synchronous generator damping enhancement control strategy based on a nested state feedback controller. Background Art

[0002] With the large-scale integration of distributed renewable energy into the power grid, inverter control strategies based on virtual synchronous generators (VSGs) offer significant advantages in improving system inertia, frequency change rate, and maintaining system frequency stability. However, due to the simulation of the synchronous generator rotor motion equations, VSG control inevitably introduces oscillation characteristics similar to those of synchronous generators. When subjected to changes in power reference or grid-side frequency, VSGs are prone to active power oscillations. During oscillations, due to the low overcurrent capability of inverter resources, large transient currents can cause protective devices to operate or even damage them. Furthermore, in AC / DC microgrids, source-load power balance is a fundamental requirement for system stability. Active power oscillations in VSGs can easily cause source-load power imbalances, undermining microgrid stability.

[0003] The active power output oscillations of virtual synchronous generators are caused by a small system damping coefficient. However, conventional virtual synchronous generator damping coefficients have a small design margin. A larger damping coefficient can effectively suppress power oscillations, but it increases the steady-state error of active power. A smaller damping coefficient can reduce the steady-state error, but it cannot effectively suppress power oscillations. Therefore, conventional damping methods are inconsistent in regulating the steady-state and dynamic characteristics of virtual synchronous generators. Summary of the Invention

[0004] To address the aforementioned issues and the conflicting problem of VSG-based renewable energy power systems failing to balance the system's dynamic and steady-state characteristics when adjusting the inertia coefficient J and damping coefficient D during active power oscillations, this paper proposes a transient damping enhancement control strategy that leverages nested state feedback theory to reshape the VSG active loop structure, effectively suppressing inverter output power oscillations under power command disturbances and grid-side frequency variations. The technical solution is as follows:

[0005] This will be supplemented later

[0006] The beneficial effects of the present invention are as follows:

[0007] 1. Improve the dynamic performance of active power response: The improved control method accelerates the active power response speed, shortens the active power output stabilization time, and enhances the dynamic adaptability of the system without affecting the original steady-state characteristics.

[0008] 2. Improve robustness under grid disturbances: In the face of external disturbances such as grid frequency fluctuations and sudden load changes, the invented control strategy can maintain a more stable active power output, effectively suppress active power oscillations caused by external disturbances, and improve the robustness of the system's stable operation.

[0009] 3. Through improved control methods, grid-connected system output is smoother, reducing the impact and fluctuations on the grid caused by inverter output power oscillations during grid connection, helping to improve overall power quality and enhance microgrid stability. The control strategy is applicable to different VSG structures and grid connection conditions, and has a certain degree of versatility to adapt to the needs of various future renewable energy integration scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 Grid-connected conventional VSG network control structure and VSG active loop small signal model, (a) is the grid-connected conventional VSG network control structure, (b) is the conventional VSG active loop small signal model.

[0011] Figure 2 Nested state feedback control two-layer structure

[0012] Figure 3 Block diagram of VSG active loop control based on nested state feedback controller

[0013] Figure 4 Zero-pole distribution of VSG active power closed-loop small signal model based on nested state feedback controller

[0014] Figure 5 Zero-pole distribution of the closed-loop small signal model of the nested state feedback controller VSG active power when h1 decreases from -0.1 to -1.1

[0015] Figure 6 Root locus of the closed-loop small signal model of the nested state feedback controller VSG active power when h1 = -1.1

[0016] Figure 7 Changes in zero-pole distribution of the small signal model of the nested state feedback controller VSG active loop when h3 changes from -0.4 to -0.5

[0017] Figure 8 Changes in zero-pole distribution of the small signal model of the nested state feedback controller VSG active loop when h4 changes from 2π to 40π

[0018] Figure 9 Bode diagram of the VSG system with nested state feedback controller when h2 changes from -8.1 to -0.1

[0019] Figure 10Bode diagram of the VSG system with nested state feedback controller when h2 changes from -0.9 to -0.1

[0020] Figure 11 Bode diagram of the VSG system with nested state feedback controller when h2 changes from -1.1 to -0.1

[0021] Figure 12 Bode diagram of the VSG system with nested state feedback controller when h4 changes from 2π to 38π

[0022] Figure 13 Bode diagram of the VSG system with nested state feedback controller when h3 changes from -0.5 to -4.5

[0023] Figure 14 Bode diagram of the VSG system with nested state feedback controller when h1 changes from -0.01 to -0.81

[0024] Figure 15 In the power command P ref Four ways of VSG output active power diagram under step

[0025] Figure 16 At the grid side frequency ω g Four VSG output active power diagrams under sudden changes

[0026] Figure 17 Example implementation process conventional VSG output current

[0027] Figure 18 Example implementation process coordinated adaptive VSG output current

[0028] Figure 19 Example implementation process frequency feedback VSG output current

[0029] Figure 20 The example implementation process is based on the nested state feedback controller VSG output current DETAILED DESCRIPTION

[0030] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0031] Conventional grid-connected VSG topology and control structure such as Figure 1 (a), where U dc is the DC side voltage, R f , L f , C f The resistance, inductance and capacitance of the RLC filter, L g is the grid-side equivalent inductance, E∠δ is the VSG output terminal voltage, U l∠0° is the voltage at the PCC and is used as the reference phase. δ is the VSG "power angle" which can be regarded as the phase difference between the VSG output terminal voltage and the PCC voltage. abc is the output three-phase current of VSG, u abc is the three-phase voltage at PCC. Figure 1 (a), considering the P-ω droop characteristic, the conventional VSG active control and reactive control are shown in formula (1):

[0032]

[0033] Where J is the virtual moment of inertia, ω is the VSG output angular frequency, ω n is the rated angular frequency, D is the virtual damping coefficient, k p is the newly defined droop coefficient, P ref is the reference active power, P e is the VSG output power. ref is the reference voltage, E m is the electromotive force inside the VSG, k q is the reactive power proportional coefficient, k i is the reactive integral coefficient. It should be noted that the droop coefficient k in the formula p Compared with the conventional definition of droop coefficient k ω (Formula (2)) differs by ω times, that is, k p =k ω / ω.

[0034] P ref -P m =k ω (ω-ω n ) (2)

[0035] Since the VSG output angular frequency ω is different from the rated angular frequency ω n The difference is very small. To simplify the analysis, formula (1) can be expressed as formula (3).

[0036]

[0037] According to the typical structure diagram and circuit theory, since L f It is an LC type filter inductor, so wL f Much larger than R f The value of VSG power angle is small under normal operation. Therefore, the inverter outputs active power and the inverter terminal voltage E∠δ under the VSG control strategy, and the voltage at the common coupling point PCC is U l The relationship between ∠0° can be simplified to the following formula (4):

[0038]

[0039] By linearizing equation (4), we can obtain:

[0040]

[0041] Where U l0 ,E0,δ0 are respectively U l ,E,δ steady-state value. When δ is small enough, the coefficient K cf Much larger than the other two (U l0 / X f )sinδ0,(E0 / X f )sinδ0, so we can ignore it and only keep K cf is a constant. In addition, the power angle δ is related to the grid side angular frequency ω g The small signal model of the relationship is shown below:

[0042]

[0043] According to equations (3) to (6), the small signal model of the conventional VSG active loop is as follows: Figure 1 (b) shown.

[0044] The present invention aims to solve the contradiction between the inertia coefficient J and the damping coefficient D of the new energy power system using VSG and the inability to take into account both the dynamic characteristics and the steady-state characteristics of the system under the active power oscillation of the system, and reconstructs the active loop of the conventional VSG. When the grid-connected VSG is running stably without disturbance, the grid-side angular frequency ω g Equal to the rated angular frequency ω n According to the conventional VSG small signal model, the VSG output power disturbance can be regarded as the superposition of the reference power disturbance and the grid-side frequency disturbance. The conventional VSG swing equation is reconstructed as follows:

[0045]

[0046] Where D p =D+k p The correction terms A and B are the output power feedback compensation term and the angular frequency feedback compensation term, respectively. By performing an equivalent transformation on Equation (7), we can obtain the expressions for the output angular frequency and the output angular frequency derivative, and set ω0 as the initial angular frequency of the system.

[0047]

[0048] According to formula (8), it can be seen that the reconstructed VSG output power can be regarded as the superposition of the system angular frequency derivative after the proportional link, the integral link, the constant term and the correction term. Convert formula (8) into the frequency domain form of the reconstructed VSG small signal model:

[0049]

[0050] Compared with the conventional VSG small signal model, the reconstructed VSG output power disturbance ΔP e (s) can be regarded as ΔP ref (s), d(Δω-Δω g ) / dt(s), ΔA(s) and dΔB / dt(s) are related frequency domain functions. Using nested state feedback controller, the dynamic information of the system is effectively integrated to improve the damping performance of VSG. The reconstructed VSG state variables are selected and combined to design a two-layer state feedback control. The nested state feedback controller is used to adjust two control objectives: P A The swing equation of the VSG active loop is represented by P B The nested state feedback controller structure consists of two control loops: the external controller C A (Including pre-stage controller C1 A and post-stage controller C2 A ) and internal controller C B The external state feedback controller is used to adjust the dynamic deviation, and the internal state feedback controller is used to enhance the frequency response capability. The outer loop state feedback controller is based on ΔA, dΔP ref / dt, and d(Δω-Δω g +ΔB) / dt is the state variable, where d(Δω-Δω g +ΔB) / dt is used as the composite state variable connecting the inner and outer loops. The inner loop state feedback controller uses ΔB and dΔP e / dt is a state variable, such as Figure 2 As shown in the figure, y is the output variable, C1 and C2 are their combination coefficients, and x1 and x2 are the state variable sets of the outer loop state feedback controller and the inner loop state feedback controller respectively.

[0051] In the outer loop state feedback controller, in order to analyze the dynamic coupling between the reference disturbance and the reconstructed VSG internal dynamics, dΔP is added to the state variable ref / dt. dΔP ref / dt responds to the change of reference power. In order to introduce frequency compensation ΔB through inner loop feedback, the composite state variable d(Δω-Δωg+ΔB) / dt is defined. ΔA feeds back the grid-side frequency disturbance and the reference power deviation caused by the load from the inner loop to the active loop. Therefore, according to the frequency domain form (9) of the reconstructed VSG small signal model, the state equation of the external state feedback controller is:

[0052]

[0053] Where Δμ = dA / dt, and vector x1 is the set of system state variables. Comparing the state-space equation in Equation (10) with the standard state-space equation, the control signal Δμ is considered as the control input of the external state controller system. According to the control form of the standard state feedback controller, the control input Δμ can be designed as:

[0054] Δμ=[h1,h2,-h4]x1 (11)

[0055] Where h1, h2, and h4 are feedback gain coefficients. According to the designed control input Δμ expression (11), the time domain expression of the relationship between ΔA and ΔB can be obtained:

[0056]

[0057] Performing Laplace transform and equivalent transform on Equation (12) yields the frequency domain expression of the relationship between ΔA and ΔB:

[0058]

[0059] According to the relationship between the output power and angular frequency of the reconstructed VSG small signal model, the following formula can be obtained:

[0060]

[0061] In the inner loop state feedback controller, according to the above analysis, ΔP ref and Δω g P e Interference occurs due to ΔP e is close to zero in steady state, so dΔP is chosen e / dt is used as the state variable. And the variable dΔP e / dt can improve the system's ability to respond to output power deviation. The variable ΔB converts the power deviation into a frequency compensation signal and transmits it to the outer loop state feedback controller through the composite state variable. Therefore, dΔP is selected e / dt and ΔB are used as the state variables of the inner loop state feedback controller. According to the frequency domain forms (9) and (14) of the reconstructed VSG small signal model, the state equation of the inner loop state feedback controller can be written as follows:

[0062]

[0063] Where Δν = dΔB / dt. Equation (15) is not a standard state space equation. Its matrix C contains some state variables of the external state feedback. To simplify the analysis, the internal state feedback observer is assumed to have a high bandwidth. That is, under the time scale of the internal state feedback observer, the external state feedback variables change slowly and can be regarded as invariants. In this case, Δν is the system control input, and the vector x2 is the set of internal observer system state variables. According to the state feedback theory, the control form of the controller can be obtained as follows:

[0064] Δν=[h3,-h4]x2 (16)

[0065] Where h3 and h4 are the feedback gain coefficients. According to equation (16), the time domain expression of ΔB can be obtained as follows:

[0066]

[0067] Performing Laplace transform on (17) yields the frequency domain expression of ΔB:

[0068]

[0069] Substituting the frequency domain expression (18) of ΔB into the frequency domain expression (13) of the relationship between ΔA and ΔB, we can obtain:

[0070]

[0071] According to formula (19), the correction term is added to the conventional VSG active loop small signal model to obtain the VSG active loop control model based on the nested state feedback controller as shown in the attached figure. Figure 3 As shown, that is, the first high-frequency component ΔP of the active power instruction is obtained through the first high-pass filter h1s / (s+h4) ref (s)h1s / (s+h4), the first high-frequency component is fed back to the reference power input node; the second high-pass filter h3s / (s+h4) is used to obtain the second high-frequency component ΔP of the VSG output active power e (s)h3s / (s+h4) is fed back to the output angular frequency node, and the second high frequency component is combined with the angular frequency change (Δω-Δω g ) is superimposed and passed through the third high-pass filter h2s / (s+h4) to obtain the corrected angular frequency (Δω-Δω g +h3s / (s+h4)ΔP e ), and feeds back the third high-frequency component to the feedback node of the VSG output power, where the feedback coefficients h1, h2, and h3 are the gains of the first, second, and third high-pass filters in the VSG active loop control model based on the nested state feedback controller, and h4 is the cutoff frequency of these three high-pass filters.

[0072] The process of the technical solution of the present invention is summarized and explained below.

[0073] Step 1: Establish a conventional VSG active loop small signal model.

[0074] Step 2: According to the nested state feedback control theory, the VSG active loop structure based on the nested state feedback controller is obtained.

[0075] Step 3: Parameter design. According to the VSG active loop control model based on the nested state feedback controller, its transfer function is solved. The feedback coefficients h1, h2, h3, and h4 are designed using the root locus, zero-pole configuration, and amplitude-phase characteristics to ensure that they achieve the ideal effect on active power oscillation suppression.

[0076] The specific method for parameter design in step 3 is as follows:

[0077] According to the VSG active loop control small signal model of the nested state feedback controller designed in step 2 ( Figure 3 As shown), the transfer function is calculated as follows:

[0078]

[0079] Where:

[0080]

[0081] First, determine the stability of the method proposed by the present invention: Draw the root locus according to the Figure 4 As shown, the poles are all distributed in the left half plane, and the root locus curve does not cross the right half plane. Therefore, the system remains stable after using the method proposed in the present invention.

[0082] Next, we design the parameters for the gain coefficients h1, h2, h3, and h4 based on the zero-pole distribution. Since the VSG active loop system contains conjugate poles and negative real poles after the correction step, the conjugate poles determine system oscillations. The absolute value of the real part of the conjugate pole affects the oscillation decay rate, while the absolute value of the imaginary part affects the oscillation frequency. The real pole determines the system response speed. The larger the real part of the pole, the faster the system response speed. Based on this theory, the gain coefficients h1, h2, h3, and h4 are adjusted. The goal is to adjust the parameters around the initial values ​​to achieve a balance between the system oscillation decay rate, oscillation frequency, and response speed.

[0083] When the gain coefficient h1 changes, the system root locus, that is, the zero-pole transformation, is as follows Figure 5 As shown in Figure 2, h1 mainly affects the system zero point. When h1 decreases from -0.1 to -1.1, the system zero point moves to the right half plane, and after h1 is less than -1, the zero point crosses the imaginary axis, which can easily cause system instability. The root locus of h1 = -1.01 is shown in Figure 2. Figure 6 As shown, the value of h1 should be greater than -1.

[0084] The zero-pole change trend when the gain coefficient h3 changes from -0.4 to -0.5 is as follows Figure 7 As shown, its main impact is that a pair of conjugate poles move toward the real axis and a negative real pole moves away from the imaginary axis. In this change process, the absolute value of the imaginary part of the conjugate pole gradually decreases and the system oscillation decreases. At the same time, the absolute value of the real part of the negative real pole increases and the system reaction speed increases. When the h3 change value is -0.5, its pair of conjugate poles move to the imaginary part of the real axis is 0. At this point, all poles with h3<-0.5 can be regarded as having an imaginary part of 0, which can effectively suppress the active power oscillation of the system.

[0085] The gain coefficient h4 is used as the cutoff frequency of the high-pass filter, which introduces the high-frequency components of the three state quantities. For the high-pass filter, the 20Hz standard is usually considered to filter out the low-frequency components below 20Hz. The root locus changes when h4 is gradually increased from 2π to 40π. Figure 8 As shown, it can be found that in the process of increasing h4, the conjugate pole of the system gradually moves away from the real axis, and the negative real pole moves away from the imaginary axis, that is, as h4 increases, the rapid response capability of the system is enhanced while the system oscillation is intensified. Therefore, in order to ensure that the system oscillation attenuation speed, oscillation frequency, and response speed reach equilibrium, the value of h4 should not be too large or too small, and should be kept at an intermediate value.

[0086] The gain coefficient h2 has little effect on the system zeros and poles. Therefore, in order to further adjust the gain coefficient, the Bode diagram is drawn under different parameter changes according to formulas (20) and (21) and analyzed and designed. The specific method is shown below.

[0087] The influence of gain coefficient h2 on the frequency domain characteristics of the system is as follows: Figure 9 , 10, 11. Specifically, when h2 gradually increases from –8.1 to –2.1, the low-frequency gain K of the system low It shows a clear downward trend, indicating that the system's ability to suppress low-frequency oscillations has been enhanced. At the same time, the system's bandwidth Φ BW The system gain gradually decreases, and the response speed decreases, indicating that while the system improves steady-state performance, its dynamic response capability has weakened. In addition, within the frequency range of 1Hz to 1.26Hz, the overall system gain is greater than 0dB, and as h2 increases, the gain decreases, further suppressing the low-frequency oscillations that may occur in this frequency range.

[0088] In addition, according to Figure 9 The phase-frequency characteristic curve shows that when h2 gradually increases from -8.1 to -0.1, the system phase curve suddenly changes. Figure 11It is found that when h2<-1, the system phase curve undergoes a sudden change and the system phase increases significantly. Therefore, the analysis of the h2 phase-frequency characteristic curve should also be conducted according to h2>-1 and h2<-1. Figure 10 , when h2 gradually increases from -0.9 to -0.1, the system phase margin first decreases and then increases, and the system phase is always near -180°, and the phase margin is small. Figure 11 , in the process of gradually increasing from -8.1 to -2.1, the system phase margin gradually increases and the phase margin is greater than 50°. Therefore, in order to ensure that the phase margin is always greater than 30°, the value of h2 should be less than -1. Taking into account the impact of h2 on the low-frequency gain and bandwidth of the system, although a smaller h2 value can improve the response speed of the system, when the h2 value is too small, the low-frequency oscillation phenomenon of the system will be significantly enhanced, thereby affecting the stability and robustness of the system. Therefore, the selection of h2 should not be too small. In the present invention, based on the comprehensive trade-off between dynamic performance and low-frequency oscillation suppression, h2 is set to -2.1.

[0089] The influence of gain coefficient h4 on the frequency domain characteristics of the system is as follows: Figure 12 As shown in Figure 2. In conventional VSG, a resonance peak greater than 0 dB appears in the low frequency band, resulting in power oscillation. Figure 12 Under nested state feedback VSG control, the resonance peak in the low-frequency band is smaller than that of a conventional VSG. As h4 increases, the system bandwidth increases, indicating faster dynamic response. However, larger resonance peaks are also observed, increasing the risk of oscillation. Furthermore, as h4 increases, the peak on the phase-frequency curve around 100 Hz gradually turns into a downward peak, and the system phase margin gradually decreases. Therefore, considering the system phase margin, dynamic response speed, and oscillation frequency, the value of h4 in the design of this invention is 14π.

[0090] The influence of gain coefficient h3 on the frequency domain characteristics of the system is as follows: Figure 13 As shown. Combined with the above analysis of the zero-pole distribution under the change of gain coefficient h3, when h3<-0.5, all poles can be regarded as having an imaginary part of 0, which can effectively suppress the active power oscillation of the system. Therefore, consider the change of the system Bode diagram when h3 gradually decreases from -0.5 to -4.5. When h3 continues to decrease, the low-frequency gain K low decreases, indicating that low-frequency oscillation suppression is improved. However, the system bandwidth Φ BW Slightly decreased, the system's rapid response capability decreased. In the process of h3 decreasing, the phase margin θ corresponding to the system's 0dB gain crossover frequency PM Increasing is beneficial to the stability of the system. Comprehensively considering the system bandwidth Φ BW and the phase margin θ PM In the design of the present invention, the value of h3 is -1.5.

[0091] The influence of gain coefficient h1 on the frequency domain characteristics of the system is as follows: Figure 14 As shown. Combined with the above analysis of the zero-pole distribution under the change of gain coefficient h1, in order to maintain system stability h1 should be greater than -1. Considering h1 decreases from -0.01 to -0.81, the low-frequency gain K low It remains almost unchanged, indicating that h1 has limited influence on the steady-state gain of the system. However, in the mid-frequency range, the gain is significantly attenuated, especially near the resonant frequency. This means that the amplification of external interference is reduced and the robustness of the system is improved. In addition, the system bandwidth gradually decreases, indicating that the dynamic response slows down as the absolute value of h1 increases. The phase diagram shows that as h1 becomes smaller, the phase margin decreases significantly. In particular, in the frequency range of 5-20Hz, the phase drops faster, resulting in a significant reduction in the phase margin. When h1 = -0.81, the phase at the gain crossover frequency is close to -180°, indicating that the system operates near the stability limit, which will reduce the stability of the system. Therefore, increasing the absolute value of h1 will lead to reduced dynamic responsiveness and a smaller phase margin, which will affect the stability of the system. In the design of the present invention, the value of h1 is -0.01.

[0092] Examples:

[0093] During the operation of the grid-connected inverter, at t=1s, the VSG active power command value P ref From 150kW to 170kW, the grid frequency ω at t = 2s g The grid frequency ω drops suddenly from 50Hz to 49.6Hz at t=3s g Restore 50Hz. In this example, the grid line voltage is 380V.

[0094] VSG output power curve is as follows Figure 15-16 shown. Figure 17-20 The corresponding output current curve is shown.

[0095] Figure 15 The active power output diagram of the nested state feedback controller VSG under the active power instruction step provided by the embodiment of the present invention and the VSG of the comparative method, when t=1s is set, P ref When the power supply is stepped up to 170kW, compared to three other typical approaches (conventional VSG, coordinated adaptive VSG, and frequency feedback VSG), the power response waveform under the nested state feedback controller (VSG) exhibits no significant oscillation or overshoot, achieving ideal control results, with no sudden slope increase during the entire response process. This fully demonstrates the effectiveness of the proposed control strategy in suppressing active power oscillations in the presence of sudden power command changes.

[0096] Figure 16The VSG output active power diagram of the nested state feedback controller VSG under the grid side frequency mutation provided by the embodiment of the present invention and the comparative method, setting t=2s grid side frequency ω g The grid frequency ω drops suddenly from 50Hz to 49.6Hz at t=3s g Restore to 50Hz. Compared with the other three typical methods, the nested state feedback controller VSG effectively suppresses the power oscillation amplitude during step changes in grid frequency, has the shortest stabilization time and the ability to quickly recover to the steady-state value, and its power change rate is also smaller than the other three typical methods. In contrast, the nested state feedback controller VSG has a smooth power response throughout the entire process, with almost no oscillation and overshoot. This fully verifies the effectiveness of the control strategy proposed in this invention in suppressing active power oscillations in the case of sudden changes in grid-side frequency.

[0097] Figure 17-20 Figures showing the inverter output current under control of a conventional VSG, a coordinated adaptive VSG, a frequency feedback VSG, and a nested state feedback controller VSG, respectively, according to an embodiment of the present invention. During the entire 0-4s implementation period, the nested state feedback controller VSG achieved the best output current oscillation suppression. This demonstrates the effectiveness of the proposed control strategy in suppressing active power oscillations under sudden power command and grid-side frequency changes.

Claims

1. A virtual synchronous generator damping enhancement control method based on a nested state feedback controller, characterized in that: The following steps are involved: Step 1: Convert the grid-side angular frequency reference value of the virtual synchronous generator (VSG) and add a correction term to obtain a reconstructed VSG swing equation and a reconstructed VSG small signal model in frequency domain. The correction term includes an output power feedback compensation term and an angular frequency feedback compensation term. Step 2: Design a double-layer state feedback control loop using nested state feedback control and select the state variables of the nested state feedback double-layer control loop; Step 3: Write the state equation according to the state variables of the inner and outer loops of the nested state feedback and the frequency domain form of the reconstructed VSG small signal model and obtain the expression of the correction term, thereby obtaining the VSG active loop control model based on the nested state feedback controller; Step 4, calculate the transfer function of the VSG active loop control model based on the nested state feedback observer; Step 5, design the parameters of the feedback coefficient according to the transfer function using the zero-pole distribution and the amplitude-phase characteristics. The method is as follows: using the zero-pole configuration method, determine its stable range according to the influence of the feedback coefficient on the zero-pole distribution of the transfer function, adjust the feedback coefficient so that the conjugate pole gradually moves to the real axis so that it is transformed into a negative real pole; use the amplitude-phase characteristic curve to design the feedback coefficient from four perspectives: system bandwidth, phase margin, low-frequency gain, and resonance peak.

2. The virtual synchronous generator damping enhancement control method of nested state feedback controller according to claim 1 is characterized in that: The method for step 1 is: When the grid-connected VSG is running stably without disturbance, the grid-side angular frequency ω g Equal to the rated angular frequency ω n According to the VSG small signal model, the VSG output power disturbance can be regarded as the superposition of the reference power disturbance and the grid-side frequency disturbance. Reconstruct the VSG swing equation: Where D p =D+k p , J is the virtual moment of inertia, D is the virtual damping coefficient, ω is the VSG output angular frequency, k p is the droop coefficient and ω n The ratio, P ref is the reference active power, P e is the VSG output power, and the correction terms A and B serve as the output power feedback compensation term and the angular frequency feedback compensation term; Convert the reconstructed VSG swing equation (1) into the frequency domain form of the reconstructed VSG small signal model:

3. The virtual synchronous generator damping enhancement control method of nested state feedback controller according to claim 2 is characterized in that: The method for step 2 is: According to the reconstructed VSG small signal model frequency domain form, its output power disturbance ΔP e (s) is determined by ΔP ref (s), d(Δω-Δω g ) / dt(s), ΔA(s) and dΔB / dt(s). By using nested state feedback control, a double-layer state feedback control is designed. The outer loop state feedback controller is based on ΔA, dΔP ref / dt, and d(Δω-Δω g +ΔB) / dt is the state variable, where d(Δω-Δω g +ΔB) / dt is used as the composite state variable connecting the inner and outer loops. The inner loop state feedback controller uses ΔB and dΔP e / dt is the state variable.

4. The virtual synchronous generator damping enhancement control method of nested state feedback controller according to claim 1 is characterized in that: The method for step 3 is According to formula (2), the state equation of the outer loop state feedback controller is written as: In the formula, the intermediate variable Δμ = dA / dt is introduced, and the vector x1 is the state variable set of the outer loop state feedback controller. Comparing formula (3) with the standard state space equation, the control signal Δμ is the control input of the outer loop state feedback controller system. According to the control form of the standard state feedback controller, the control signal Δμ is designed as Δμ = [h1,h2,-h4]x1, where h1,h2,h4 are feedback coefficients. Based on this, the expression of the correction term small disturbance ΔA is obtained: According to the reconstructed VSG small signal model in the frequency domain, the relationship between the output power, output angular frequency and grid-side angular frequency ΔP e =K cf / s(Δω-Δω g +ΔB) and formula (2) to write the state equation of the inner loop state feedback controller: In the formula, the intermediate variable Δν=dΔB / dt; K cf =(E0U l0 / X f )cosδ0,U l0 , E0, δ0 are the common coupling point voltage, VSG terminal voltage and the steady-state value of power angle respectively, X f is the equivalent line reactance; In formula (5), the matrix C contains some state variables of the external state feedback. Under the time scale of the internal state feedback observer, the changes of the external state feedback variables are ignored. At this time, Δν is used as the system control input, and the vector x2 is used as the set of internal observer system state variables. According to the control form of the standard state feedback controller, the control input Δν is designed as: Δν = [h3, -h4]x2, where h3 and h4 are feedback coefficients. Based on this, the expression (6) of the correction term small disturbance ΔB is obtained, and the complete expression (7) of ΔA is further obtained: According to formula (7), the correction term is added to the VSG active loop small signal model to obtain the VSG active loop control model based on the nested state feedback controller, that is: the first high-frequency component ΔP of the active power instruction is obtained through the first high-pass filter h1s / (s+h4) ref (s)h1s / (s+h4), the first high-frequency component is fed back to the reference power input node; the second high-pass filter h3s / (s+h4) is used to obtain the second high-frequency component ΔP of the VSG output active power e (s)h3s / (s+h4) is fed back to the output angular frequency node, and the second high frequency component is combined with the angular frequency change (Δω-Δω g ) is superimposed and passed through the third high-pass filter h2s / (s+h4) to obtain the corrected angular frequency (Δω-Δω g +h3s / (s+h4)ΔP e ), and feeds back the third high-frequency component to the feedback node of the VSG output power, where the feedback coefficients h1, h2, and h3 are the gains of the first, second, and third high-pass filters in the VSG active loop control model based on the nested state feedback controller, and h4 is the cutoff frequency of these three high-pass filters.

5. The virtual synchronous generator damping enhancement control method of nested state feedback controller according to claim 4 is characterized in that: The method for step 4 is: Where:

Images