A virtual synchronous machine parallel networking optimization method based on advance-lag compensation

Abstract

In view to the dynamic oscillation problem of output active power and output frequency of a traditional virtual synchronous machine parallel networking system under load disturbance, the application discloses a virtual synchronous machine parallel networking optimization control method based on lead-lag compensation. The control method utilizes an angular frequency deviation to pass through a lead-lag compensation link to construct a phase angle feedforward compensation amount, optimizes the dynamic response performance of output active power and output frequency of the virtual synchronous machine parallel networking system by adjusting lead-lag compensation parameters, has the advantages of flexible control parameter selection, no influence on active power steady-state uniform distribution effect, elimination of active power dynamic oscillation, suppression of output frequency dynamic oscillation, and can be applied to the virtual synchronous machine parallel networking control field in power electronics technology.

Description

Technical Field

[0001] The present invention relates to the field of virtual synchronous generator control, and in particular to a virtual synchronous machine parallel networking optimization method based on lead-lag compensation, which is applicable to virtual synchronous machine parallel networking operation scenarios in power electronics technology and microgrid scenarios including such virtual synchronous machine parallel networking systems. Background Art

[0002] The Virtual Synchronous Generator (VSG) simulates the rotor motion equations of a synchronous generator, giving the VSG a certain degree of inertia support capability. However, this inevitably introduces the problem of dynamic oscillation of the output active power and output frequency, which is particularly prominent in VSG parallel network systems. Furthermore, the VSG can directly utilize the virtual damping control link and increase the virtual damping parameters to suppress or eliminate the dynamic oscillation of the output active power and output frequency of the VSG parallel network system. However, due to the coupling between its virtual damping parameters and the primary frequency modulation parameters, its virtual damping control performance and primary frequency modulation characteristics cannot be adjusted independently, which can easily affect the steady-state balancing effect of the active power of the VSG parallel network system.

[0003] To this end, people have conducted various studies, such as the article entitled "Analysis of Small Signal Model of Parallel Virtual Synchronous Generator System Based on Consensus Algorithm", published in the Proceedings of the Chinese Society of Electrical Engineering, Vol. 47, No. 02, 2022, pp. 2427-2438. Based on the analysis of the interaction mechanism of the traditional virtual synchronous generator (TVSG) parallel networking system, the sensitivity of the line structure and control structure related parameters to the dynamic oscillation of the system's active power, this article proposes a secondary frequency modulation control strategy for the TVSG parallel networking system based on the consistency algorithm, which effectively suppresses the dynamic oscillation of the system's active power. However, data communication lines need to be introduced to realize information transmission and interaction between different TVSGs, and data communication lines limit the layout maneuverability and control flexibility of the TVSG grid-connected networking system.

[0004] An article entitled "Active Power Oscillation Suppression Strategy of VSG Parallel System Based on Transient Electromagnetic Power Compensation" is published in "Power System Technology" Volume 47, Issue 01, 2023, pages 23-33. This article introduces a transient electromagnetic power compensation link into the traditional virtual synchronous generator (TVSG) parallel network control system without the need for data communication lines, and uses a first-order lag link to construct transient electromagnetic power, thereby increasing the equivalent damping of the parallel network system under transient conditions and achieving active power oscillation suppression of the TVSG parallel network system. However, the transient electromagnetic power compensation link increases the order of the control system, and the differential operation link contained therein will also introduce high-frequency harmonic interference signals, thereby affecting the operating stability of the TVSG parallel network system.

[0005] The article entitled “Active power oscillation and suppression techniques between two parallel synchronverters during load fluctuations”, SHUAI ZK, HUANG W, SHEN ZJ, et al., IEEE Transactions on Power Electronics, 2020, 35(4), 4127-4142 (“Active power oscillation and suppression techniques between two parallel synchronous inverters during load fluctuations”, IEEE Journal of Power Electronics, Vol. 35, No. 4, 2020, pp. 4127-4142). This article introduces a transient damping control strategy based on the difference of the first-order lag link of active power into the traditional virtual synchronous generator (TVSG) parallel network control system. It can effectively suppress the dynamic oscillation of the active power of the TVSG parallel network system and avoid the differential operation process. However, it has the disadvantages of increasing the order of the control system, making the design of the system control parameters complex, and the control freedom still needs to be further improved.

[0006] An article entitled "Analysis of Active Power Oscillation Characteristics of Energy Storage VSG Parallel Network System and Its Improvement Strategy" is published in "Electric Power Automation Equipment" Vol. 44, No. 05, 2024, pp. 51-57. This article uses a first-order low-pass filter link containing virtual inertia parameters and primary frequency modulation parameters to construct an active power dynamic feedforward term, and introduces the active power dynamic feedforward term into the traditional virtual synchronous generator (TVSG) parallel network control system. By adjusting the feedforward parameters, the parallel network system's ability to suppress active power dynamic oscillations is improved. Without relying on communication or requiring differential operations, the parallel network system's steady-state active power averaging effect is not affected. However, there are disadvantages in that the open-loop gain of the control system decreases in the high-frequency band, and the control freedom still needs to be further improved.

[0007] From the above, it can be seen that although the existing technology has solved the problems of dynamic oscillation and power overshoot of active power in the traditional virtual synchronous machine parallel network system, it still has disadvantages such as the need for data communication lines, high-frequency harmonic interference caused by differential operations, and the increase in system order resulting in complex parameter design and the need to further improve the control freedom. Summary of the Invention

[0008] In order to overcome the limitations of various technical solutions given in the background technology, the present invention provides a virtual synchronous machine parallel networking optimization method based on lead-lag compensation for the problem that the output active power of the traditional virtual synchronous machine parallel networking system is difficult to balance the steady-state balancing effect and dynamic oscillation due to the mutual coupling of its primary frequency modulation parameters and virtual damping parameters. This control method can effectively suppress the dynamic oscillation of the output active power and output frequency of the virtual synchronous machine grid-connected networking system without affecting the primary frequency modulation characteristics, without the need for differential operations, without the need for data communication lines, without increasing the control system order, and with the advantages of increased control freedom and more flexible control parameter selection.

[0009] In order to achieve the above object, the technical solution adopted by the present invention is:

[0010] A virtual synchronous machine parallel networking optimization method based on lead-lag compensation includes the following steps:

[0011] Step 1, in the rotor motion equation part, the active reference instruction P of the virtual synchronous machine is ref Subtract the output active power P of the virtual synchronous machine e The power difference ΔP of the virtual synchronous machine is obtained, and the power difference ΔP is converted into the power difference ΔP by the virtual inertia parameter J and the primary frequency modulation parameter k. p , and the first-order low-pass filtering link of the virtual damping parameter D to obtain the angular frequency deviation Δω, and the angular frequency deviation Δω is added to the rated angular frequency ω0 to obtain the output angular frequency ω of the virtual synchronous machine;

[0012] Step 2: Calculate the lead-lag feedforward compensation. The angular frequency deviation Δω obtained in step 1 is passed through the lead-lag compensation link including the lead compensation parameter a and the lag compensation parameter T to obtain the phase angle feedforward compensation δ of the lead-lag compensation control. d ;

[0013] Step 3: Subtract the angular frequency ω of the virtual synchronous machine parallel network system from the virtual synchronous machine output angular frequency ω obtained in step 1 bus After the integration operation, the power angle δ is obtained b , the power angle δ b Add the phase angle feedforward compensation δ obtained in step 2 d Get the power factor angle δ of the virtual synchronous machine;

[0014] Step 4: Multiply the power factor angle δ of the virtual synchronous machine obtained in step 3 by the synchronous voltage coefficient K to obtain the output active power P of the virtual synchronous machine. e ;

[0015] Step 5: Integrate the virtual synchronous machine output angular frequency ω obtained in step 1 and add the phase angle feedforward compensation δ obtained in step 2. d The output phase θ of the virtual synchronous machine is obtained, and the output phase θ is used as the phase used for dq coordinate transformation to obtain the voltage reference instruction E in the dq coordinate system for the reactive control loop. d With E q Perform dq coordinate transformation to obtain the three-phase voltage modulation signal E in the abc coordinate system a 、E b With E c , and then the three-phase voltage modulation signal E a 、E b With E c The driving signal of the virtual synchronous machine switch tube is generated through the SVPWM modulation link.

[0016] Preferably, the power difference ΔP in step 1 is calculated using the following formula:

[0017] ΔP=P ref -P e ,

[0018] The calculation formula for the angular frequency deviation Δω is:

[0019]

[0020] The calculation formula for the output angular frequency ω of the virtual synchronous machine is:

[0021] ω=Δω+ω0,

[0022] Where s is the Laplace operator.

[0023] Preferably, the phase angle feedforward compensation amount δ in step 2 is calculated by the following formula: d :

[0024]

[0025] wherein s is a Laplace operator.

[0026] Preferably, the power angle δ in step 3 is calculated by the following formula: b

[0027]

[0028] The calculation formula of the power factor angle δ of the virtual synchronous machine is as follows:

[0029] δ = δ b + δ d ,

[0030] wherein s is a Laplace operator.

[0031] Preferably, the output active power P of the virtual synchronous machine in step 4 is calculated by the following formula: e

[0032]

[0033] wherein U bus is the system bus voltage amplitude, E is the output voltage amplitude of the virtual synchronous machine, and X is the equivalent inductive reactance of the line.

[0034] Preferably, the output phase θ of the virtual synchronous machine in step 5 is calculated by the following formula:

[0035]

[0036] The calculation formulas of the voltage reference instruction E d and E q in the dq coordinate system are as follows:

[0037]

[0038] wherein Q ref is the reactive reference instruction of the virtual synchronous machine, Q e is the output reactive power of the virtual synchronous machine, k p is a primary voltage regulation parameter, E0 is the voltage reference instruction of the virtual synchronous machine, and s is a Laplace operator.

[0039] Compared with the prior art, the present application has the following beneficial effects:

[0040] ​​The application provides a virtual synchronous machine parallel networking optimization method based on lead-lag compensation. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 Figure 1 is a virtual synchronous machine control structure diagram of an embodiment of the application.

[0042] Figure 2 Figure 2 is a compensation quantity calculation schematic diagram of the lead-lag feedforward compensation control.

[0043] Figure 3 Figure 3 is a virtual synchronous machine parallel networking system circuit topology structure diagram of an embodiment of the application.

[0044] Figure 4 Figure 4 is a virtual synchronous machine parallel networking system output active closed-loop equivalent control structure diagram of an embodiment of the application.

[0045] Figure 5 Figure 5 is a coordinate transformation and modulation schematic diagram of an embodiment of the application.

[0046] Figure 6 Figure 6 is a virtual synchronous machine parallel networking system simulation waveform comparison diagram before and after the application.

[0047] Figure 7 Figure 7 is a virtual synchronous machine parallel networking system experimental waveform comparison diagram before and after the application. DETAILED DESCRIPTION

[0048] The following detailed description will be further illustrated with reference to the above drawings, and the specific description is as follows:

[0049] Please refer to Figure 1 The application provides a virtual synchronous machine parallel networking optimization method based on lead-lag compensation, which comprises the following steps:

[0050] Step 1, rotor motion equation part, subtract the output active power P ref of the virtual synchronous machine from the active reference instruction P e of the virtual synchronous machine to obtain the power difference ΔP of the virtual synchronous machine, and pass the power difference ΔP through the virtual inertia parameter J, the primary frequency modulation parameter k p, and the first-order low-pass filtering link of the virtual damping parameter D to obtain the angular frequency deviation Δω, and the angular frequency deviation Δω is added to the rated angular frequency ω0 to obtain the output angular frequency ω of the virtual synchronous machine.

[0051] The power difference ΔP in step 1 is calculated using the following formula:

[0052] ΔP=P ref -P e ,

[0053] The calculation formula for the angular frequency deviation Δω is:

[0054]

[0055] The calculation formula for the output angular frequency ω of the virtual synchronous machine is:

[0056] ω=Δω+ω0,

[0057] Where s is the Laplace operator.

[0058] Step 2: Calculate the lead-lag feedforward compensation, as follows: Figure 2 As shown, the angular frequency deviation Δω obtained in step 1 is passed through the lead-lag compensation link including the lead compensation parameter a and the lag compensation parameter T to obtain the phase angle feedforward compensation δ of the lead-lag compensation control. d .

[0059] Among them, the lead-lag feedforward compensation δ d The calculation formula used is:

[0060]

[0061] Where s is the Laplace operator.

[0062] Step 3: Subtract the angular frequency ω of the virtual synchronous machine parallel network system from the virtual synchronous machine output angular frequency ω obtained in step 1 bus After the integration operation, the power angle δ is obtained b , the power angle δ b Add the phase angle feedforward compensation δ obtained in step 2 d The power factor angle δ of the virtual synchronous machine is obtained.

[0063] Among them, the power angle δ b The calculation formula used is:

[0064]

[0065] The calculation formula for the power factor angle δ of the virtual synchronous machine is:

[0066] δ=δ b +δ d,

[0067] Where s is the Laplace operator.

[0068] Step 4: Multiply the power factor angle δ of the virtual synchronous machine obtained in step 3 by the synchronous voltage coefficient K to obtain the output active power P of the virtual synchronous machine. e .

[0069] Among them, the output active power of the virtual synchronous machine P e The calculation formula used is:

[0070]

[0071] Where U bus is the system bus voltage amplitude, E is the output voltage amplitude of the virtual synchronous machine, and X is the line equivalent inductive reactance.

[0072] Combine Figure 1 The power transfer model and Figure 3 The circuit topology diagram of the virtual synchronous machine parallel network system is shown in Figure 1. Based on the above control, the output active closed-loop equivalent control structure diagram of the virtual synchronous machine parallel network system can be obtained as shown in Figure 1. Figure 4 shown. Figure 4 In, P load The active power consumed by the load of the virtual synchronous machine parallel network system. The subscripts "1" and "2" are used to distinguish the first and second virtual synchronous machines in the virtual synchronous machine parallel network system, respectively.

[0073] Step 5, such as Figure 5 As shown, the virtual synchronous machine output angular frequency ω obtained in step 1 is integrated and added with the phase angle feedforward compensation δ obtained in step 2. d The output phase θ of the virtual synchronous machine is obtained, and the output phase θ is used as the phase used for dq coordinate transformation to obtain the voltage reference instruction E in the dq coordinate system for the reactive control loop. d With E q Perform dq coordinate transformation to obtain the three-phase voltage modulation signal E in the abc coordinate system a 、E b With E c , and then the three-phase voltage modulation signal E a 、E b With E c The driving signal of the virtual synchronous machine switch tube is generated through the SVPWM modulation link.

[0074] The calculation formula for the output phase θ of the virtual synchronous machine is:

[0075]

[0076] Voltage reference instruction E in dq coordinate systemd With E q The calculation formulas used are:

[0077]

[0078] Where Q ref is the reactive power reference instruction of the virtual synchronous machine, Q e is the output reactive power of the virtual synchronous machine, k p is the primary voltage regulation parameter, E0 is the voltage reference instruction of the virtual synchronous machine, and s is the Laplace operator.

[0079] Example

[0080] In order to verify the control effect of the virtual synchronous machine (LLC-VSG) parallel networking optimization method based on lead-lag compensation proposed in the present invention, the LLC-VSG parallel networking optimization method proposed in the present invention is compared with the existing traditional virtual synchronous machine (TVSG) parallel networking control method (the TVSG parallel networking control method is mentioned in the background technology and is given in the article entitled "Analysis of active oscillation characteristics of energy storage VSG parallel networking system and its improvement strategy", "Electric Power Automation Equipment" Vol. 44, No. 05, 2024, pp. 51-57) under two conditions of whether the proposed lead-lag feedforward compensation is introduced or not. Simulation and experimental comparison tests are mainly compared in terms of responding to the active power P consumed by the load. load Under the condition of step disturbance, the output frequency f1 and f2 of the virtual synchronous machine parallel network system and the output active power P e1 With P e2 , and the response performance of the system frequency f. The details are as follows:

[0081] First, set the relevant parameters. In this embodiment, the relevant parameters in the LLC-VSG parallel networking optimization method of the present invention are set as follows:

[0082] The virtual synchronous machine parallel network system mainly includes a VSG1 with a rated capacity of 100kVA and a VSG2 with a rated capacity of 50kVA. refl The active reference instruction P is 40kW. ref2 The rated angular frequency ω0 is 314.16 rad / s, the virtual inertia parameter J1 is 6 kg·m2, and the virtual inertia parameter J2 is 3 kg·m 2 , the primary frequency modulation parameter k p1 is 15915.5J / rad, and the primary frequency modulation parameter k p2 is 7957.8J / rad, and the system bus voltage amplitude U bus311V, the output voltage amplitude E1 and E2 of the virtual synchronous machine is 311V, the line equivalent inductance X1 is 0.2Ω, the line equivalent inductance X2 is 0.1Ω, the virtual damping parameter D1 is 200J / rad, the virtual damping parameter D2 is 100J / rad, then the synchronization voltage coefficient K1=1.5U bus E1 / X1=725407.5, the synchronization voltage coefficient K2=1.5U bus E2 / X2=1450815, at this time, the hysteresis compensation parameter T1 is set to 6, the hysteresis compensation parameter T2 is set to 3, the advance compensation parameter a1 is set to 0.07, and the advance compensation parameter a2 is set to 0.035.

[0083] Based on the above parameter setting, simulation and experimental comparison test are carried out, and the specific conditions are as follows:

[0084] The simulation working condition is set as follows: the system load P load At the initial moment, from 60kW to 90kW.

[0085] The experimental working condition is set as follows: at the initial moment, VSG1 and VSG2 are connected in parallel to form a network system, and the system is stably operated by driving 60kW of resistive load, at 1s, 60kW of resistive step load is input, and at 1.6s, 60kW of resistive step load is cut off.

[0086] According to the above simulation working condition, the output frequency simulation waveform comparison diagram of the virtual synchronous machine parallel network system before and after the application is shown in Figure 6

[0087] According to Figure 6 (a), it can be seen that the parallel network system corresponding to the existing TVSG parallel network control method has a large dynamic oscillation in the process of the system load P load from 60kW to 90kW, and the TVSG1 output frequency f1 and the TVSG2 output frequency f2 also have a large dynamic oscillation; and according to Figure 6 (b), it can be seen that the parallel network system corresponding to the LLC-VSG parallel network control method of the application has no dynamic oscillation in the process of the system load P load from 60kW to 90kW, and the LLC-VSG1 output frequency f1 and the LLC-VSG2 output frequency f2 also have no dynamic oscillation. Therefore, the LLC-VSG parallel network control method of the application can effectively solve the dynamic oscillation problem of the output frequency of the parallel network system corresponding to the existing TVSG parallel network control method under the system load step disturbance.

[0088] According to the above experimental working condition, the output active power and system frequency experimental waveform comparison diagram of the virtual synchronous machine parallel network system before and after the application is shown in Figure 7 ​​

[0089] According to Figure 7 (a) can be seen, I, the existing TVSG parallel networking control method corresponding to the parallel networking system in the system load P load TVSG1 output active power P e1 TVSG2 output active power P e2 There is a large dynamic oscillation; two, the LLC-VSG parallel networking control method proposed in the present application corresponding to the parallel networking system in the system load P load LLC-VSG1 output active power P e1 LLC-VSG2 output active power P e2 There is no dynamic oscillation. It can be seen that the LLC-VSG parallel networking control method proposed in the present application can effectively solve the dynamic oscillation problem of the output active power of the parallel networking system corresponding to the existing TVSG parallel networking control method under the system load step disturbance.

[0090] According to Figure 7 (b) can be seen, I, the existing TVSG parallel networking control method corresponding to the parallel networking system in the system load P load There is a large dynamic oscillation of the system frequency; two, the LLC-VSG parallel networking control method proposed in the present application corresponding to the parallel networking system in the system load P load There is no dynamic oscillation of the system frequency f. It can be seen that the LLC-VSG parallel networking control method proposed in the present application can effectively solve the dynamic oscillation problem of the system frequency of the parallel networking system corresponding to the existing TVSG parallel networking control method under the system load step disturbance.

[0091] The results of Figure 7 and Figure 6 It is not difficult to see that the experimental test comparison results in the embodiment Figure 7 can be one-to-one corresponding with the simulation test comparison results in Figure 6 Both fully show that the LLC-VSG parallel networking control method proposed in the present application can effectively solve the dynamic oscillation problem of the output active power, the output frequency and the system frequency of the parallel networking system corresponding to the existing TVSG parallel networking control method under the system load P load Step disturbance, therefore, compared with the existing TVSG parallel networking control method, the control effect of the LLC-VSG parallel networking control method proposed in the present application is better.

[0092] The above description is a detailed description of the preferred embodiments of the present invention, but the embodiments are not intended to limit the scope of the patent application of the present invention. Any equivalent changes or modifications completed under the technical spirit suggested by the present invention should fall within the patent scope covered by the present invention.

Claims

1. A virtual synchronous machine parallel networking optimization method based on lead-lag compensation, characterized in that: The steps include: Step 1, in the rotor motion equation part, the active reference instruction P of the virtual synchronous machine is ref Subtract the output active power P of the virtual synchronous machine e The power difference ΔP of the virtual synchronous machine is obtained, and the power difference ΔP is converted into the power difference ΔP by the virtual inertia parameter J and the primary frequency modulation parameter k. p , and the first-order low-pass filtering link of the virtual damping parameter D to obtain the angular frequency deviation Δω, and the angular frequency deviation Δω is added to the rated angular frequency ω0 to obtain the output angular frequency ω of the virtual synchronous machine; Step 2: Calculate the lead-lag feedforward compensation. The angular frequency deviation Δω obtained in step 1 is passed through the lead-lag compensation link including the lead compensation parameter a and the lag compensation parameter T to obtain the phase angle feedforward compensation δ of the lead-lag compensation control. d ; Step 3: Subtract the angular frequency ω of the virtual synchronous machine parallel network system from the virtual synchronous machine output angular frequency ω obtained in step 1 bus After the integration operation, the power angle δ is obtained b , the power angle δ b Add the phase angle feedforward compensation δ obtained in step 2 d Get the power factor angle δ of the virtual synchronous machine; Step 4: Multiply the power factor angle δ of the virtual synchronous machine obtained in step 3 by the synchronous voltage coefficient K to obtain the output active power P of the virtual synchronous machine. e ; Step 5: Integrate the virtual synchronous machine output angular frequency ω obtained in step 1 and add the phase angle feedforward compensation δ obtained in step 2. d The output phase θ of the virtual synchronous machine is obtained, and the output phase θ is used as the phase used for dq coordinate transformation to obtain the voltage reference instruction E in the dq coordinate system for the reactive control loop. d With E q Perform dq coordinate transformation to obtain the three-phase voltage modulation signal E in the abc coordinate system a 、E b With E c , and then the three-phase voltage modulation signal E a 、E b With E c The driving signal of the virtual synchronous machine switch tube is generated through the SVPWM modulation link.

2. The virtual synchronous machine parallel networking optimization method based on lead-lag compensation according to claim 1 is characterized in that: The power difference ΔP in step 1 is calculated using the following formula: ΔP=P ref -P e , The calculation formula for the angular frequency deviation Δω is: The calculation formula for the output angular frequency ω of the virtual synchronous machine is: ω=Δω+ω0, Where s is the Laplace operator.

3. The virtual synchronous machine parallel network optimization method based on lead-lag compensation according to claim 1 is characterized in that: Phase angle feedforward compensation δ of lead-lag compensation control in step 2 d : Where s is the Laplace operator.

4. The virtual synchronous machine parallel network optimization method based on lead-lag compensation according to claim 1 is characterized in that: The power angle δ in step 3 b The calculation formula used is: The calculation formula for the power factor angle δ of the virtual synchronous machine is: d=d b +d d , Where s is the Laplace operator.

5. The virtual synchronous machine parallel network optimization method based on lead-lag compensation according to claim 1 is characterized in that: The output active power P of the virtual synchronous machine in step 4 e The calculation formula used is: Where U bus is the system bus voltage amplitude, E is the output voltage amplitude of the virtual synchronous machine, and X is the line equivalent inductive reactance.

6. The virtual synchronous machine parallel network optimization method based on lead-lag compensation according to claim 1, characterized in that: The calculation formula for the output phase θ of the virtual synchronous machine in step 5 is: Voltage reference instruction E in dq coordinate system d With E q The calculation formulas used are: Where Q ref is the reactive power reference instruction of the virtual synchronous machine, Q e is the output reactive power of the virtual synchronous machine, k q is the primary voltage regulation parameter, E0 is the voltage reference instruction of the virtual synchronous machine, and s is the Laplace operator.

Images