Self-adaptive control method for accurately distributing reactive power for multiple network-forming converters

Abstract

The invention discloses a self-adaptive control method for accurately distributing reactive power for a plurality of network-building type converters, and belongs to the field of parallel control of the plurality of network-building type converters. According to the method, an output voltage phase angle reference value is obtained by adopting virtual synchronous generator active loop control; a d-axis voltage component instruction value and a q-axis voltage component instruction value are obtained through virtual synchronous generator reactive loop control and an optimal distribution algorithm, then modulation waves needed by PWM control are obtained through voltage and current double-closed-loop control, and after PWM modulation is conducted, switching tube driving signals are output to complete control over a converter. On the premise that line impedance information is not obtained, voltage drop is accurately compensated through a self-adaptive control method, reactive power distribution errors are reduced, reactive power is accurately distributed according to rated capacity, and system stability is improved.

Description

Technical Field

[0001] The invention belongs to the field of parallel control of multiple grid-forming converters, and in particular relates to an adaptive control method for accurately distributing reactive power of multiple grid-forming converters. Background Art

[0002] Common grid-type converter control technologies include droop control technology and virtual synchronous generator control technology. Among them, virtual synchronous generator technology simulates the rotor inertia, damping characteristics and droop external characteristics of the synchronous generator, so that the inverter has the same steady-state and dynamic characteristics as the synchronous generator, providing damping and inertial support for the system during operation, thereby improving system stability. However, when a system composed of multiple grid-type converters in parallel operates in island mode, due to the random distribution of the geographical locations of each converter, there are differences in the line impedance from the converter to the PCC point. Therefore, the system has the problem that the actual output reactive power is difficult to match the rated capacity, and reasonable reactive power distribution cannot be achieved.

[0003] At present, many scholars at home and abroad have proposed a variety of improvement methods for the problem of reasonable allocation of reactive power, including line impedance parameter identification method, adaptive virtual impedance method, etc., which can achieve reasonable allocation of reactive power and improve system stability by obtaining corresponding information to reasonably compensate for voltage drop. For example:

[0004] 1. In the article entitled “Distributed Control Strategy Based on Line Impedance Identification for Reactive Power Sharing in Microgrid, 2023 IEEE Energy Conversion Congress and Exposition (ECCE), 2023, 1160-1167, the line parameter values ​​are accurately observed by the observer, and then the line voltage drop of each converter is calculated using the observed line parameters, and the reasonable distribution of system reactive power is achieved by accurately compensating the voltage. However, the line parameter identification scheme has high requirements on system communication and data measurement accuracy, and is difficult to be applied in actual engineering.

[0005] 2. The article entitled "Improved droop control strategy for reactive power sharing of inverters in low-voltage microgrids" by Bai Xiaodan, Miao Hong, Zeng Chengbi, and Most, "High Voltage Technology", Vol. 46, No. 04, 2020, pp. 1310-1318, collects information from each converter through a central controller for calculation, which is used to reasonably design the dynamic virtual impedance part of the adaptive virtual impedance, compensate for the voltage drop caused by line impedance differences, and thus reasonably distribute reactive power. However, the centralized control method using a central controller has high requirements on communication quality, and is prone to inaccurate collection information due to communication interruptions or delays, so the virtual impedance parameters cannot accurately adapt to the system.

[0006] 3. "Research on Reactive Power Distribution of Isolated AC Microgrid Based on Improved Droop Control" Yan Li, Mi Yang, Sun Wei, Cai Hangyi, Niu Qingsong, Qi Fujun, Acta Energiae Solaris Sinica, Vol. 42, No. 08, 2021, pp. 07-15. The article collects reactive power information of adjacent converters, obtains the parameters required to build an adaptive virtual impedance algorithm through a consistency algorithm, and realizes reasonable reactive power distribution while reducing dependence on global communication. However, it is only applicable to the case where the capacity of each converter is equal. When the capacity is unequal, the reactive power cannot be reasonably distributed according to the capacity ratio.

[0007] In summary, existing research mainly focuses on parameter identification methods and adaptive virtual impedance methods, and there are still the following technical problems:

[0008] 1. The recognition effect requires high data measurement accuracy, which is difficult to measure and difficult to apply in actual engineering;

[0009] 2. Larger virtual impedance will reduce the system voltage control accuracy and affect the dynamic characteristics of the system output voltage;

[0010] 3. The system has high requirements for communication quality and is overly dependent on global communication. Fluctuations in communication quality will lead to differences in control effects.

[0011] 4. With respect to the restriction that the capacity of each converter in the system is equal, if the capacity of each converter is not equal, the target control effect cannot be achieved. Summary of the invention

[0012] The technical problem to be solved by the present invention is: when multiple grid-type converters based on virtual synchronous generator technology are connected in parallel to the load, the problem of accurate distribution of reactive power output by each converter. The present invention provides an adaptive control method for accurately distributing reactive power of multiple grid-type converters, so that each converter can accurately share the reactive load and ensure the stable operation of the system.

[0013] The technical solution of the present invention is as follows.

[0014] An adaptive control method for accurately distributing reactive power of multiple grid-type converters, the adaptive control method involves a system composed of X grid-type converters based on virtual synchronous generator technology, each converter is connected to a load in parallel, any one of the grid-type converters is recorded as grid-type converter i, i∈[1,X]; the number of grid-type converters adjacent to the grid-type converter i is Y, and any one of them is recorded as adjacent grid-type converter j, j∈[1,Y], Y<X; a distributed communication network of X grid-type converters is constructed in the system, specifically, the grid-type converter i can send information to the adjacent grid-type converter j, and receive information sent by the adjacent grid-type converter j;

[0015] The adaptive control method comprises the following steps:

[0016] Step 1: Sample the three-phase output voltage u at the output end of the grid-connected converter i ai ,u bi ,u ci And the three-phase output current i ai ,i bi ,i ci After coordinate transformation, the dq axis output voltage component U in the synchronous rotating coordinate system is obtained di ,U qi and dq axis output current component I di ,I qi Then, the output active power P of the grid-connected converter i is obtained by power calculation. ei And the output reactive power Q ei ;

[0017] Step 2: According to the output active power P of the grid-connected converter i ei And the output reactive power Q ei Perform virtual synchronous generator control. Specifically, the output voltage angular frequency reference value ω is obtained through the active loop control equation i , and get the output voltage phase angle reference value θ i ,θ i =∫ω i ; Given reactive power-voltage droop coefficient k ni , the voltage reference value U is obtained through the reactive loop control equation refi ;

[0018] Step 3: The grid-connected converter i obtains the output reactive power Q of the adjacent grid-connected converter j through communication. ej and reactive power-voltage droop coefficient k nj , combined with the output reactive power Q of the grid-connected converter i ei And reactive power-voltage droop coefficient k ni, the virtual capacitor droop coefficient compensation value Δn is obtained through the multi-agent consensus algorithm ci and virtual inductance compensation value ΔL vi ;

[0019] Step 4: According to the voltage reference value U refi and virtual capacitor droop coefficient compensation value Δn ci , the d-axis virtual voltage standard value component U is calculated by the virtual capacitance algorithm rdi ; Set the q-axis virtual voltage standard value component U rqi Set to 0, and then according to the d-axis virtual voltage standard value component U rdi and virtual inductance compensation value ΔL vi , the dq axis voltage component command value U is calculated by the virtual inductance algorithm clrdi ,U clrqi ;

[0020] Step 5: According to the dq axis voltage component command value U clrdi ,U clrqi and the output voltage phase angle reference value θ i , perform voltage and current double closed-loop control of the grid-type converter i. Specifically, the output I of the dq axis output voltage outer loop control is obtained by calculating the dq axis output voltage outer loop control equation. rdi ,I rqi , and then the output U of the dq axis output current inner loop control equation is obtained pdi ,U pqi , and then the three-phase modulation wave signal u is obtained through coordinate transformation pai ,u pbi ,u pci , and generate the switch tube driving signal of the grid-type converter i.

[0021] Preferably, the output active power P of the grid-connected converter i in step 1 is ei And output reactive power Q ei The calculation formulas are:

[0022]

[0023] Preferably, the active loop control equation in step 2 is:

[0024]

[0025] Among them, ω 0i is the rated angular frequency of the system voltage, ω i ′ is the output voltage angular frequency reference value of the previous control cycle, k ωi is the primary frequency modulation coefficient, P refi is the active power command value, J iis the virtual inertia;

[0026] The reactive loop control equation is:

[0027] U refi =U 0i -k ni (Q ei -Q refi )

[0028] Among them, U 0i is the system rated output voltage amplitude, Q refi is the reactive power command value.

[0029] Preferably, the expression of the multi-agent consensus algorithm in step 3 is:

[0030]

[0031] Among them, C Qi is the control gain, a ij is the weight coefficient set according to whether the grid-type converter i can communicate with the adjacent grid-type converter j. If communication is possible, then a ij =1, if communication is unavailable then a ij =0, k pyi is the proportional coefficient in the multi-agent consensus algorithm, k iyi is the integral coefficient of the multi-agent consensus algorithm, and s is the Laplace operator.

[0032] Preferably, the d-axis virtual voltage component standard value U in step 4 is rdi The calculation formula is:

[0033]

[0034] Among them, U maxi is the maximum voltage amplitude allowed by the system, Q ratei is the rated reactive power of grid-connected converter i, n rci is the fixed value of the virtual capacitor droop coefficient, K Qci is the virtual capacitor proportional gain.

[0035] The dq axis voltage component command value U clrdi ,U clrqi The calculation formula is:

[0036]

[0037] Among them, ω 0i is the rated angular frequency of the system voltage, L rvi is the fixed value of virtual inductance, K Qli is the virtual inductor proportional gain.

[0038] Preferably, the dq axis output voltage outer loop control equation in step 5 is:

[0039]

[0040] In the formula, k spvi is the proportional coefficient of the output voltage outer loop proportional-integral controller, k sivi is the integral coefficient of the output voltage outer loop proportional-integral controller, and s is the Laplace operator;

[0041] The inner loop control equation of the dq axis output current is:

[0042]

[0043] Among them, k spii is the proportional coefficient of the output current inner loop proportional-integral controller, k siii is the integral coefficient of the output current inner loop proportional-integral controller.

[0044] Compared with the existing reactive power distribution scheme based on line impedance identification and the reactive power distribution scheme based on adaptive virtual impedance algorithm, the adaptive control method for accurately distributing reactive power of multiple grid-forming converters disclosed in the present invention has the following beneficial effects:

[0045] 1. This adaptive control method adds virtual capacitor algorithm and virtual inductor algorithm to the conventional grid-connected converter control based on virtual synchronous generator technology to compensate for the system voltage drop, thus realizing reactive power distribution according to the rated capacity of the system;

[0046] 2. This adaptive control method uses a multi-agent consensus algorithm to perform adaptive control on the virtual capacitor algorithm and the virtual inductor algorithm. Without obtaining any line impedance information, only the reactive-voltage droop coefficient and reactive information of the adjacent grid-type converter are obtained to obtain the algorithm parameters of the matching system. The parameter design method is simple, accurate and reliable, which is conducive to engineering application.

[0047] 3. This adaptive control method can reduce the dependence on global communication, and the reactive power distribution effect is not affected by communication delays and load mutations, and has high reliability in dealing with environmental and system changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a control structure diagram of the X-type grid-connected converter in the present invention connected in parallel to the load.

[0049] Figure 2 This is a control structure diagram of a grid-type converter based on adaptive control according to an embodiment of the present invention.

[0050] Figure 3The output reactive power Q of the parallel system of two grid-connected converters is calculated without introducing the virtual capacitor algorithm and the virtual inductor algorithm in the embodiment of the present invention. e1 ,Q e2 Waveform and output reactive power distribution error Q error Waveform diagram, Qerror = (Qrate2Qe1 - Qrate1Qe2) / Qload.

[0051] Figure 4 When the virtual capacitor algorithm and virtual inductor algorithm are introduced into the embodiment of the present invention and the multi-agent consensus algorithm is used for adaptive control, the output reactive power Q of the parallel system of two grid-connected converters without any environmental and system changes is e1 ,Q e2 Waveform and output reactive power distribution error Q error Waveform graph.

[0052] Figure 5 When the virtual capacitor algorithm and virtual inductor algorithm are introduced into the embodiment of the present invention and the multi-agent consensus algorithm is used for adaptive control, the output reactive power Q of the parallel system of two grid-connected converters is e1 ,Q e2 Waveform and output reactive power distribution error Q error Waveform graph.

[0053] Figure 6 The output reactive power Q of the parallel system of two grid-connected converters is Q1 when the virtual capacitor algorithm and virtual inductor algorithm are introduced into the embodiment of the present invention and the multi-agent consensus algorithm is used for adaptive control. e1 ,Q e2 Waveform and output reactive power distribution error Q error Waveform graph. DETAILED DESCRIPTION

[0054] The present embodiment is described in detail below with reference to the accompanying drawings.

[0055] Figure 1 The control structure diagram of the X-station grid-type converter connected in parallel to the load in the present invention is as follows: Figure 1 It can be seen that the present invention provides an adaptive control method for accurately distributing reactive power of multiple grid-type converters, and the adaptive control method involves a system composed of X grid-type converters based on virtual synchronous generator technology, each converter is connected to the load in parallel, and any one of the grid-type converters is recorded as grid-type converter i, i∈[1,X]. Let the number of grid-type converters adjacent to the grid-type converter i be Y, and any one of them is recorded as adjacent grid-type converter j, j∈[1,Y], Y<X.

[0056] A distributed communication network of X meshed converters is constructed in the system. Specifically, meshed converter i can send information to an adjacent meshed converter j, and receive information sent by an adjacent meshed converter j.

[0057] In addition, by Figure 1 It can be seen that a CL filter and a line impedance are connected in series between each grid-connected converter and the load. fi is the inductance of the filter inductor, C fi is the capacitance of the filter capacitor, L li is the inductance value in the line impedance, r li is the resistance value of the line impedance.

[0058] In this embodiment, X=2 and Y=1 are selected. That is, there are two grid-type converters in the system, where the DC side voltage of grid-type converter 1 is 600V, and L fi 0.5mH, C fi The line impedance is 0.09mF, and the inductance value is L li is 0.65mH, and the resistance value r in the line impedance li The inductance value L in the line impedance of the grid-type converter 2 is 0.65Ω. li is 0.7mH, and the resistance value r in the line impedance li is 0.7Ω, and other parameters are the same. Rated capacity Q of grid-type converter 1 rate1 100kVA, grid-type converter 2 rated capacity Q rate2 It is 80kVA.

[0059] Figure 2 FIG. 1 is a control structure diagram of a grid-type converter based on adaptive control according to an embodiment of the present invention. Figure 1 and Figure 2 It can be seen that the adaptive control method of the present invention comprises the following steps:

[0060] Step 1: Sample the three-phase output voltage u at the output end of the grid-connected converter i ai ,u bi ,u ci And the three-phase output current i ai ,i bi ,i ci After coordinate transformation, the dq axis output voltage component U in the synchronous rotating coordinate system is obtained di ,U qi and dq axis output current component I di ,I qi Then, the output active power P of the grid-connected converter i is obtained by power calculation. ei And the output reactive power Q ei.

[0061] In this embodiment, the output active power P of the grid-connected converter i is ei And output reactive power Q ei The calculation formulas are:

[0062]

[0063] Step 2: According to the output active power P of the grid-connected converter i ei And the output reactive power Q ei Perform virtual synchronous generator control. Specifically, the output voltage angular frequency reference value ω is obtained through the active loop control equation i , and get the output voltage phase angle reference value θ i ,θ i =∫ω i ; Given reactive power-voltage droop coefficient k ni , the voltage reference value U is obtained through the reactive loop control equation refi .

[0064] In this embodiment, the active loop control equation is:

[0065]

[0066] Among them, ω 0i is the rated angular frequency of the system voltage, ω i ′ is the output voltage angular frequency reference value of the previous control cycle, k ωi is the primary frequency modulation coefficient, P refi is the active power command value, J i is the virtual inertia;

[0067] The reactive loop control equation is:

[0068] U refi =U 0i -k ni (Q ei -Q refi )

[0069] Among them, U 0i is the system rated output voltage amplitude, Q refi is the reactive power command value.

[0070] In this embodiment 0i =314.159rad / s,k ωi =31831Ws / rad,P refi =0kW,J i =1kgm 2 .U 0i =311V, i=1 when kni =0.00012V / kVA, i=2 when k ni =0.00015V / kVA,Q refi =0kVA.

[0071] Step 3: The grid-connected converter i obtains the output reactive power Q of the adjacent grid-connected converter j through communication. ej and reactive power-voltage droop coefficient k nj , combined with the output reactive power Q of the grid-connected converter i ei And reactive power-voltage droop coefficient k ni , the virtual capacitor droop coefficient compensation value Δn is obtained through the multi-agent consensus algorithm ci and virtual inductance compensation value ΔL vi .

[0072] In this embodiment, the expression of the multi-agent consensus algorithm is:

[0073]

[0074] Among them, C Qi is the control gain, a ij is the weight coefficient set according to whether the grid-type converter i can communicate with the adjacent grid-type converter j. If communication is possible, then a ij =1, if communication is unavailable then a ij =0, k pyi is the proportional coefficient in the multi-agent consensus algorithm, k iyi is the integral coefficient of the multi-agent consensus algorithm, and s is the Laplace operator.

[0075] In this embodiment, when i=1, k nj =0.00015V / kVA, when i=2, k nj =0.00012V / kVA. ij =1,C Qci =10,k pyi =0.005,k iyi =1.

[0076] Step 4: According to the voltage reference value U refi and virtual capacitor droop coefficient compensation value Δn ci , the d-axis virtual voltage standard value component U is calculated by the virtual capacitance algorithm rdi ; Set the q-axis virtual voltage standard value component U rqi Set to 0, and then according to the d-axis virtual voltage standard value component U rdi and virtual inductance compensation value ΔL vi , the dq axis voltage component command value U is calculated by the virtual inductance algorithmclrdi ,U clrqi .

[0077] In this embodiment, the d-axis virtual voltage component standard value U rdi The calculation formula is:

[0078]

[0079] Among them, U maxi is the maximum voltage amplitude allowed by the system, Q ratei is the rated reactive power of grid-connected converter i, n rci is the fixed value of the virtual capacitor droop coefficient, K Qci is the virtual capacitor proportional gain.

[0080] The dq axis voltage component command value U clrdi ,U clrqi The calculation formula is:

[0081]

[0082] Among them, ω 0i is the rated angular frequency of the system voltage, L rvi is the fixed value of virtual inductance, K Qli is the virtual inductor proportional gain.

[0083] In this embodiment, U maxi =231.48V, i=1, n rci =1F / kVA, L rvi =0.65mH, K Qci =0.001, K Qli =0.001; when i=2, n rci =0.8F / kVA, L rvi =0.2mH, K Qci =0.001, K Qli =0.001.

[0084] Step 5: According to the dq axis voltage component command value U clrdi ,U clrqi and the output voltage phase angle reference value θ i , perform voltage and current double closed-loop control of the grid-type converter i. Specifically, the output I of the dq axis output voltage outer loop control is obtained by calculating the dq axis output voltage outer loop control equation. rdi ,I rqi , and then the output U of the dq axis output current inner loop control equation is obtained pdi ,U pqi , and then the three-phase modulation wave signal u is obtained through coordinate transformation pai ,upbi ,u pci , and generate the switch tube driving signal of the grid-type converter i.

[0085] In this embodiment, the dq axis output voltage outer loop control equation is:

[0086]

[0087] In the formula, k spvi is the proportional coefficient of the output voltage outer loop proportional-integral controller, k sivi is the integral coefficient of the output voltage outer loop proportional-integral controller, and s is the Laplace operator;

[0088] The inner loop control equation of the dq axis output current is:

[0089]

[0090] Among them, k spii is the proportional coefficient of the output current inner loop proportional-integral controller, k siii is the integral coefficient of the output current inner loop proportional-integral controller.

[0091] In this embodiment, k spvi =0.6,k sivi =800,k spii =0.8, k siii =0.

[0092] The invention in this embodiment is applicable to a parallel system of multiple grid-connected converters based on virtual synchronous generator technology. When the system operates in off-grid mode, the reactive power output of each grid-connected converter is reasonably distributed. Figure 2 A parallel system of two grid-connected converters with the control method shown was simulated.

[0093] Figure 3 When the virtual capacitor algorithm and virtual inductor algorithm are not introduced for the two grid-connected converters, the system output reactive power Q ei and reactive power sharing error (Q rate2 Q e1 -Q rate1 Q e2 / Q load ). It can be seen that due to the difference in line impedance between the two grid-type converters connected to the load, in this case, the ratio of the reactive power output of converter 1 and converter 2 is approximately 1.09, which is different from the rated reactive power ratio of 1.25, and the reactive power distribution error Q error The value is roughly -3500Var.

[0094] Figure 4The virtual capacitor algorithm and virtual inductor algorithm are introduced into both grid-connected converters, and the multi-agent consensus algorithm is used for adaptive control. The system outputs reactive power Q ei and reactive power sharing error (Q rate2 Q e1 -Q rate1 Q e2 / Q load ). It can be seen that at this time, the ratio of the reactive power output of converter 1 and converter 2 is close to 1.25, which is equal to the rated reactive power ratio, and the reactive power distribution error Q error The value is close to 0, which realizes the accurate distribution of output reactive power according to the rated capacity. The virtual capacitor algorithm and virtual inductor algorithm optimized based on the multi-agent consensus algorithm effectively solve the problem that reactive power cannot be reasonably distributed.

[0095] Figure 5 After introducing virtual capacitor algorithm and virtual inductor algorithm for both grid-connected converters and using multi-agent consensus algorithm for adaptive control, when the reactive load suddenly changes from 80kVA to 60kVA in 3s, the system output reactive power Q ei and reactive power sharing error Q error It can be seen that the ratio of the reactive power output of converter 1 and converter 2 before and after the load mutation is close to 1.25, which is the same as its rated reactive power ratio of 1.25, and the reactive power distribution error Q error The values ​​of are close to 0, which realizes the accurate distribution of output reactive power according to rated capacity, and the optimization distribution effect is not affected by load changes.

[0096] Figure 6 The virtual capacitor algorithm and virtual inductor algorithm are introduced for both grid-connected converters, and the multi-agent consensus algorithm is used for adaptive control. When there is a 30ms delay in system communication, the system outputs reactive power Q ei and reactive power sharing error Q error It can be seen that the ratio of the reactive power output of converter 1 and converter 2 is close to 1.25, which is the same as its rated reactive power ratio of 1.25, and the reactive power distribution error Q error The value of is close to 0, which realizes the accurate distribution of output reactive power according to the rated capacity, and the optimization distribution effect is not affected by communication delay.

Claims

1. An adaptive control method for accurately distributing reactive power of multiple grid-connected converters, characterized in that: The adaptive control method involves a system consisting of X grid-type converters based on virtual synchronous generator technology, each converter is connected to a load in parallel, and any one of the grid-type converters is recorded as grid-type converter i, i∈[1,X]; the number of grid-type converters adjacent to the grid-type converter i is Y, and any one of them is recorded as an adjacent grid-type converter j, j∈[1,Y], Y<X; a distributed communication network of X grid-type converters is constructed in the system, specifically, the grid-type converter i can send information to the adjacent grid-type converter j, and receive information sent by the adjacent grid-type converter j; The adaptive control method comprises the following steps: Step 1: Sample the three-phase output voltage u at the output end of the grid-connected converter i ai ,u bi ,u ci And the three-phase output current i ai ,i bi ,i ci After coordinate transformation, the dq axis output voltage component U in the synchronous rotating coordinate system is obtained di ,U qi and dq axis output current component I di ,I qi Then, the output active power P of the grid-connected converter i is obtained by power calculation. ei And the output reactive power Q ei ; Step 2: According to the output active power P of the grid-connected converter i ei And the output reactive power Q ei Perform virtual synchronous generator control. Specifically, the output voltage angular frequency reference value ω is obtained through the active loop control equation i , and get the output voltage phase angle reference value θ i ,θ i =∫ω i ; Given reactive power-voltage droop coefficient k ni , the voltage reference value U is obtained through the reactive loop control equation refi ; Step 3: The grid-connected converter i obtains the output reactive power Q of the adjacent grid-connected converter j through communication. ej and reactive power-voltage droop coefficient k nj , combined with the output reactive power Q of the grid-connected converter i ei And reactive power-voltage droop coefficient k ni , the virtual capacitor droop coefficient compensation value Δn is obtained through the multi-agent consensus algorithm ci and virtual inductance compensation value ΔL vi ; Step 4: According to the voltage reference value U refi and virtual capacitor droop coefficient compensation value Δn ci , the d-axis virtual voltage standard value component U is calculated by the virtual capacitance algorithm rdi ; Set the q-axis virtual voltage standard value component U rqi Set to 0, and then according to the d-axis virtual voltage standard value component U rdi and virtual inductance compensation value ΔL vi , the dq axis voltage component command value U is calculated by the virtual inductance algorithm clrdi ,U clrqi ; Step 5: According to the dq axis voltage component command value U clrdi ,U clrqi and the output voltage phase angle reference value θ i , perform voltage and current double closed-loop control of the grid-type converter i. Specifically, the output I of the dq axis output voltage outer loop control is obtained by calculating the dq axis output voltage outer loop control equation. rdi ,I rqi , and then the output U of the dq axis output current inner loop control equation is obtained pdi ,U pqi , and then the three-phase modulation wave signal u is obtained through coordinate transformation pai ,u pbi ,u pci , and generate the switch tube driving signal of the grid-type converter i.

2. The adaptive control method for accurately distributing reactive power of multiple grid-connected converters according to claim 1 is characterized in that: The output active power P of the grid-connected converter i in step 1 is ei And output reactive power Q ei The calculation formulas are:

3. The adaptive control method for accurately distributing reactive power of multiple grid-connected converters according to claim 1 is characterized in that: The active loop control equation in step 2 is: Among them, ω 0i is the rated angular frequency of the system voltage, ω i ′ is the output voltage angular frequency reference value of the previous control cycle, k ωi is the primary frequency modulation coefficient, P refi is the active power command value, J i is the virtual inertia; The reactive loop control equation is: U refi =U 0i -k ni (Q ei -Q refi ) Among them, U 0i is the system rated output voltage amplitude, Q refi is the reactive power command value.

4. The adaptive control method for accurately distributing reactive power of multiple grid-connected converters according to claim 1 is characterized in that: The expression of the multi-agent consensus algorithm in step 3 is: Among them, C Qi is the control gain, a ij is the weight coefficient set according to whether the grid-type converter i can communicate with the adjacent grid-type converter j. If communication is possible, then a ij =1, if communication is unavailable then a ij =0, k pyi is the proportional coefficient in the multi-agent consensus algorithm, k iyi is the integral coefficient of the multi-agent consensus algorithm, and s is the Laplace operator.

5. The adaptive control method for accurately distributing reactive power of multiple grid-connected converters according to claim 1 is characterized in that: Step 4: The standard value U of the d-axis virtual voltage component rdi The calculation formula is: Among them, U maxi is the maximum voltage amplitude allowed by the system, Q ratei is the rated reactive power of grid-connected converter i, n rci is the fixed value of the virtual capacitor droop coefficient, K Qci is the virtual capacitor proportional gain. The dq axis voltage component command value U clrdi ,U clrqi The calculation formula is: Among them, ω 0i is the rated angular frequency of the system voltage, L rvi is the fixed value of virtual inductance, K Qli is the virtual inductor proportional gain.

6. The adaptive control method for accurately distributing reactive power of multiple grid-connected converters according to claim 1 is characterized in that: The outer loop control equation of the dq axis output voltage in step 5 is: In the formula, k spvi is the proportional coefficient of the output voltage outer loop proportional-integral controller, k sivi is the integral coefficient of the output voltage outer loop proportional-integral controller, and s is the Laplace operator; The inner loop control equation of the dq axis output current is: Among them, k spii is the proportional coefficient of the output current inner loop proportional-integral controller, k siii is the integral coefficient of the output current inner loop proportional-integral controller.

Images