Multi-inverter parallel system control method combining robust and adaptive control

By combining robust and adaptive control methods for multi-inverter parallel systems, the problem of insufficient robustness of traditional islanded microgrid PI controllers in the face of environmental changes and load imbalances is solved. This achieves reactive power equalization and improved system stability, ensuring power quality and load balance, and extending system life.

CN118944197APending Publication Date: 2024-11-12NORTHEAST DIANLI UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411008533.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Existing traditional islanded microgrid PI controllers are not robust enough when facing complex situations such as changes in the external environment, load changes, and grid imbalances. This results in voltage and current distortion, slow response speed, large steady-state error, and the inability to evenly distribute reactive power among multiple inverters, leading to overload or low efficiency in some power generation units.

Method used

A control method combining robust and adaptive control for a multi-inverter parallel system is adopted. By establishing an adaptive virtual impedance controller, combined with a robust H∞ controller and distributed adaptive virtual impedance design, reactive power is evenly distributed and system stability is improved.

Benefits of technology

It improves the robustness and stability of microgrid systems, enabling them to adapt to different operating conditions without frequent parameter adjustments, extending system lifespan, and achieving precise control of parameters such as frequency and voltage, ensuring power quality and load balance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118944197A_ABST
    Figure CN118944197A_ABST
Patent Text Reader

Abstract

The invention discloses a multi-inverter parallel system control method combining robustness and adaptive control, belongs to the technical field of power system island micro-grid control, and establishes an adaptive virtual impedance controller. The H-infinity controller is famous in excellent robust performance, and can effectively process the influence caused by factors such as uncertainty of a system model, external disturbance and parameter change. And parameters do not need to be frequently adjusted to adapt to different working conditions. When the MG runs in the island mode, certain disturbance inevitably exists in the system, and the control performance and stability of the system can be improved by minimizing the sensitivity of the system to various uncertainties. According to the method, the overall frequency response of the system can be considered during design, so that accurate control over important parameters such as frequency and voltage in the micro-grid system is better achieved. And the distributed adaptive virtual impedance is added, so that the load of the micro-grid can be better shared, and the service life of the MG can be prolonged.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention belongs to the technical field of island microgrid control of power systems, and in particular relates to a multi-inverter parallel system control method combining robust and adaptive control. Background Art

[0002] With the growth of global energy demand and the increasing concern for environmental sustainability, the role of renewable energy in power systems has become increasingly important. In this context, microgrids (MGs) have gradually become the focus of academic and industrial attention as an emerging energy supply solution. An island microgrid consists of distributed energy resources and energy storage facilities. Its unique feature is that it can operate independently when it loses connection with the main grid, or connect to the main grid when needed. The traditional power system architecture relies on centralized power generation and distributed power transmission networks. This structure faces challenges such as unstable energy supply, power load fluctuations, and environmental impacts. In contrast, an island microgrid provides users with more flexible and continuous power options by managing energy production and consumption more integratedly. One of its core advantages is that it can utilize local renewable energy resources, such as solar photovoltaic and wind power generation, to reduce dependence on traditional fossil energy, thereby reducing carbon emissions and environmental impacts.

[0003] The voltage and current dual closed loop in the traditional primary control of MG is based on the proportional integral (PI) controller. The PI controller is responsible for regulating and controlling the current or voltage output of distributed energy resources in the primary control loop of the microgrid. The PI controller ensures that the micro MG system can effectively manage and optimize power supply. Its high flexibility and robustness enable the microgrid to adapt to different operating modes and environmental conditions, thereby improving the reliability, availability and economy of the system. Through the design and implementation of effective PI controller proportional integral parameters, the microgrid can ensure the stability and sustainable development of the power system while providing clean energy.

[0004] The existing traditional primary control method of isolated island microgrid has the following problems:

[0005] (1) The existing traditional island microgrid PI controller cannot effectively respond to external environmental changes such as load changes, weather fluctuations, and grid imbalances. The output voltage and current will be distorted to a certain extent, affecting the power quality of the MG output;

[0006] (2) The performance of the existing traditional island microgrid PI controller is highly dependent on the accurate adjustment of parameters. If the parameters are not adjusted properly, it may lead to slow system response, large steady-state error or frequent oscillation, which is particularly important for applications such as microgrids that require high stability and flexibility;

[0007] (3) In the existing traditional island microgrid, when the load line impedance from the power generation unit to the common bus is inconsistent, the reactive power generated by multiple inverters cannot be evenly distributed, resulting in a certain power generation unit bearing too much or too little reactive power. This imbalance may cause some power generation units to be overloaded or inefficient.

[0008] Therefore, the prior art urgently needs a new technical solution to solve the above problems. Summary of the invention

[0009] The technical problem to be solved by the present invention is to provide a multi-inverter parallel system control method combining robust and adaptive control. In order to enable each power generation unit to share the reactive power evenly, an adaptive virtual impedance controller is established, which has stronger robustness than traditional control methods.

[0010] A multi-inverter parallel system control method combining robust and adaptive control includes the following steps:

[0011] Step 1: Establish the DG structure of the distributed generation unit in the meshed island microgrid, and mathematically model the topology of the inverter based on the DG structure and the filter reference value, combined with the circuit KCL and KVL equations; perform Park transformation on the model to obtain the dq value of the inverter output voltage and output current in the direct axis-orthogonal axis;

[0012] Step 2: To improve the power quality and minimize the disturbance fluctuation of the busbar, a robust H ∞ Controlled voltage outer loop H ∞ Controller modeling, and the current inner loop H ∞ Controller modeling;

[0013] Step 3: Perform distributed adaptive virtual impedance design to obtain virtual inductance, virtual resistance and voltage values.

[0014] The mathematical model of the inverter topology is:

[0015]

[0016] In the formula, I f , V f are the current and voltage values ​​output by the three-phase inverter respectively; R f ,L f ,C f They are the filter resistance, filter inductance, and filter capacitance corresponding to the filter; I o , V o They respectively represent the three-phase current and voltage connected to the busbar of the isolated microgrid.

[0017] The Park transformation formula of the model is:

[0018]

[0019] In the formula, I fd , I fq , V fd , V fq are the output current and voltage values ​​of the three-phase inverter on the dq axis respectively; I od , I oq , V od , V oq They represent the values ​​of the three-phase current and voltage on the dq axis respectively. In the present invention, ω represents the angular frequency of the inverter.

[0020] The voltage outer loop H ∞ The controller modeling method is:

[0021] The reference voltage is tracked as a constant, and the voltage d-axis and q-axis reference values ​​are processed with the components of the output voltage d-axis and q-axis to obtain the voltage error. The voltage error formula is shown in (4):

[0022]

[0023] In the formula, e vod , e voq is the reference value of the voltage difference on the dq axis. od,ref , V oq,ref is the bus voltage reference value on the dq axis;

[0024] The voltage is used as a reference value for tracking, and the derivative of the reference voltage is considered to be 0; differentiating (4) yields:

[0025]

[0026] In the formula, is the derivative of the voltage difference on the dq axis. are the derivatives of the bus voltage on the dq axis respectively;

[0027] The inverter system includes a droop controller and a voltage-current dual closed-loop controller; the droop controller uses a power measurement module to obtain the output active power and reactive power of the micro-inverter unit, and calculates the target values ​​of frequency and voltage according to a preset droop relationship curve; the target value is used to determine the output reference voltage, and the voltage-current dual closed-loop controller performs closed-loop tracking control on the reference voltage; the established expression is:

[0028]

[0029] In the formula, is the second-order differential of the bus voltage on the dq axis; is the current differential of the inverter output on the dq axis; is the differential of the bus current on the dq axis; ω is the angular frequency of the inverter; ΔC is the uncertain capacitance value existing in the system during operation;

[0030] By augmenting formula (6), we obtain the state space expression (7):

[0031]

[0032] Where, X v is the state space of the system; u v is the input of the controller; W v is the disturbance of the system.

[0033] And its specific variables are

[0034] According to formula (6), each system matrix of the state space in (7) can be obtained:

[0035]

[0036] In the formula, the parameter matrix ΔB v1 =UTG (8)

[0037] Where U and G are known constant matrices, where U∈R 4×2 , G∈R 2×2 ; T is an unknown matrix, and the elements in T satisfy T T T≤I; where I represents the identity matrix.

[0038] Build H ∞ The expected output matrix Z of the performance index adjustment in the controller v :

[0039] Z v =C v X v +D v u v +D v11 W v (9)

[0040] In the formula, D v11 =0;

[0041] Based on H ∞ The control gain K of the state feedback robust controller of the control system v for:

[0042] u v =K v Xv (10)

[0043] In the formula, K v =[k vI ,k v ],K v ∈R 4×2 , according to the u set in the state space v , get u v The specific mathematical formula is:

[0044]

[0045] In the formula, is the current differential of the inverter output on the dq axis; is the differential of the bus voltage on the dq axis;

[0046] Calculate the current reference value I of the voltage loop output fd,ref and I fq,ref , as the input of the current loop, the formula is:

[0047]

[0048] In the formula, k vI ∈R 2×2 ,k v ∈R 2×2 is the gain K v two parts.

[0049] The robust H ∞ The robustness requirement satisfied by the control is:

[0050] H ∞ The performance index output matrix Z of robust control v The disturbance signal W should be robust, Z v The performance index should be less than γ, expressed as:

[0051]

[0052] In the formula, is the system's H ∞ norm;

[0053] According to the inverter system performance design, the constraint condition of γ is γ>0. According to the Schur complement property, H is obtained by using formula (7) and formula (9): ∞ The optimization problem is to establish the linear matrix inequality LMIs as:

[0054]

[0055] Where N is a positive definite matrix and N∈R 4×4,Z∈R 2×4 , obtain feasible solutions N and Z through calculation; through H ∞ Gain K of robust control feedback control system v =ZN -1 .

[0056] The inner current loop H ∞ The controller modeling method is:

[0057] Formula (3) is used to calculate the robust H ∞ The control is mathematically modeled in the current loop. Set the state variable X of the system l It is expressed as follows:

[0058] X l =[x1 x2 x3 x4] T (15)

[0059] On the basis of formula (3), the fluctuation variation inside the inverter system and the external interference are added to obtain the following formula:

[0060]

[0061] Where ΔL and ΔR are the uncertain inductance and resistance values ​​in the circuit when there is external disturbance or the system itself has fluctuations; suppose the state space expression of the current loop system is as follows:

[0062]

[0063] In the formula, u l is the input of the controller; W l is the disturbance of the system;

[0064] The system state space is:

[0065] X l =[i fd -i fd,ref ,∫(i fd -i fd,ref )dt,i fq -i fq,ref ,∫(i fq -i fq,ref )dt] T (18)

[0066] The input of the system is set to u l =[V fd V fq ] T , W l =[W ld W lq ] T The specific expression is:

[0067]

[0068] The specific expression of the state space formula (17) is obtained as follows:

[0069]

[0070] According to formula (20), the system matrix of the state space is obtained as follows:

[0071]

[0072] Where, ΔB l1 is a parameter uncertainty matrix, ΔB l1 The perturbation matrix of the system can be written as:

[0073] ΔB l1 =JOY (21)

[0074] Where J and Y are known constant matrices, where J∈R 4×2 , Y∈R 2×2 O is an unknown matrix, and the elements in O satisfy O T O≤I;

[0075] Z l H ∞ The expected output matrix of performance index adjustment in the controller is:

[0076] Z l =C l X l +D l u l +D l11 W l (twenty two)

[0077] In the formula, D l11 =0;

[0078] Establish LMIs and solve for controller gain K l1 , the output is:

[0079]

[0080] The distributed adaptive virtual impedance design method is:

[0081] Each DG unit is set to communicate only with the DG connected to it. This sets the communication network structure of the MG. According to the communication structure, the adjacency matrix is ​​represented as A * ;

[0082]

[0083] In the formula, Indicates the communication link connection form between the i-th DG unit and the j-th DG; the communication matrix A * Transformed into Markov matrix A:

[0084]

[0085] In the formula, a ij It represents the connection form between the i-th DG unit and the j-th DG on the Markov communication link; the matrix A is a right random matrix, that is, it has only one maximum eigenvalue 1 and corresponds to a right eigenvector; the continuously running system conforms to the stationary distribution of the Markov chain;

[0086] Establish reactive power control equalization controller:

[0087] Q e =A×Q errors (26)

[0088] In the formula,

[0089] In the formula, Q i represents the reactive power of the ith DG unit; since the matrix A is used in the Markov steady-state distribution b j is a real number, that is, the elements in each column are the same; make the reactive error Q errors,ij After passing through the controller, the consensus Q e ;

[0090] Q e The virtual inductance value L is calculated in the distributed adaptive controller DAC v ; It is designed to:

[0091]

[0092] Where Kk is the adaptive controller parameter;

[0093] The virtual resistance is:

[0094]

[0095] The voltage drop generated by the virtual impedance on the dq axis is expressed as:

[0096]

[0097] Where V d,drop With V q,drop is the voltage drop value on the d-axis and q-axis.

[0098] Through the above design scheme, the present invention can bring the following beneficial effects: combining the multi-inverter parallel system control method with robust and adaptive control, an adaptive virtual impedance controller is established. ∞ The controller is known for its excellent robust performance and can effectively handle the impact of factors such as system model uncertainty, external disturbances and parameter changes. There is no need to adjust parameters frequently to adapt to different working conditions. When the system inevitably experiences certain disturbances when the MG is operating in island mode, it can improve the control performance and stability of the system by minimizing the sensitivity of the system to various uncertainties. It can consider the overall frequency response of the system during design, so as to better achieve precise control of important parameters such as frequency and voltage in the microgrid system. The addition of distributed adaptive virtual impedance allows the microgrid to better balance the load and extend the life of the MG. BRIEF DESCRIPTION OF THE DRAWINGS

[0099] The present invention is further described below with reference to the accompanying drawings and specific embodiments:

[0100] Figure 1 It is a schematic diagram of the topological structure of the three-phase inverter of the present invention;

[0101] Figure 2 The present invention is based on H ∞ Schematic diagram of voltage and current dual-loop control structure;

[0102] Figure 3 is the H of the voltage loop of the present invention ∞ Schematic diagram of the control structure.

[0103] Figure 4 is the H of the current loop of the present invention ∞ Schematic diagram of the control structure.

[0104] Figure 5 This is a schematic diagram of the Mesh Network structure of the present invention.

[0105] Figure 6 The specific implementation method of the present invention adds disturbance based on H ∞ Schematic diagram of the system output voltage of the controller.

[0106] Figure 7 It is a schematic diagram of the system output voltage based on the PI controller after adding disturbance according to a specific embodiment of the present invention.

[0107] Figure 8 When adding variable load to the specific implementation mode of the present invention, based on H ∞ Schematic diagram of the system output voltage of the controller.

[0108] Fig. 9 When adding variable load to the specific implementation mode of the present invention, based on H∞ Schematic diagram of the system output current of the controller.

[0109] Fig.10 The specific embodiment of the present invention is based on H when disconnecting the variable load. ∞ Schematic diagram of the system output current of the controller.

[0110] Fig.11 It is a schematic diagram of reactive power based on distributed adaptive virtual impedance multi-inverters in parallel according to a specific implementation mode of the present invention.

[0111] Fig.12 It is a schematic diagram of reactive power of conventional primary control of multiple inverters in parallel according to a specific implementation mode of the present invention.

[0112] Fig.13 The present invention is a flow chart of a control method for a multi-inverter parallel system combining robust and adaptive control. DETAILED DESCRIPTION

[0113] The present invention combines robust and adaptive control with a multi-inverter parallel system control method, such as Fig.13 and Figure 1-Figure 5 As shown, it includes: the establishment of the inverter mathematical model, the voltage outer loop H ∞ Controller design, current inner loop H ∞ Controller design, distributed adaptive virtual impedance design, the specific steps are:

[0114] 1) Establishment of inverter mathematical model

[0115] according to Figure 1 The structure of the distributed generation unit (DG) in the isolated microgrid is established. According to its structure and the reference value of the filter, the topology of the inverter is mathematically modeled by combining the KCL and KVL equations in the circuit knowledge. The equation (1) is obtained as follows:

[0116]

[0117] Among them I f , V f are the current and voltage values ​​output by the three-phase inverter respectively; R f ,L f ,C f They are the filter resistance, filter inductance, and filter capacitance corresponding to the filter; I o , V oRespectively represent the three-phase current and voltage connected to the busbar of the isolated microgrid. By performing Park transformation on the above formula (1), we can obtain the mathematical formulas (2) and (3) of the inverter output voltage and output current on the direct-quadrature axis (dq).

[0118]

[0119] Among them, I fd , I fq , V fd , V fq are the output current and voltage values ​​of the three-phase inverter on the dq axis respectively; I od , I oq , V od , V oq They represent the values ​​of the three-phase current and voltage on the dq axis respectively. In the present invention, ω represents the angular frequency of the system.

[0120] 2) Voltage outer loop H ∞ Controller Design

[0121] When the microgrid is operating in island mode, the system will inevitably have certain disturbances. In order to ensure that the entire microgrid system has a certain robustness when the inverter is disturbed by the outside world, to improve the power quality and minimize the fluctuation of the busbar, a robust H-based ∞ Controlled voltage loop design: In the voltage loop, formula (2) is mainly used for mathematical modeling. Figure 2 Based on H ∞ Voltage and current dual-loop control structure diagram.

[0122] The reference voltage is regarded as a constant and tracked. The voltage error can be obtained by performing difference processing on the reference values ​​of the voltage d-axis and q-axis and the components of the output voltage d-axis and q-axis. The voltage error formula is shown in (4).

[0123]

[0124] where e vod , e voq is the reference value of the voltage difference on the dq axis. od,ref , V oq,ref is the bus voltage reference value on the dq axis. Since the voltage is used as a reference value for tracking, the derivative of the reference voltage can be regarded as 0. Differentiating (4) yields:

[0125]

[0126] in is the derivative of the voltage difference on the dq axis. They are the derivatives of the bus voltage on the dq axis respectively.

[0127] The droop controller and voltage-current dual closed-loop controller in the inverter system are its core components. The droop controller uses the power measurement module to obtain the output active power and reactive power of the micro-inverter unit, and calculates the target values ​​of frequency and voltage according to the preset droop relationship curve. These target values ​​are further used to determine the output reference voltage, and then the voltage-current dual closed-loop controller performs closed-loop tracking control on the reference voltage. The following expression is established:

[0128]

[0129] in is the second-order differential of the bus voltage on the dq axis; is the current differential of the inverter output on the dq axis; is the differential of the bus current on the dq axis; ω is the angular frequency of the inverter; ΔC is the uncertain capacitance value that exists in the system during operation.

[0130] By augmenting formula (6), we can obtain the state space expression (7).

[0131]

[0132] Where X v is the state space of the system; u v is the input of the controller; W v is the disturbance of the system. And its specific variables are According to formula (6), each system matrix of the state space in (7) can be obtained as follows.

[0133]

[0134] In the above matrix, the uncertainty parameter matrix ΔB v1 It can be written as follows;

[0135] ΔB v1 =UTG (8)

[0136] Where U and G are known constant matrices, where U∈R 4×2 , G∈R 2×2 . T is an unknown matrix, and the elements in T satisfy T T T≤I. Where I represents the identity matrix.

[0137] Z v As H ∞ The expected output matrix of the performance index adjustment in the controller is defined by the mathematical formula as follows:

[0138] Z v =C v X v +D v u v +D v11 W v (9)

[0139] According to the different parameters of each inverter, the performance index output matrix Z we designed is v The parameters of the performance indicators adopted are as follows:

[0140] D v11 =0

[0141] Based on H ∞ The control gain K of the state feedback robust controller of the control system v as follows:

[0142] u v =K v X v (10)

[0143] Where K v =[k vI ,k v ],K v ∈R 4×2 According to the u set in the state space v , we can get u v The specific mathematical formula is as follows:

[0144]

[0145] in is the current differential of the inverter output on the dq axis; It is the differential of bus voltage on dq axis.

[0146] The current reference value I of the voltage loop output can be obtained fd,ref and I fq,ref , as the input of the current loop, the specific formula is as follows:

[0147]

[0148] where k vI ∈R 2×2 ,k v ∈R 2×2 is the gain K v two parts.

[0149] This voltage loop uses H ∞Robust control needs to meet the requirements of robust control, which are mainly as follows:

[0150] H ∞ The performance index output matrix Z of robust control v The disturbance signal W should be robust, Z v The performance index should be less than γ, and its specific expression formula is as follows:

[0151]

[0152] in is the system's H ∞ Norm.

[0153] Considering the performance of the inverter system, the constraint condition of γ is γ>0. According to the Schur complement property, the mathematical model formula (7) and the performance index output matrix Z v Formula (9) can be used as a model to obtain H ∞ The optimization problem is established. The linear matrix inequalities (LMIs) are expressed as follows:

[0154]

[0155] Where N is a positive definite matrix and N∈R 4×4 ,Z∈R 2×4 , after calculation, feasible solutions N and Z can be obtained. Therefore, H ∞ Gain K of robust control feedback control system v =ZN -1 Its control structure is as follows: Figure 3 shown.

[0156] 3) Current inner loop H ∞ Controller Design

[0157] Using formula (3) to calculate the robust H ∞ The control is mathematically modeled in the current loop. Set the state variable X of the system l It is expressed as follows:

[0158] X l =[x1 x2 x3 x4] T (15)

[0159] On the basis of formula (3), the fluctuation variation inside the inverter system and the external interference are added to obtain the following formula:

[0160]

[0161] Among them, ΔL and ΔR are the uncertain inductance and resistance values ​​of the circuit when there is external disturbance in the system or the system itself has fluctuations. Assume that the state space expression of the current loop system is as follows:

[0162]

[0163] where u l is the input of the controller; W l is the disturbance of the system. The system state space is specifically expressed as:

[0164] X l =[i fd -i fd,ref ,∫(i fd -i fd,ref )dt,i fq -i fq,ref ,∫(i fq -i fq,ref )dt] T (18)

[0165] The input of the system is set to u l =[V fd V fq ] T , W l =[W ld W lq ] T The specific expressions are as follows:

[0166]

[0167] The specific expression of the system state space (17) can be obtained as follows:

[0168]

[0169] According to formula (20), the system matrix of the state space can be obtained as follows:

[0170]

[0171] ΔB l1 is a matrix with uncertain parameters, and ΔB l1 The perturbation matrix of the system can be written as follows:

[0172] ΔB l1 =JOY (21)

[0173] Where J and Y are known constant matrices, where J∈R 4×2 , Y∈R 2×2 O is an unknown matrix, and the elements in O satisfy O TO≤I.

[0174] Z l As H ∞ The expected output matrix of the performance index adjustment in the controller is defined by the mathematical formula as follows:

[0175] Z l =C l X l +D l u l +D l11 W l (twenty two)

[0176] in D l11 =0.

[0177] Establish LMIs in the same way as the voltage design method and solve the controller gain K l1 . The following output is obtained:

[0178]

[0179] Its control structure is as follows Figure 4 shown.

[0180] 4) Distributed adaptive virtual impedance design

[0181] The MG structure involved in the present invention is a mesh structure (MeshNetwork), which is significantly different from the general single bus structure (Single Bus). The mesh structure can better reflect the common structure of the current microgrid, which contains multiple buses. Figure 5 shown.

[0182] Each DG unit is set to communicate only with the DG connected to it. This sets the communication network structure of the MG. According to the communication structure, the adjacency matrix is ​​represented as A * .

[0183]

[0184] in Represents the communication link connection form between the i-th DG unit and the j-th DG. * Transformed into Markov matrix A, the form is as follows:

[0185]

[0186] Among them, a ijIt represents the connection form between the i-th DG unit and the j-th DG on the Markov communication link. The above matrix belongs to a right random matrix, that is, it has only one largest eigenvalue 1 and corresponds to a right eigenvector. According to the Perron-Frobenius theorem, it can be known that the system is a stationary distribution of the Markov chain when it is continuously running.

[0187] The proposed reactive power control equalization controller is expressed as:

[0188] Q e =A×Q errors (26)

[0189] in:

[0190]

[0191] Among them, Q i represents the reactive power of the ith DG unit. Since the matrix A can be used in the Markov steady-state distribution b j is a real number. That is, the elements in each column are the same. So that the reactive error Q errors,ij After passing through the controller, we can get the consensus Q e .

[0192] The obtained Q e The virtual inductance value L is calculated in the distributed adaptive controller (DAC) v . It is designed to:

[0193]

[0194] Where Kk is the adaptive controller parameter. The virtual resistance is:

[0195]

[0196] The voltage drop generated by the virtual impedance on the dq axis is expressed as follows:

[0197]

[0198] Where V d,drop With V q,drop is the voltage drop value on the dq axis. Specific embodiment:

[0200] By changing the output impedance near the nominal value as a disturbance of small signal, corresponding to R f ,L f ,C fThe parameters fluctuate by 50%. Among them, the DC source voltage is 700V; the frequency is 50Hz; the active power droop coefficient is 5e-5; the reactive power droop coefficient is 3e-3; R f =0.4Ω; L f =1.4e-4mH; C f =1e-4μF. Kk is 1e-5; the impedances of the three lines are 0.05+j0.1Ω, 0.1+j0.21Ω, and 0.06+j0.13Ω respectively.

[0201] Firstly, the operation of the proposed controller is compared with that of the traditional PI controller without parameter disturbance. Figure 6 and Figure 7 shown.

[0202] When two controllers control the voltage outer loop and current inner loop, the output waveform of the inverter is greatly different. Figure 7 The output three-phase voltage is very unstable, and its peak value greatly exceeds the specified 311.127V. Figure 6 The H ∞ The output effect of the inverter under robust control does not change due to the addition of external interference, and the waveform of the system output can still stably reach the reference value. Since robust control can adjust the system interference, this can be referred to the H ∞ By establishing a mathematical model of the inverter during control, it can be found that external interference exists in its state space expression, and the interference is controlled to minimize the infinite norm of the transfer function from the interference to the desired output.

[0203] The variable load is set to be connected to the circuit at 1s and disconnected at 2s. The controller response proposed by the present invention is as follows Figure 8 , Fig. 9 , Fig.10 As shown. It can be found that the present invention can fully adapt to load measurement changes. When the load is connected, the voltage will not fluctuate. The current will increase with the increase of the load. At 2s, as the load is cut off, the current value decreases accordingly.

[0204] Fig.11 and Fig.12 The comparison results of reactive power sharing between the control proposed in the present invention and the traditional primary control are shown respectively. It can be obviously found that the control strategy proposed in the present invention can adapt to the inconsistent line impedance from the DG unit to the bus, thereby achieving reactive power sharing, so that each power generation unit can operate stably while extending its service life.

[0205] The specific implementation methods of the present invention are not exhaustive and do not constitute a limitation on the scope of protection of the claims. Based on the inspiration gained from the embodiments of the present invention, those skilled in the art can think of other substantially equivalent alternatives without creative work, all of which are within the scope of protection of the present invention.

Claims

1. A multi-inverter parallel system control method combining robust and adaptive control, characterized by: The following steps are included: Step 1: Establish the DG structure of the distributed generation unit in the meshed island microgrid, and mathematically model the topology of the inverter based on the DG structure and the filter reference value, combined with the circuit KCL and KVL equations; perform Park transformation on the model to obtain the dq value of the inverter output voltage and output current in the direct axis-orthogonal axis; Step 2: To improve the power quality and minimize the disturbance fluctuation of the busbar, a robust H ∞ Controlled voltage outer loop H ∞ Controller modeling, and the current inner loop H ∞ Controller modeling; Step 3: Perform distributed adaptive virtual impedance design to obtain virtual inductance, virtual resistance and voltage values.

2. The multi-inverter parallel system control method combining robust and adaptive control according to claim 1 is characterized by: The mathematical model of the inverter topology is: In the formula, I f , V f are the current and voltage values ​​output by the three-phase inverter respectively; R f , L f , C f They are the filter resistance, filter inductance, and filter capacitance corresponding to the filter; I o , V o They respectively represent the three-phase current and voltage connected to the busbar of the isolated microgrid. The Park transformation formula of the model is: In the formula, I fd , I fq , V fd , V fq are the output current and voltage values ​​of the three-phase inverter on the dq axis respectively; I od , I oq , V od , V oq They respectively represent the values ​​of the three-phase current and voltage on the dq axis that are connected to the busbar of the isolated microgrid; in the present invention, ω represents the angular frequency of the inverter.

3. The multi-inverter parallel system control method combining robust and adaptive control according to claim 1, characterized in that: The voltage outer loop H ∞ The controller modeling method is: The reference voltage is tracked as a constant, and the voltage d-axis and q-axis reference values ​​are processed with the components of the output voltage d-axis and q-axis to obtain the voltage error. The voltage error formula is shown in (4): In the formula, e vod , e voq is the reference value of the voltage difference on the dq axis. od,ref , V oq,ref is the bus voltage reference value on the dq axis; The voltage is used as a reference value for tracking, and the derivative of the reference voltage is considered to be 0; differentiating (4) yields: In the formula, is the derivative of the voltage difference on the dq axis. are the derivatives of the bus voltage on the dq axis respectively; The inverter system includes a droop controller and a voltage-current dual closed-loop controller; the droop controller uses a power measurement module to obtain the output active power and reactive power of the micro-inverter unit, and calculates the target values ​​of frequency and voltage according to a preset droop relationship curve; the target value is used to determine the output reference voltage, and the voltage-current dual closed-loop controller performs closed-loop tracking control on the reference voltage; the established expression is: In the formula, is the second-order differential of the bus voltage on the dq axis; is the current differential of the inverter output on the dq axis; is the differential of the bus current on the dq axis; ω is the angular frequency of the inverter; ΔC is the uncertain capacitance value existing in the system during operation; By augmenting formula (6), we obtain the state space expression (7): Where, X v is the state space of the system; u v is the input of the controller; W v is the disturbance of the system. And its specific variables are According to formula (6), each system matrix of the state space in (7) can be obtained: In the formula, the parameter matrix ΔB v1 =UTG (8) Where U and G are known constant matrices, where U∈R 4×2 , G∈R 2×2 ; T is an unknown matrix, and the elements in T satisfy T T T≤I, where I represents the identity matrix; Build H ∞ The expected output matrix Z of the performance index adjustment in the controller v : Z v =C v X v +D v u v +D v11 W v (9) In the formula, Based on H ∞ The control gain K of the state feedback robust controller of the control system v for: u v =K v X v (10) In the formula, K v =[k vI ,k v ],K v ∈R 4×2 , according to the u set in the state space v , get u v The specific mathematical formula is: In the formula, is the current differential of the inverter output on the dq axis; is the differential of the bus voltage on the dq axis; Calculate the current reference value I of the voltage loop output fd,ref and I fq,ref , as the input of the current loop, the formula is: In the formula, k vI ∈R 2×2 ,k v ∈R 2×2 is the gain K v two parts.

4. The control method for a multi-inverter parallel system combining robust and adaptive control according to claim 1, characterized in that: The robust H ∞ The robustness requirement satisfied by the control is: H ∞ The performance index output matrix Z of robust control v The disturbance signal W should be robust, Z v The performance index should be less than γ, expressed as: In the formula, is the system's H ∞ norm; According to the inverter system performance design, the constraint condition of γ is γ>0. According to the Schur complement property, H is obtained by using formula (7) and formula (9): ∞ The optimization problem is to establish the linear matrix inequality LMIs as: Where N is a positive definite matrix and N∈R 4×4 ,Z∈R 2×4 , obtain feasible solutions N and Z through calculation; through H ∞ Gain K of robust control feedback control system v =ZN -1 .

5. The control method for a multi-inverter parallel system combining robust and adaptive control according to claim 1, characterized in that: The inner current loop H ∞ The controller modeling method is: Formula (3) is used to calculate the robust H ∞ The control is mathematically modeled in the current loop. Set the state variable X of the system l It is expressed as follows: X l =[x1 x2 x3 x4] T (15) On the basis of formula (3), the fluctuation variation inside the inverter system and the external interference are added to obtain the following formula: Where ΔL and ΔR are the uncertain inductance and resistance values ​​in the circuit when there is external disturbance or the system itself has fluctuations; suppose the state space expression of the current loop system is as follows: In the formula, u l is the input of the controller; W l is the disturbance of the system; The system state space is: X l =[i fd -i fd,ref ,∫(i fd -i fd,ref )dt,i fq -i fq,ref ,∫(i fq -i fq,ref )dt] T (18) The input of the system is set to u l =[V fd V fq ] T , W l =[W ld W lq ] T The specific expression is: The specific expression of the state space formula (17) is obtained as follows: According to formula (20), the system matrix of the state space is obtained as follows: Where, ΔB l1 is a parameter uncertainty matrix, ΔB l1 The perturbation matrix of the system can be written as: ΔB l1 =JOY (21) Where J and Y are known constant matrices, where J∈R 4×2 , Y∈R 2×2 O is an unknown matrix, and the elements in O satisfy O T O≤I; Z l H ∞ The expected output matrix of performance index adjustment in the controller is: Z l =C l X l +D l u l +D l11 W l (22) In the formula, Establish LMIs and solve for controller gain K l1 , the output is:

6. The control method for a multi-inverter parallel system combining robust and adaptive control according to claim 1, characterized in that: The distributed adaptive virtual impedance design method is: Each DG unit is set to communicate only with the DG connected to it. This sets the communication network structure of the MG. According to the communication structure, the adjacency matrix is ​​represented as A * ; In the formula, Indicates the communication link connection form between the i-th DG unit and the j-th DG; the communication matrix A * Transformed into Markov matrix A: In the formula, a ij It represents the connection form between the i-th DG unit and the j-th DG on the Markov communication link; the matrix A is a right random matrix, that is, it has only one maximum eigenvalue 1 and corresponds to a right eigenvector; the continuously running system conforms to the stationary distribution of the Markov chain; Establish reactive power control equalization controller: Q e =A×Q errors (26) In the formula, In the formula, Q i represents the reactive power of the ith DG unit; since the matrix A is used in the Markov steady-state distribution b j is a real number, that is, the elements in each column are the same; make the reactive error Q errors,ij After passing through the controller, the consensus Q e ; Q e The virtual inductance value L is calculated in the distributed adaptive controller DAC v ; It is designed to: Where Kk is the adaptive controller parameter; The virtual resistance is: The voltage drop generated by the virtual impedance on the dq axis is expressed as: Where V d,drop With V q,drop is the voltage drop value on the d-axis and q-axis.