Fan power decoupling method based on adaptive virtual impedance and feedforward compensation

Abstract

The invention belongs to the technical field of new energy active support stability control, and provides a fan power decoupling method based on adaptive virtual impedance and feedforward compensation. According to the method, dynamic characteristics of active power and reactive power in the system are analyzed by combining a topological structure of a doubly-fed fan grid-connected system and fan internal control, and a small signal model is constructed based on power transmission characteristics. A self-adaptive virtual impedance strategy based on power grid impedance characteristic adjustment is introduced for the problem of power coupling caused by high line impedance in a weak power grid environment and dynamic and control loop characteristics of flux linkages in a fan. Meanwhile, through a disturbance component extraction and feedforward control strategy, a disturbance signal in rotor current is introduced into a power ring for adjustment, and power decoupling control is realized. The method overcomes the limitation that an existing decoupling control technology is poor in effect, the control principle is clear, engineering implementation is easy, and the stability and the dynamic response capability of the doubly-fed fan in the weak power grid environment can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of new energy active support stability control, and particularly to a method for decoupling fan power based on adaptive virtual impedance and feed-forward compensation. Background Technique

[0002] To achieve the "dual carbon" goal, China urgently needs to improve the flexibility and stability of wind power grid connection. The doubly-fed induction generator (DFIG) has become an important choice for wind power development due to its high efficiency and good grid connection characteristics. However, the power coupling problem during the grid connection process of DFIG has become a key challenge restricting system stability. When the doubly-fed fan is connected to the grid, the rotor-side converter realizes active and reactive power regulation through current control. However, due to the dynamic interaction between the grid impedance and the rotor current, there is a strong coupling effect in the power output. This coupling will cause problems such as voltage fluctuations and transient oscillations at the grid connection point, especially in weak grids or fault conditions, seriously affecting the low voltage ride-through ability and power quality of the system.

[0003] Traditional decoupling methods mostly use virtual impedance compensation with fixed parameters. However, such methods are insufficient in adaptability under time-varying grid parameters or complex working conditions, and it is difficult to dynamically eliminate the coupling effect, resulting in a decrease in power control accuracy and even the risk of sub-synchronous oscillation. Although the active compensation strategy based on virtual impedance can partially alleviate the coupling effect, its performance highly depends on the static design of the impedance parameters and cannot adapt to the dynamic changes of the grid impedance and non-linear interference. In addition, it is difficult to completely eliminate the coupling residue only through single virtual impedance compensation. Especially in the multi-objective control scenario where the fan participates in frequency modulation and voltage regulation, the interaction coupling between the power loop and the current loop is further intensified. Summary of the Invention

[0004] The purpose of the present invention is to solve the defects existing in the prior art, and to propose a method for decoupling fan power based on adaptive virtual impedance and feed-forward compensation.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions:

[0006] In the first aspect, a method for decoupling fan power based on adaptive virtual impedance and feed-forward compensation provided by the present invention includes the following steps:

[0007] S1. Combine the equivalent circuit of the voltage-controlled DFIG-VSG system and the grid and the internal control structure of the system to analyze the active power and reactive power output by the system;

[0008] S2. Construct a power small-signal model according to the analysis results of the active power and reactive power of the system;

[0009] S3. For the power coupling caused by different factors, considering the coupling caused by line impedance under weak grid conditions, an adaptive virtual impedance decoupling strategy is adopted;

[0010] S4. After the system adopts the adaptive virtual impedance decoupling strategy and there is high-frequency coupling, a rotor current feedforward decoupling strategy is adopted within the loop for power decoupling.

[0011] In a possible implementation manner, the step S1 specifically includes the following sub-steps:

[0012] S11. According to the voltage source characteristics exhibited by the voltage-controlled DFIG-VSG, the amplitude and phase angle of the voltage source are the virtual electromotive force and virtual impedance under VSG control, and their magnitudes are consistent with the external voltage on the stator side. An equivalent circuit between the doubly-fed wind turbine system and the power grid is constructed. At the same time, using the flux linkage relationship inside the doubly-fed wind turbine, a dynamic equation between the stator voltage and the flux linkage is constructed;

[0013] S12. Using the stator voltage as a bridge, an influence equation of current considering the power grid and the wind turbine is constructed;

[0014] S13. Through the power calculation expression, an equation of active power and reactive power with the rotor-side current, which is the direct control quantity of the wind turbine, is constructed to analyze the active power and reactive power output by the system.

[0015] In a possible implementation manner, the calculation formula of the dynamic equation described in step S11 is:

[0016] The equivalent circuit equation of the doubly-fed system equivalent to the voltage source and the power grid side is expressed as:

[0017]

[0018] In the formula, ω s is the operating frequency of the VSG, R g and L g are the equivalent resistance and equivalent reactance on the line respectively;

[0019] The stator and rotor flux linkage and voltage equations of the doubly-fed wind turbine are respectively:

[0020]

[0021] In the formula, the subscripts d and q are the components of the electrical quantity on the d-axis and q-axis in the two-phase rotating coordinate system respectively, and the subscript s is the electrical quantity of the stator; p is the differential operator, and L m is the mutual inductance between the stator and rotor of the doubly-fed wind turbine.

[0022] In a possible implementation manner, the power calculation formula described in step S13 is:

[0023]

[0024] According to the power calculation formula, it can be obtained that:

[0025]

[0026] In the formula, P s and Q s are the active power and reactive power respectively transmitted by the doubly-fed fan to the power grid under control. U gd and U gq are the dq values of the grid-side voltage respectively. ||I s || 2 is the square of the stator-side current modulus, where I sd and I sq can be expressed by the rotor-side current and grid-side factors respectively.

[0027] In a possible implementation manner, the power small-signal model representation formula described in step S2 is:

[0028]

[0029] In the formula, represents the influence of I rd on Q, represents the influence of I rq on Q, represents the influence of I rd on P, represents the influence of I rq on P.

[0030] In a possible implementation manner, the adaptive virtual impedance decoupling described in step S3 includes:

[0031] For the power coupling caused by the line impedance in the system, a virtual impedance is introduced for compensation. Through the negative resistance characteristic in the virtual impedance, the resistive component on the line is eliminated, making it present a high inductive reactance externally. After the introduction, the power is expressed as follows:

[0032]

[0033] Considering that the line impedance will change with multiple factors, in order to achieve better power decoupling, an adaptive control is introduced. By detecting the target power fluctuation amplitude, with the minimum power fluctuation amplitude as the objective function and the fluctuation as the penalty function factor, specifically as the formula:

[0034] ΔP PQ = P PQ - P PQE

[0035]

[0036] where: ΔP PQ is the generated power fluctuation, P PQ and P PQE respectively represent the actual power value and the rated value; α and β are weight coefficients respectively; dZ v / dt represents the virtual impedance change rate, which is used to avoid frequent jumps of the virtual impedance; ΔP RMS represents the root mean square value of the power fluctuation within the sliding window, which is used to penalize the power fluctuation. After obtaining the penalty factor, the gradient descent algorithm is used to dynamically adjust Z v , η represents the learning rate, which controls the parameter update speed; the gradient is approximately calculated by the numerical perturbation method;

[0037] According to the above formula, the size of the virtual impedance is dynamically adjusted so that the change of the virtual impedance can adapt to the change of the line impedance, and finally a better power decoupling effect is achieved.

[0038] In a possible implementation manner, the feedforward decoupling strategy described in step S4 is as follows:

[0039] After the system adds a virtual impedance module for power decoupling, the DFIG-VSG system weakens the power coupling caused by adjusting the active power or reactive power in the transient state of the system. For the expressions of the active power and reactive power containing the coupling variables I rd and I rq , it shows that after introducing the virtual impedance, due to the control of the doubly-fed wind turbine itself, power coupling will be caused. In order to achieve more precise power control and accurate voltage and frequency support, a feedforward decoupling strategy is further introduced on the basis of the virtual impedance;

[0040] Taking the fluctuation control amount of the target power as the feedforward object, by injecting the compensation amount into the error node of the power control loop, the coupling path is actively cancelled, the active-reactive interaction is reduced, and the decoupling expression after adding the compensation is as follows:

[0041]

[0042] The system goal is to make the irrelevant quantities hardly fluctuate during support, and it is necessary to keep each power perturbation at 0 during the control process. It can be deduced that the compensator is:

[0043] BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention.

[0045] Figure 1A flowchart of a wind turbine power decoupling method based on adaptive virtual impedance and feedforward compensation proposed by the present invention;

[0046] Figure 2 This is a block diagram of an adaptive virtual impedance control method for a wind turbine power decoupling method based on adaptive virtual impedance and feedforward compensation proposed by the present invention;

[0047] Figure 3 This is a feedforward compensation control block diagram of a fan power decoupling method based on adaptive virtual impedance and feedforward compensation proposed by the present invention. DETAILED DESCRIPTION

[0048] In order to better understand the technical solution of the present application, the embodiments of the present application are described in detail below with reference to the accompanying drawings.

[0049] An embodiment of a wind turbine power decoupling method based on adaptive virtual impedance and feedforward compensation is disclosed. This embodiment demonstrates the power regulation process of the system during transient operation. Firstly, the equivalent circuit of the voltage-controlled DFIG-VSG system and the power grid is used to construct an equivalent circuit equivalent equation. The internal equation is constructed according to the internal flux control of the wind turbine. The stator voltage is used as a bridge to simultaneously analyze the active power and reactive power output by the system. Then, according to the analysis results of the active power and reactive power of the system, a power small signal model of the system is constructed for quantitative analysis. Secondly, considering the coupling caused by line impedance under weak power grid conditions, virtual impedance is used for compensation to offset the influence of line impedance. However, considering the time-varying nature of line impedance, an adaptive virtual impedance strategy is designed using power deviation. Finally, in view of the high-frequency coupling component that still exists after adopting the above strategy, a rotor current feedforward decoupling strategy is adopted in the loop to ultimately achieve more accurate power decoupling.

[0050] Figure 1 This is a flowchart of a fan power decoupling method based on adaptive virtual impedance and feedforward compensation proposed by the present invention, such as Figure 1 As shown, the system includes four steps: S1, S2, S3, and S4:

[0051] S1. Combine the equivalent circuit of the voltage-controlled DFIG-VSG system and the power grid and the internal control structure of the system to analyze the active power and reactive power output of the system;

[0052] S2. Construct a power small signal model based on the analysis results of the system active power and reactive power;

[0053] S3. Aiming at the power coupling caused by different factors, the coupling caused by line impedance under weak grid conditions is considered, and an adaptive virtual impedance decoupling strategy is adopted;

[0054] After the system adopts the adaptive virtual impedance decoupling strategy and there is high-frequency coupling, a rotor current feedforward decoupling strategy is adopted within the loop for power decoupling.

[0055] Figure 2 This is the adaptive virtual impedance control block diagram of a fan power decoupling method based on adaptive virtual impedance and feedforward compensation proposed by the present invention. In the figure, P m and P e represent the mechanical active power and electromagnetic active power of the system, ω0 and ω * represent the rated angular velocity and the reference angular velocity of production respectively, Q ref and Q e represent the reactive power rated value and the electromagnetic reactive power of the system respectively, V0, and represent the rated voltage and the generated voltage reference value respectively, D P and D q represent the droop coefficients of active power and reactive power, 1 / s represents an integrator, I s represents the stator output current, Z v represents the virtual impedance introduced by the system. The system voltage compensation value ΔV d and ΔV q is calculated through the stator current;

[0056] As Figure 2 shown, this module introduces a virtual impedance, uses the negative resistance characteristic of the virtual impedance to cancel the resistive component on the line, multiplies it with the stator-side current to obtain the magnitude of the voltage to be compensated, and adds the voltage compensation value to the voltage loop of the converter control loop to achieve virtual impedance compensation. Considering the time-varying nature of the line impedance, an adaptive virtual impedance strategy considering power deviation as a penalty factor is introduced to achieve power decoupling.

[0057] In the embodiment of the invention, the specific steps of the adaptive virtual impedance strategy are as follows:

[0058] In the system, for the power coupling caused by the line impedance, a virtual impedance is introduced for compensation. Through the negative resistance characteristic in the virtual impedance, the resistive component on the line is eliminated, making it present a high inductive reactance externally. The power after introduction is expressed as follows:

[0059]

[0060] Considering that the line impedance will change with multiple factors, in order to achieve better power decoupling, adaptive control is introduced. By detecting the target power fluctuation amplitude, with the minimum power fluctuation amplitude as the objective function, the fluctuation is used as the penalty function factor. Specifically, the formula is:

[0061] ΔP PQ = P PQ - PPQE

[0062]

[0063] Where: ΔP PQ is the generated power fluctuation, P PQ and P PQE respectively represent the actual power value and the rated value; α and β are weight coefficients respectively; dZ v / dt represents the virtual impedance change rate, which is used to avoid frequent jumps of the virtual impedance; ΔP RMS represents the root mean square value of the power fluctuation within the sliding window, which is used to punish the power fluctuation. After obtaining the punishment factor, the gradient descent algorithm is used to dynamically adjust Z v , η represents the learning rate, which controls the parameter update speed; the gradient is approximately calculated by the numerical perturbation method;

[0064] According to the above formula, the size of the virtual impedance is dynamically adjusted so that the change of the virtual impedance can adapt to the change of the line impedance, and finally the power decoupling effect is achieved.

[0065] Figure 3 This is the feed-forward compensation control block diagram of a wind turbine power decoupling method based on adaptive virtual impedance and feed-forward compensation proposed by the present invention. In the figure, ΔQ ref , ΔP ref represent the change amount of the power reference value, ΔQ and ΔP represent the change amount of the power output, G 11 and G 22 respectively represent the controllers for generating the rotor current in the active loop and the reactive loop, G P21 and G P12 respectively represent the feed-forward compensation coefficients;

[0066] As Figure 3 shown, after applying the adaptive virtual impedance strategy, it can be analyzed that there are still high-frequency coupling components in the system. The reason is the dynamic change of the magnetic flux inside the wind turbine, and the virtual impedance can only eliminate the coupling caused by the line impedance. Therefore, on this basis, feed-forward compensation decoupling is carried out, taking the fluctuation control amount of the target power as the feed-forward object, and by injecting the compensation amount into the error node of the power control loop, the coupling path is actively cancelled, and the active-reactive interaction is reduced. The decoupling expression after adding the compensation is as follows:

[0067]

[0068] The system goal is to make the irrelevant quantities hardly fluctuate during support, and it is necessary to keep each power perturbation at 0 during the control process. It can be deduced that the compensator is:

[0069]

[0070] Through the above strategies, precise power decoupling of the doubly-fed wind turbine system can be achieved, which is beneficial to the stable operation of the system.

Claims

1. A wind turbine power decoupling method based on adaptive virtual impedance and feedforward compensation, characterized in that: The method comprises the following steps: S1. Combine the equivalent circuit of the voltage-controlled DFIG-VSG system and the power grid and the internal control structure of the system to analyze the active power and reactive power output of the system; S2. Construct a power small signal model based on the analysis results of the system active power and reactive power; S3. Aiming at the power coupling caused by different factors, the coupling caused by line impedance under weak grid conditions is considered, and an adaptive virtual impedance decoupling strategy is adopted; S4. After the system adopts the adaptive virtual impedance decoupling strategy, if there is high-frequency coupling, the rotor current feedforward decoupling strategy is adopted in the loop for power decoupling.

2. The method according to claim 1, characterized in that The step S1 specifically includes the following sub-steps: S11. According to the voltage source characteristics displayed by the voltage-controlled DFIG-VSG, the amplitude and phase angle of the voltage source displayed are the virtual potential and virtual impedance under the control of VSG, and its size is consistent with the external voltage performance of the stator side. The equivalent circuit between the doubly fed wind turbine system and the power grid is constructed. At the same time, the magnetic flux relationship inside the doubly fed wind turbine is used to construct the dynamic equation between the stator voltage and magnetic flux; S12. Using the stator voltage as a bridge, construct an equation that considers the current impact of the grid and the wind turbine; S13. Through the power calculation expression, construct the equation of active power, reactive power and the rotor side current of the wind turbine direct control quantity, and analyze the active power and reactive power output by the system.

3. The method according to claim 2, characterized in that The dynamic equation calculation formula described in step S11 is: The equivalent circuit equation of the doubly fed system, which is equivalent to the voltage source and the grid side structure, is expressed as: In the formula, ω s is the operating frequency of VSG, R g and L g are the equivalent resistance and equivalent reactance on the line respectively; The stator and rotor flux and voltage equations of the doubly fed wind turbine are: In the formula, the subscripts d and q are the components of the electrical quantity on the d-axis and q-axis in the two-phase rotating coordinate system, respectively; the subscript s is the electrical quantity of the stator; p is the differential operator, and L m is the mutual inductance between the stator and rotor of the doubly fed wind turbine.

4. The method according to claim 2, characterized in that: The power calculation formula described in step S13 is: According to the power calculation formula: Where P s and Q s are respectively the active power and reactive power transmitted to the grid by the double-fed wind turbine under control, U gd and U gq are the dq values ​​of the grid-side voltage, ||I s || 2 is the square of the current modulus value on the stator side, where I sd and I sq They can be expressed by rotor side current and grid side factors respectively.

5. The method according to claim 1, characterized in that The power small signal model expression formula described in step S2 is: In the formula, Indicates I rd Impact on Q, Indicates I rq Impact on Q, Indicates I rd The impact on P Indicates I rq The impact on P.

6. The method according to claim 1, characterized in that The adaptive virtual impedance decoupling described in step S3 includes: In the system, virtual impedance is introduced to compensate for the power coupling caused by line impedance. The negative resistance characteristic of the virtual impedance eliminates the resistive component on the line, making it present a high inductive reactance to the outside. After the introduction, the power is expressed as follows: Considering that line impedance will change with multiple factors, in order to achieve better power decoupling, adaptive control is introduced. By detecting the target power fluctuation amplitude, the minimum power fluctuation amplitude is taken as the objective function, and the fluctuation is used as the penalty function factor, as shown in the formula: ΔP PQ =P PQ -P PQE Where: ΔP PQ is the power fluctuation generated, P PQ and P PQE Respectively represent the actual value and rated value of power; α and β are weight coefficients; dZ v / dt represents the rate of change of virtual impedance, which is used to avoid frequent jumps of virtual impedance; ΔP RMS Represents the root mean square value of the power fluctuation in the sliding window, which is used to penalize the power fluctuation. After obtaining the penalty factor, the gradient descent algorithm is used to dynamically adjust Z v , η represents the learning rate, which controls the parameter update speed; gradient Approximate calculation by numerical perturbation method; The size of the virtual impedance is dynamically adjusted according to the above formula, so that the change of the virtual impedance can adapt to the change of the line impedance, and finally achieve the power decoupling effect.

7. The method according to claim 1, characterized in that The feedforward decoupling strategy described in step S4 is as follows: After the system added a virtual impedance module for power decoupling, the DFIG-VSG system weakened the power coupling caused by the system adjusting active power or reactive power in transient conditions. The expressions of active power and reactive power contain coupling variables Ird and Irq, which means that after the introduction of virtual impedance, the control of the doubly fed wind turbine itself will cause power coupling. In order to achieve more accurate power control and precise voltage and frequency support, a feedforward decoupling strategy is further introduced on the basis of virtual impedance. Taking the target power fluctuation control amount as the feedforward object, the compensation amount is injected into the error node of the power control loop to actively offset the coupling path and reduce the active-reactive interaction. The decoupling expression after adding compensation is as follows: The system goal is to make the irrelevant quantity almost non-volatile during support, and each power disturbance needs to be kept at 0 during the control process. The compensator can be derived as:

Images