Phase advance compensation iterative learning control method and device for dual-mode flyback inverter

Abstract

The present invention discloses a dual-mode flyback inverter phase advance compensation iterative control method and device, the method comprising: based on grid-connected current feedback control, the flyback inverter is operated in DCM and CCM dual modes through segmented feedforward compensation; the switching rules of DCM and CCM dual modes are formulated, so that the DCM and CCM modes appear alternately within half a power frequency cycle; a dynamic mathematical model of the grid-connected process of the dual-mode flyback inverter is established, and the dynamic mathematical model is discretized to obtain a series of discrete time iterative linear systems; a phase advance compensation iterative learning controller is constructed according to the discrete time iterative linear system; the phase advance compensation iterative learning controller parameters are designed according to the convergence conditions that meet the stability of the closed-loop system, and the optimization control of the dual-mode flyback inverter is realized according to the grid-connected current feedback control process. The present invention ensures global convergence in CCM and DCM modes and avoids the RHP zero point compensation problem.

Description

Technical Field

[0001] The present invention belongs to the technical field of micro inverter control, and in particular relates to a dual-mode flyback inverter phase advance compensation iterative learning control method and device. Background Art

[0002] In distributed energy systems, small and medium-sized energy systems such as photovoltaic panels or thermoelectric power modules generate electricity from primary energy sources. These systems are combined with grid-connected inverters to transmit the generated electricity to the utility grid. In recent years, the use of module integrated converters (MICs) at the inverter stage has become increasingly popular because it allows each module such as photovoltaic panels to work independently and reduces power losses caused by mismatches between modules. In addition, it is easy to maintain, avoids system failures due to single-point failures, and provides "plug and play" functionality. In single-phase MICs, flyback microinverters have the characteristics of small size, high integration, independent tracking of maximum power, and convenient installation. They are widely used in scenarios such as photovoltaic curtain walls and home roofs. In order to increase power density, flyback microinverters usually adopt an interleaved flyback structure.

[0003] Traditional interleaved flyback inverters usually use peak current control to indirectly modulate the grid-connected current by controlling the primary side peak current or on-time. The inverter usually works in a mixed mode of DCM and BCM. Although it is simple and efficient, the reference current or on-time formula involved is not accurate, which increases the total harmonic distortion (THD). In addition, in order to solve the RHP zero point problem of the flyback inverter in CCM mode, a closed-loop control scheme with output current feedback is usually adopted. Traditional PI and PR controllers all need to design compensation networks to suppress the inherent RHP zero point of the system to limit the system gain, and the control performance is largely limited by the compensation network. Summary of the invention

[0004] The purpose of the present invention is to provide a dual-mode flyback inverter phase advance compensation iterative learning control method and device. In order to solve the above problems, the present invention, based on traditional grid-connected current feedback control, uses segmented feedforward compensation to make the inverter work in a mixed mode of DCM and CCM, and constructs a dynamic mathematical model according to different switching modes. With the help of the dynamic mathematical model, a phase advance compensation iterative learning controller is constructed. By verifying the asymptotic stability of the closed-loop system, the global convergence of the system in different working modes is guaranteed.

[0005] The present invention is implemented according to the following technical solutions:

[0006] The present invention provides a dual-mode flyback inverter phase advance compensation iterative learning control method, the method comprising:

[0007] Based on grid-connected current feedback control, the flyback inverter is made to work in DCM and CCM dual modes through segmented feedforward compensation; the DCM mode means that the inductor current flowing through the flyback transformer is in a discontinuous state during the entire switching cycle, and the CCM mode means that the inductor current flowing through the flyback transformer is in a continuous state during the entire switching cycle;

[0008] Formulate the switching rules between DCM and CCM modes so that DCM and CCM modes appear alternately within half of the power frequency cycle;

[0009] The dynamic mathematical model of the grid-connected process of the dual-mode flyback inverter is established, and the dynamic mathematical model is discretized to obtain a series of discrete-time iterative linear systems;

[0010] According to the discrete time iterative linear system, a phase advance compensation iterative learning controller is constructed;

[0011] The parameters of the phase advance compensation iterative learning controller are designed according to the convergence conditions that meet the stability of the closed-loop system, and the optimal control of the dual-mode flyback inverter is achieved according to the grid-connected current feedback control process.

[0012] In one implementation, in order to facilitate the control of the output current in each mode, the nominal duty cycles in the two different operating modes of DCM and CCM are used as feedforward inputs. The nominal duty cycles of the DCM and CCM modes are:

[0013]

[0014] Where V pv is the photovoltaic input voltage, P o is the output power of the dual-mode flyback inverter, L m is the flyback transformer magnetizing inductance, f s is the switching frequency, f g is the grid frequency, V g is the effective value of the grid voltage, and N is the transformer ratio.

[0015] The switching rules between DCM and CCM are as follows: DCM <D CCM The flyback inverter operates in DCM mode. DCM >D CCM When , the flyback inverter operates in CCM mode.

[0016] In one implementation, the specific steps of establishing a dynamic mathematical model of the dual-mode flyback inverter grid-connected process include:

[0017] According to the simplified equivalent circuit of a flyback converter, considering the whole switching cycle, the state space equation in the whole switching cycle can be obtained by using Kirchhoff's voltage and current laws.

[0018] Through the state space equation, the small signal model of DCM and CCM mode is obtained:

[0019] According to the small signal model of DCM and CCM modes, and the flyback inverter works alternately in DCM and CCM modes in one grid cycle, its dynamic mathematical model is expressed as a linear switching system:

[0020]

[0021] y(t)=Cx(t)

[0022] Since the dual-mode flyback inverter is based on the grid cycle T g The same task is repeated to describe the dynamic mathematical model as a series of iterative linear switching systems:

[0023]

[0024] y i (t) = Cx i (t)

[0025] Where i represents the number of iterations, x i (t), y i (t), d i (t) are the state variables, output, and input at the i-th iteration respectively.

[0026] In one embodiment, the discrete-time iterative linear system is expressed as:

[0027] x i (k+1)=Fx i (k)+Gu c (k)

[0028] y i (k) = Hx i (k)

[0029] Among them, x i (k)∈R 4 ×1,y i (k)∈R,u c (k)∈R are the state vector, output and control input when i=1,2…, respectively, F∈R 4 ×4,G∈R 4 ×1,H∈R 1 ×4 is the discrete time system matrix.

[0030] In one embodiment, constructing a phase advance compensation iterative learning controller according to a discrete time iterative linear system comprises the following steps:

[0031] (1) The construction of the phase advance compensation iterative learning controller must meet two assumptions:

[0032] Assumption 1: The resulting discrete-time iterative linear system is reversible, that is, given the desired output trajectory y d (k) There exists a unique bounded input u d (k) such that:

[0033] x d (k+1)=Fx d (k)+Gu d (k)

[0034] y d (k) = Hx d (k),k∈[0,k T ],

[0035] where x d (k)∈R 4 ×1 is the desired state vector;

[0036] Assumption 2: All iterations satisfy the initial condition: x i (0) = x d (0), y i (0) = y d (0), i=1, 2…;

[0037] (2) Based on the two assumptions, a phase advance compensation iterative learning control framework is constructed, and its control input can be expressed as:

[0038] u c (k) = u fb (k)+u ff (k)+u i (k)

[0039] Among them, u fb (k)∈R is the output value of the state feedback controller, u ff (k)∈R is the nominal duty cycle in two different switching modes, u i (k)∈R is the output value of the iterative learning controller;

[0040] (3) According to the switching rules of the DCM and CCM dual modes, the constraints of the nominal duty cycle are set, and the state feedback controller adopts the form of proportional control, and the iterative learning controller adopts the form of P-type iterative learning control law to set the control satisfaction conditions of the phase lead compensation iterative learning controller.

[0041] In one implementation, the control of the phase advance compensation iterative learning controller satisfies the condition, which is specifically expressed as:

[0042] u fb (k) = k p e(k)

[0043]

[0044] where k p is the proportional gain; λ d ,λ c ≥1 is the phase advance step length; e i (k) = y d -y i (k) is the output error; k l is the learning controller gain; i = 0, 1… is the iteration index; k = 0,…, k T is a discrete-time index.

[0045] In one embodiment, the convergence condition satisfying the stability of the closed-loop system is obtained by:

[0046] Consider the discrete-time iterative linear system x i (k+1)=Fx i (k)+Gu c (k) and y i (k) = Hx i (k), if the two assumptions for constructing the phase lead compensation iterative learning controller hold and the inequality ||E i+1 ||≤||Ik l P||||E i || is established, which can make the output tracking error of the closed-loop system of the phase advance compensation iterative learning controller converge monotonically to 0;

[0047] Among them, E in the inequality i+1 is the error matrix of the i+1th iteration, I is k N ×k N The k-th order identity matrix, l To learn the controller gain, It can be expressed as:

[0048]

[0049] F s and G s Satisfy the following formula:

[0050] F s =(Fk p GH) λ

[0051]

[0052] Where F, G, H are discrete time system matrices, λ is the iterative learning order, j = 0, 1, …λ-1.

[0053] In one embodiment, after constructing the phase advance compensation iterative learning controller according to the discrete time iterative linear system, the method further includes verifying the closed-loop stability of the established phase advance iterative learning controller, and the specific verification steps include:

[0054] (1) Applying the phase advance compensation iterative learning controller to the discrete time iterative linear system, the relationship between the control input at time k+1 in the i-th iteration and the control output at time k in the i-th iteration is obtained;

[0055] (2) In the i-th iteration, k = 0, 1…λ-1 is brought into the discrete-time iterative linear system to which the phase advance compensation iterative learning controller is applied, and the relationship between the control input at each moment in the i-th iteration process and the control output at the previous moment can be obtained;

[0056] (3) From the control conditions of the phase advance compensation iterative learning controller, it can be known that the control input u i (k) is the same for all k = 0, 1…λ-1, so the control input at time λ in the i-th iteration can be derived from the control output at the initial time;

[0057] Applying the above steps (1) to (3) to each sampling, a discrete-time iterative linear system after applying a phase-advance compensation iterative learning controller to any control input signal can be obtained;

[0058] Under the condition of satisfying Assumption 2, the newly obtained discrete-time iterative linear system is solved and the output error of the i+1th iteration is subtracted from the output error of the i-th iteration to obtain the condition for the output error to converge to 0, thereby completing the verification of the closed-loop stability of the phase-advance iterative learning controller.

[0059] The present invention also provides a dual-mode flyback inverter phase advance compensation iterative learning control device, the device comprising:

[0060] The dual-mode control module is used to make the flyback inverter work in DCM and CCM dual modes based on grid-connected current feedback control through segmented feedforward compensation; and formulate the switching rules between DCM and CCM dual modes so that the DCM and CCM modes appear alternately within half a power frequency cycle;

[0061] The discrete time iterative linear system module is used to establish a dynamic mathematical model of the grid-connected process of the dual-mode flyback inverter and discretize the dynamic mathematical model to obtain a series of discrete time iterative linear systems;

[0062] A phase advance compensation iterative learning controller building module is used to build a phase advance compensation iterative learning controller according to a discrete time iterative linear system;

[0063] The parameter design and output control module performs the phase advance compensation iterative learning controller parameter design according to the convergence conditions that meet the stability of the closed-loop system, and realizes the optimal control of the dual-mode flyback inverter according to the grid-connected current feedback control process.

[0064] In one embodiment, the phase advance compensation iterative learning controller includes: a state feedback controller, a feedforward input controller and an iterative learning controller; wherein,

[0065] The state feedback controller adopts a proportional control form to adjust the system gain of the dual-mode flyback inverter to speed up the response speed;

[0066] The feedforward input controller is used to calculate the nominal duty cycle under different switching modes and reduce the control gain of the state feedback controller;

[0067] The iterative learning controller adopts the form of a P-type iterative learning control law, which is used to eliminate periodic interference, improve tracking accuracy, and avoid control system instability caused by RHP zero point in CCM mode.

[0068] Beneficial effects:

[0069] Based on the traditional grid-connected current feedback control, the present invention enables the inverter to operate in a mixed mode of DCM and CCM through segmented feedforward compensation, without the need to calculate the reference current, while retaining the advantage of traditional peak current control that single-channel operation can be performed under light load, and there is no need to change the original switching frequency in the CCM mode. Compared with the variable switching frequency in the peak current control BCM mode, the switching loss is reduced; in addition, the ILC controller constructed with the help of dynamic mathematical models under different switching modes ensures the global convergence of the system in CCM and DCM modes, avoiding the RHP zero-point compensation problem. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] The accompanying drawings are part of the present invention and are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention, but do not constitute an improper limitation of the present invention. Obviously, the drawings described below are only some embodiments. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0071] Figure 1 A flow chart of an iterative learning control method for phase advance compensation of a dual-mode flyback inverter provided by an embodiment of the present invention;

[0072] Figure 2 A topological structure diagram of an interleaved flyback micro-inverter provided in an embodiment of the present invention;

[0073] Figure 3 It is a schematic diagram of the main operating mode waveforms and mode switching of the dual-mode flyback inverter DCM / CCM provided in an embodiment of the present invention;

[0074] Figure 4 A simplified equivalent circuit of a flyback inverter provided in an embodiment of the present invention;

[0075] Figure 5 It is a dual-mode flyback inverter phase advance compensation iterative learning control framework provided in an embodiment of the present invention;

[0076] Figure 6 The control scheme and peak current control waveform of the primary and secondary currents of the two-way flyback converter provided in the embodiment of the present invention are shown; Figure 6 (a) is the primary and secondary current waveforms of the two-way flyback converter using the method of the present invention, Figure 6 (b) The primary and secondary current waveforms of the two-way flyback converter in the peak current control scheme;

[0077] Figure 7 The control scheme and peak current control grid-connected voltage and grid-connected current waveforms provided in the embodiment of the present invention are shown in FIG. Figure 7 (a) is the grid-connected voltage and current waveform using the method of the present invention, Figure 7 (b) is the peak current control grid-connected voltage and current waveform diagram;

[0078] Figure 8 are the grid-connected voltage, grid-connected current and grid-connected current THD at full load provided in the embodiment of the present invention, wherein: Figure 8 (a) is the grid-connected voltage and current at full load using the method of the present invention, Figure 8 (b) is the grid-connected current THD at full load using the method of the present invention, Figure 8 (c) is the traditional current feedback using PI control when the grid-connected voltage and current are fully loaded. Figure 8 (d) The grid-connected current THD at full load using PI control with traditional current feedback;

[0079] Fig. 9 are the grid-connected voltage, grid-connected current and grid-connected current THD at medium load provided in the embodiment of the present invention, wherein Fig. 9 (a) is the load-time grid-connected voltage and current using the method of the present invention, Fig. 9(b) is the load-time grid-connected current THD using the method of the present invention, Fig. 9 (c) The traditional current feedback uses PI to control the grid-connected voltage and current at medium load. Fig. 9 (d) The conventional current feedback adopts PI control for medium load grid current THD;

[0080] Fig.10 Grid-connected voltage, grid-connected current and grid-connected current THD under light load provided in the embodiment of the present invention, wherein: Fig.10 (a) is the grid-connected voltage and current when light load is used using the method of the present invention, Fig.10 (b) is the grid-connected current THD when the method of the present invention is used at light load, Fig.10 (c) The traditional current feedback uses PI to control the grid-connected voltage and current when the load is light. Fig.10 (d) The traditional current feedback uses PI to control the grid-connected current THD at light load;

[0081] Fig.11 The pulse response time in the DCM and CCM states provided in the embodiment of the present invention;

[0082] Fig.12 A schematic structural diagram of a dual-mode flyback inverter phase lead compensation iterative learning control system provided by an embodiment of the present invention.

[0083] It should be noted that these drawings and textual descriptions are not intended to limit the conceptual scope of the present invention in any way, but are intended to illustrate the concept of the present invention for those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0084] In order to deepen the recognition and understanding of the present invention, the technical solution of the present invention is further introduced below in conjunction with the accompanying drawings and specific implementation methods. The following embodiments are explanations of the present invention and the present invention is not limited to the following embodiments.

[0085] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.

[0086] Reference Figure 1 As shown, the embodiment of the present invention provides a dual-mode flyback inverter phase advance compensation iterative learning control method, the method comprising the following steps:

[0087] Step S100: Based on grid-connected current feedback control, the flyback inverter is made to operate in DCM and CCM dual modes through segmented feedforward compensation. The DCM mode means that the inductor current flowing through the flyback transformer is in a discontinuous state during the entire switching cycle, and the CCM mode means that the inductor current flowing through the flyback transformer is in a continuous state during the entire switching cycle.

[0088] Interleaved Flyback Microinverter Figure 2 As shown in the figure, it consists of three parts: DC-DC circuit, DC-AC circuit and CL filter circuit. The DC-DC circuit adopts an interleaved flyback structure, including a voltage stabilizing capacitor C pv , main switch tubes S1, S2, auxiliary switch tube S a1 , S a2 , flyback transformer and other devices; the DC-AC circuit adopts a full-bridge structure composed of thyristors S3, S4, S5, and S6, usually called an industrial frequency expansion bridge; C f , L f It constitutes the filter circuit on the AC side.

[0089] Figure 3 (a) shows the main operating waveforms of the flyback inverter working in DCM / CCM mode. In the DCM state, the excitation current i m In the CCM state, the excitation current is continuous throughout the switching cycle. Figure 3 (b) shows the principle of DCM / CCM mode switching. The DCM and CCM intervals appear alternately within half the power frequency cycle. In order to conveniently control the output current in each mode, the nominal duty cycle is used as the feedforward input. The nominal duty cycles in DCM and CCM states are:

[0090]

[0091] Step S200: Formulate a switching rule between the DCM and CCM dual modes so that the DCM and CCM modes appear alternately within half a power frequency cycle.

[0092] Furthermore, the switching rule between DCM and CCM is as follows: DCM <D CCM The flyback inverter operates in DCM mode. DCM >D CCM When , the flyback inverter operates in CCM mode.

[0093] Step S300: establishing a dynamic mathematical model of the grid-connected process of the dual-mode flyback inverter, and discretizing the dynamic mathematical model to obtain a series of discrete-time iterative linear systems.

[0094] In the embodiment of the present application, the specific steps of establishing a dynamic mathematical model of the dual-mode flyback inverter grid-connected process include:

[0095] Since the two flyback converters are staggered in parallel and work 180° apart, only one of them can be analyzed. Figure 4 A simplified equivalent circuit of a flyback converter is given (the active clamping circuit is only for feeding back leakage inductance energy and is not considered here).

[0096] Considering the entire switching cycle according to the simplified equivalent circuit of a flyback converter, the state space equation within the entire switching cycle can be obtained using Kirchhoff's voltage and current laws.

[0097] Furthermore, the state space equation is:

[0098]

[0099] Where d1(t) is the duty cycle of the main switch when it is turned on. In CCM working mode, d2(t)=1-d1(t).

[0100] Through the above state space equations, the small signal models of DCM and CCM are obtained.

[0101] The small signal model in DCM operation mode is:

[0102] The small signal model in CCM operation mode is:

[0103] According to the small signal model of DCM and CCM modes, and in one grid cycle, the flyback inverter works alternately in DCM and CCM modes, and its model can be expressed as a linear switching system:

[0104]

[0105] Since the dual-mode flyback inverter is based on the grid cycle T g The same task is performed repeatedly, so it can be described as a series of iterative linear switching systems:

[0106]

[0107] Where i represents the number of iterations, x i (t), y i (t), d i (t) are the state variables, output, and input at the i-th iteration respectively.

[0108] Furthermore, the above iterative linear switching system is discretized into a discrete time iterative linear system:

[0109]

[0110] x i (k)∈R 4 ×1,y i (k)∈R,u c (k)∈R are the state vector, output and control input when i=1,2…, respectively, F∈R 4 ×4,G∈R 4 ×1,H∈R 1 ×4 is the discrete time system matrix.

[0111] Step 400: construct a phase advance compensation iterative learning controller according to a discrete time iterative linear system.

[0112] In this embodiment of the present application, a phase advance compensation iterative learning controller is constructed according to a discrete time iterative linear system, including:

[0113] (1) The construction of the phase advance compensation iterative learning controller must meet two assumptions:

[0114] Assumption 1: The system is reversible, that is, given the expected output trajectory y d (k) There exists a unique bounded input u d (k) such that:

[0115]

[0116] where x d (k)∈R 4 ×1 is the desired state vector.

[0117] Assumption 2: All iterations satisfy the following initial conditions:

[0118] x i (0) = x d (0), y i (0) = y d (0), i = 1, 2…

[0119] (2) Based on the two assumptions, a phase advance compensation iterative learning control framework is constructed. Figure 5 A phase advance compensation iterative learning control framework is proposed, and its control input can be expressed as:

[0120] u c (k) = u fb (k)+u ff (k)+u i (k) (9)

[0121] Among them, u fb(k)∈R is the state feedback controller, u ff (k)∈R is the Figure 3 (b) shows the nominal duty cycle of the two different switching modes, u i (k)∈R is an iterative learning controller.

[0122] (3) According to the switching rules of the DCM and CCM dual modes, the constraints of the nominal duty cycle are set, and the state feedback controller adopts the form of proportional control, and the iterative learning controller adopts the form of P-type iterative learning control law to set the control satisfaction conditions of the phase lead compensation iterative learning controller.

[0123] Furthermore, the control of the phase advance compensation iterative learning controller meets the conditions, which are specifically expressed as:

[0124] u fb (k) = k p e(k) (10)

[0125]

[0126] where k p is the proportional gain; λ d ,λ c ≥1 is the phase advance step length; e i (k) = y d -y i (k) is the output error; k l is the learning controller gain; i = 0, 1… is the iteration index; k = 0,…, k T is a discrete-time index.

[0127] Furthermore, the convergence condition for the stability of the closed-loop system is obtained in the following way:

[0128] Consider the discrete-time iterative linear system x i (k+1)=Fx i (k)+Gu c (k) and y i (k) = Hx i (k). If assumptions 1 and 2 hold, and the following inequalities hold:

[0129] ||E i+1 ||≤||Ik l P||||E i || (11)

[0130] If it holds, then it can be guaranteed that the output tracking error of the proposed ILC controller closed-loop system converges monotonically to 0.

[0131] Among them, E in the inequality i+1is the error matrix of the i+1th iteration, I is k N ×k N The k-th order identity matrix, l To learn the controller gain, It can be expressed as:

[0132]

[0133] F s and G s Satisfy the following formula:

[0134] F s =(Fk p GH) λ (13)

[0135]

[0136] In an optional embodiment, after constructing a phase advance compensation iterative learning controller according to a discrete time iterative linear system, the method further includes performing closed-loop stability verification on the established phase advance iterative learning controller, and the specific verification steps include:

[0137] (1) Applying the phase advance compensation iterative learning controller to the discrete time iterative linear system, the relationship between the control input at time k+1 in the i-th iteration and the control output at time k in the i-th iteration is obtained;

[0138] (2) In the i-th iteration, k = 0, 1…λ-1 is brought into the discrete-time iterative linear system to which the phase advance compensation iterative learning controller is applied, and the relationship between the control input at each moment in the i-th iteration process and the control output at the previous moment can be obtained;

[0139] (3) From the control conditions of the phase advance compensation iterative learning controller, it can be known that the control input u i (k) is the same for all k = 0, 1…λ-1, so the control input at time λ in the i-th iteration can be derived from the control output at the initial time;

[0140] Applying the above steps (1) to (3) to each sampling, a discrete-time iterative linear system after applying a phase-advance compensation iterative learning controller to any control input signal can be obtained;

[0141] Under the condition of satisfying Assumption 2, the newly obtained discrete-time iterative linear system is solved and the output error of the i+1th iteration is subtracted from the output error of the i-th iteration to obtain the condition for the output error to converge to 0, thereby completing the verification of the closed-loop stability of the phase-advance iterative learning controller.

[0142] Specifically, applying controller (9) to the discrete-time iterative linear system yields:

[0143]

[0144] Substituting k=0,1…λ-1 into equation (15), we can obtain:

[0145] x i (1) = (Fk p GH)x i (0)+Gu i (0)+G(k p y d (0)+u ff (0)) (16)

[0146]

[0147] From formula (10), we can know that the input u i (k) is the same for all k = 0, 1…λ-1, so equation (18) can be rewritten as:

[0148]

[0149] In formula (19),

[0150] Similarly, applying the above steps to each sampling, we can get the following linear system:

[0151] x i (n+1)=F s x i (n)+G s u i (n)+w(n) (20)

[0152] y i (n) = Hx i (n), n∈[0,k N ] (twenty one)

[0153] In formula (20), w(n) satisfies

[0154]

[0155] Note that w(n) is the sum of the desired output and the feedforward input, where n = 0, 1, ... k N -1. This shows that w(n) is invariant in iteration. Solving equations (20) and (21) yields:

[0156]

[0157] Define the output error ei (n) = y d (n)-y i (n), using the output error of the i+1th iteration minus the output error of the i-th iteration, we have

[0158]

[0159] Substituting assumption 2 and u in equation (10) i+1 Substitute the control rate of (k) into formula (24)

[0160]

[0161] Convert equation (25) into matrix form:

[0162] E i+1 =(Ik l P)E i (26)

[0163] Taking the norm on both sides of equation (26) yields: ||E i+1 ||≤||Ik l P||||E i ||.

[0164] Therefore, as long as equation (11) is satisfied, the output error of the closed-loop system can be guaranteed to converge monotonically to 0.

[0165] Step S500: Design the parameters of the phase advance compensation iterative learning controller according to the convergence condition that satisfies the stability of the closed-loop system, and implement the optimal control of the dual-mode flyback inverter according to the grid-connected current feedback control process.

[0166] In this technical solution, a dual-mode flyback inverter phase advance compensation iterative learning control method is provided. An iterative learning control (ILC) scheme with phase advance compensation is adopted. It uses the control input of the last iteration and the output error of the current iteration, which not only retains the information of the last iteration, but also adjusts according to the current operation result to ensure that the tracking error converges to zero. Compared with the traditional DCM / BCM dual-switch mode control, the ILC scheme uses closed-loop control with output current feedback to work in DCM / CCM mode, does not require reference current calculation, has lower grid-connected current THD, and has the same switching frequency as the DCM mode, while the switching loss of the BCM mode is higher when the frequency conversion control is required. In view of the inherent control problems in the CCM state, the traditional PI and PR controllers have poor control effects and require complex compensation network design, and the repeated controller design is complex and requires a large storage space. The ILC scheme derives the asymptotic stability of the closed-loop system by modeling the dual-mode flyback inverter as a discrete-time linear switching system, thereby ensuring the global convergence of the system in different working modes.

[0167] In order to further illustrate the technical solution of the present invention, a Matlab simulation model of a bidirectional AC-DC converter with optimized phase-shift frequency conversion control and traditional single-phase-shift frequency conversion control is built below. The simulation platform parameters are shown in Table 1:

[0168] Table 1 Parameters of dual-mode flyback microinverter simulation platform

[0169]

[0170]

[0171] Note: When the peak current is controlled, two transformer excitation inductances L are required for control. m Set to 7.4uH, leakage inductance L lk Set to 0.1uH, the switching frequency in DCM mode is 100kHz, and the other control parameters remain the same.

[0172] For the convenience of comparison, the output power is guaranteed to be 200W. Figure 6 (a) represents the primary and secondary currents of the two flyback converters of the proposed control scheme, Figure 6 (b) represents the peak current control of the primary and secondary currents of the two flyback converters. The proposed control scheme works in CCM / DCM mixed mode, with the primary and secondary peak currents being 18A and 3A respectively; the peak current control works in BCM / DCM mixed mode, with the primary and secondary peak currents being 23A and 3.83A respectively. Figure 7 (a) is the grid-connected voltage and grid-connected current waveform of the proposed control scheme, Figure 7 (b) is the grid-connected voltage and grid-connected current waveform under peak current control. During peak current control, the grid-connected current is obviously distorted at the mode switching point, which seriously affects the quality of the grid-connected current.

[0173] As a specific example of the present invention, the phase advance step size is selected based on the pulse response time in the DCM and CCM states. Fig.11 The impulse responses of the DCM and CCM control systems are given. It can be seen that the maximum impulse response time in the CCM state is longer than that in the DCM state. They mainly depend on the position of the RHP zero point, which changes with the change of the operating point. Considering the sampling frequency of 50kHz, the phase advance step size in the DCM mode in the embodiment is selected as λ d =3, the phase advance step size in CCM mode is selected as λ c =4.

[0174] The proportional gain k selected in this embodiment p is 0.11, and the learning control gain k l is 0.012, and the control parameters are brought into the Figure 3 Controller shown, and follow the Figure 2 The specific simulation parameters are shown in Table 1. Figure 6-Figure 10 The control effects of the proposed control scheme and the traditional peak current control and PI control for grid-connected current feedback are compared. It can be seen that the proposed control scheme can reduce the current stress of the switch tube and improve the quality of the grid-connected current. Figure 8-10 The grid-connected current, grid-connected voltage waveform, and grid-connected current THD are shown respectively when the proposed control scheme and the conventional grid-connected current feedback adopt PI control. The grid-connected current quality of the two different control schemes at full load (200W), medium load (100W), and light load (60W) are compared in turn. The grid-connected current THD of the proposed control scheme at full load is 1.58%, and the grid-connected current THD of the conventional grid-connected current feedback adopting PI control is 1.86%; the grid-connected current THD of the proposed control scheme at medium load is 2.64%, and the grid-connected current THD of the conventional grid-connected current feedback adopting PI control is 3.01%; the grid-connected current THD of the proposed control scheme at light load is 6.44%, and the grid-connected current THD of the conventional grid-connected current feedback adopting PI control is 8.01%. It can be seen that the proposed control scheme can improve the grid current quality compared with the traditional grid-connected current feedback control.

[0175] The following is an embodiment of a dual-mode flyback inverter phase advance compensation iterative learning control device of the present invention, which can be used to execute an embodiment of a dual-mode flyback inverter phase advance compensation iterative learning control method of the present invention. For details not disclosed in the embodiment of the dual-mode flyback inverter phase advance compensation iterative learning control device of the present invention, please refer to the embodiment of the dual-mode flyback inverter phase advance compensation iterative learning control method of the present invention.

[0176] In one embodiment, a dual-mode flyback inverter phase advance compensation iterative learning control device is proposed, referring to Fig.12 As shown, the device comprises:

[0177] The dual-mode control module is used to make the flyback inverter work in DCM and CCM dual modes based on grid-connected current feedback control through segmented feedforward compensation; and formulate the switching rules between DCM and CCM dual modes so that the DCM and CCM modes appear alternately within half a power frequency cycle;

[0178] The discrete time iterative linear system module is used to establish a dynamic mathematical model of the grid-connected process of the dual-mode flyback inverter and discretize the dynamic mathematical model to obtain a series of discrete time iterative linear systems;

[0179] A phase advance compensation iterative learning controller building module is used to build a phase advance compensation iterative learning controller according to a discrete time iterative linear system;

[0180] The parameter design and output control module performs the phase advance compensation iterative learning controller parameter design according to the convergence conditions that meet the stability of the closed-loop system, and realizes the optimal control of the dual-mode flyback inverter according to the grid-connected current feedback control process.

[0181] Furthermore, the phase advance compensation iterative learning controller includes: a state feedback controller, a feedforward input controller and an iterative learning controller; wherein,

[0182] The state feedback controller adopts proportional control to adjust the system gain of the dual-mode flyback inverter and speed up the response speed;

[0183] A feedforward input controller is used to calculate the nominal duty cycle in different switching modes and reduce the control gain of the state feedback controller;

[0184] The iterative learning controller adopts the form of P-type iterative learning control law to eliminate periodic interference, improve tracking accuracy, and avoid control system instability caused by RHP zero point in CCM mode.

[0185] The various functional modules in the embodiments of the present invention may be integrated into one processing module, or each unit may exist physically separately, or two or more units may be integrated into one module. The above integrated modules may be implemented in the form of hardware or in the form of software functional modules.

[0186] The ILC control scheme in this example, based on the traditional grid-connected current feedback control, makes the inverter work in a mixed mode of DCM and CCM through segmented feedforward compensation, and builds a dynamic mathematical model according to different switching modes. With the help of the dynamic mathematical model, the ILC controller is built, and the asymptotic stability of the closed-loop system is verified to ensure the global convergence of the system in different working modes. The advantage of traditional peak current control that it can operate in a single channel when lightly loaded is retained. At the same time, the original switching frequency does not need to be changed in the CCM mode. Compared with the variable switching frequency of the peak current control BCM mode, the switching loss and switch tube current stress are reduced; in addition, the ILC controller built with the dynamic mathematical model under different switching modes ensures the global convergence of the system in CCM and DCM modes, avoids the RHP zero compensation problem, and improves the waveform quality of the grid-connected current.

[0187] The above is only a preferred embodiment of the present invention, and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this patent can make some changes or modify the technical contents suggested above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the invention without departing from the content of the technical solution of the present invention still fall within the scope of the solution of the present invention.

Claims

1. A dual-mode flyback inverter phase advance compensation iterative learning control method, characterized in that: The method comprises: Based on grid-connected current feedback control, the flyback inverter is made to work in DCM and CCM dual modes through segmented feedforward compensation; the DCM mode means that the inductor current flowing through the flyback transformer is in a discontinuous state during the entire switching cycle, and the CCM mode means that the inductor current flowing through the flyback transformer is in a continuous state during the entire switching cycle; Formulate the switching rules between DCM and CCM modes so that DCM and CCM modes appear alternately within half of the power frequency cycle; The dynamic mathematical model of the grid-connected process of the dual-mode flyback inverter is established, and the dynamic mathematical model is discretized to obtain a series of discrete-time iterative linear systems; The specific steps of establishing the dynamic mathematical model of the dual-mode flyback inverter grid-connected process include: According to the simplified equivalent circuit of a flyback converter, considering the whole switching cycle, the state space equation in the whole switching cycle can be obtained by using Kirchhoff's voltage and current laws. Through the state space equation, the small signal model of DCM and CCM mode is obtained: According to the small signal model of DCM and CCM modes, and the flyback inverter works alternately in DCM and CCM modes in one grid cycle, its dynamic mathematical model is expressed as a linear switching system: y(t)=Cx(t), Since the dual-mode flyback inverter is based on the grid cycle T g The same task is repeated to describe the dynamic mathematical model as a series of iterative linear switching systems: y i (t)=Cx i (t), Where i represents the number of iterations, x i (t), y i (t), d i (t) are the state variables, output, and input at the i-th iteration respectively; According to the discrete time iterative linear system, a phase advance compensation iterative learning controller is constructed; The parameters of the phase advance compensation iterative learning controller are designed according to the convergence conditions that meet the stability of the closed-loop system, and the optimal control of the dual-mode flyback inverter is achieved according to the grid-connected current feedback control process.

2. The dual-mode flyback inverter phase advance compensation iterative learning control method according to claim 1, characterized in that: In order to facilitate the control of the output current in each mode, the nominal duty cycle in the two different working modes of DCM and CCM is used as the feedforward input. The nominal duty cycles of DCM and CCM modes are: Among them, V pv is the photovoltaic input voltage, P o is the output power of the dual-mode flyback inverter, L m is the flyback transformer magnetizing inductance, f s is the switching frequency, f g is the grid frequency, V g is the effective value of the grid voltage, N is the transformer ratio; The switching rules between DCM and CCM are as follows: DCM <D CCM The flyback inverter operates in DCM mode. DCM >D CCM When , the flyback inverter operates in CCM mode.

3. The dual-mode flyback inverter phase advance compensation iterative learning control method according to claim 2, characterized in that: The discrete-time iterative linear system is expressed as: x i (k+1)=Fx i (k)+Gu c (k) y i (k)=Hx i (k), Among them, x i (k)∈R 4 ×1,y i (k)∈R,u c (k)∈R are the state vector, output and control input when i=1,2…, respectively, F∈R 4 ×4,G∈R 4 ×1,H∈R 1 ×4 is the discrete time system matrix.

4. The dual-mode flyback inverter phase advance compensation iterative learning control method according to claim 3, characterized in that: The step of constructing a phase advance compensation iterative learning controller according to a discrete time iterative linear system comprises: (1) The construction of the phase advance compensation iterative learning controller must meet two assumptions: Assumption 1: The resulting discrete-time iterative linear system is reversible, that is, given the desired output trajectory y d (k) There exists a unique bounded input u d (k) such that: x d (k+1)=Fx d (k)+Gu d (k) y d (k)=Hx d (k),k∈[0,k T ], where x d (k)∈R 4 ×1 is the desired state vector; Assumption 2: All iterations satisfy the initial condition: x i (0) = x d (0), y i (0) = y d (0), i=1, 2…; (2) Based on the two assumptions, a phase advance compensation iterative learning control framework is constructed, and its control input can be expressed as: in c (k)=u fb (k)+u ff (k)+u i (k) Among them, u fb (k)∈R is the output value of the state feedback controller, u ff (k)∈R is the nominal duty cycle in two different switching modes, u i (k)∈R is the output value of the iterative learning controller; (3) According to the switching rules of the DCM and CCM dual modes, the constraints of the nominal duty cycle are set, and the state feedback controller adopts the form of proportional control, and the iterative learning controller adopts the form of P-type iterative learning control law to set the control satisfaction conditions of the phase lead compensation iterative learning controller.

5. The dual-mode flyback inverter phase advance compensation iterative learning control method according to claim 4, characterized in that: The control of the phase advance compensation iterative learning controller meets the conditions, which are specifically expressed as: you fb (k)=k p e(k) where k p is the proportional gain; λ d ,λ c ≥1 is the phase advance step size in DCM and CCM modes; e i (k) = y d -y i (k) is the output error; k l is the learning controller gain; i = 0, 1… is the iteration index; k = 0,…, k T is a discrete-time index.

6. The dual-mode flyback inverter phase advance compensation iterative learning control method according to claim 5, characterized in that: The convergence condition that satisfies the stability of the closed-loop system is obtained in the following way: Consider the discrete-time iterative linear system x i (k+1)=Fx i (k)+Gu c (k) and y i (k) = Hx i (k); If the two assumptions for constructing the phase advance compensation iterative learning controller are met, and the inequality ||E i+1 ||≤||Ik l P||||E i || is established, which can make the output tracking error of the closed-loop system of the phase advance compensation iterative learning controller converge monotonically to 0; Among them, E in the inequality i+1 is the error matrix of the i+1th iteration, I is k N ×k N The k-th order identity matrix, l To learn the controller gain, It can be expressed as: F s and G s Satisfy the following formula: F s =(F-k p GH) λ Where F, G, H are discrete time system matrices, λ is the iterative learning order, j = 0, 1, …λ-1.

7. The dual-mode flyback inverter phase advance compensation iterative learning control method according to claim 1, characterized in that: After constructing the phase advance compensation iterative learning controller according to the discrete time iterative linear system, the method further includes verifying the closed-loop stability of the established phase advance iterative learning controller. The specific verification steps include: (1) Applying the phase advance compensation iterative learning controller to the discrete time iterative linear system, the relationship between the control input at time k+1 in the i-th iteration and the control output at time k in the i-th iteration is obtained; (2) In the i-th iteration, k = 0, 1…λ-1 is brought into the discrete-time iterative linear system to which the phase advance compensation iterative learning controller is applied, and the relationship between the control input at each moment in the i-th iteration process and the control output at the previous moment can be obtained; (3) From the control conditions of the phase advance compensation iterative learning controller, it can be known that the control input u i (k) is the same for all k = 0, 1…λ-1, so the control input at time λ in the i-th iteration can be derived from the control output at the initial time; Applying the above steps (1) to (3) to each sampling, a discrete-time iterative linear system after applying a phase-advance compensation iterative learning controller to any control input signal can be obtained; Under the condition of satisfying Assumption 2, the newly obtained discrete-time iterative linear system is solved and the output error of the i+1th iteration is subtracted from the output error of the i-th iteration to obtain the condition for the output error to converge to 0, thereby completing the verification of the closed-loop stability of the phase-advance iterative learning controller.

8. A dual-mode flyback inverter phase advance compensation iterative learning control device, characterized in that: The device comprises: The dual-mode control module is used to make the flyback inverter work in DCM and CCM dual modes based on grid-connected current feedback control through segmented feedforward compensation; and formulate the switching rules between DCM and CCM dual modes so that the DCM and CCM modes appear alternately within half a power frequency cycle; The discrete time iterative linear system module is used to establish a dynamic mathematical model of the grid-connected process of the dual-mode flyback inverter and discretize the dynamic mathematical model to obtain a series of discrete time iterative linear systems; A phase advance compensation iterative learning controller building module is used to build a phase advance compensation iterative learning controller according to a discrete time iterative linear system; The phase advance compensation iterative learning controller includes: a state feedback controller, a feedforward input controller and an iterative learning controller; wherein, The state feedback controller adopts a proportional control form to adjust the system gain of the dual-mode flyback inverter to speed up the response speed; The feedforward input controller is used to calculate the nominal duty cycle under different switching modes and reduce the control gain of the state feedback controller; The iterative learning controller adopts the form of a P-type iterative learning control law to eliminate periodic interference, improve tracking accuracy, and avoid control system instability caused by RHP zero point in CCM mode; The parameter design and output control module performs the phase advance compensation iterative learning controller parameter design according to the convergence conditions that meet the stability of the closed-loop system, and realizes the optimal control of the dual-mode flyback inverter according to the grid-connected current feedback control process.

Images