CN103346585A - Grid-connected inverter control method based on state observer - Google Patents

Abstract

The invention discloses a grid-connected inverter control method based on a state observer. The grid-connected inverter control method comprises the steps that (1) a bridge arm current iL1 and network voltage eg are detected, and a phase angle theta is obtained through phase locking of a phase-locked loop; (2) an observation state quantity is obtained through the state observer based on an internal model; (3) a state quantity feedback signal Xfd is obtained through a state feedback matrix K; (4) grid currents of an observational network are processed, and then grid current error signals ed and eq are obtained; (5) closed-loop processing is carried out on the grid current error signals ed and eq through a PI controller, and then through coordinate inverse transformation, wave generation voltage ui1 is obtained; (6) the wave generation voltage ui1 and the state quantity feedback signal Xfd are subtracted from each other to generate a SVPWM control signal ui of an inverter bridge switching tube, output of a three-phase full bridge inverter is controlled, and therefore a distributed power generation system is controlled to be combined to the grid. Practices show that the control method obtains good dynamic and steady-state performance for a grid-connected current.

Description

Grid-connected inverter control method based on state observer

Technical Field

The invention belongs to the technical field of power grid converter control, and particularly relates to a grid-connected inverter control method based on a state observer.

Background

With the continuous application of renewable resources such as photovoltaic power generation, wind power generation, fuel cells, and the like, distributed power generation systems have become important in recent years and are receiving much attention. The grid-connected inverter is used as a core part of a distributed power generation system, and an LCL filter structure is usually adopted in the filter design. Compared with an L filter, the LCL filter can more effectively restrain current higher harmonics and reduce the total inductance. However, LCL filters have low damping high order characteristics, which tend to cause system resonance.

In order to inhibit the resonance of the LCL grid-connected inverter, a method for indirectly controlling the network access current of the inverter by utilizing the bridge arm current of the inverter is proposed by a scholart, the stability of a system is improved by the method, but the waveform quality of the network access current cannot be ensured. Subsequently, many scholars begin to research a double-closed-loop control method adopting direct network access current control, wherein an outer loop adopts a network access current closed loop, and an inner loop adopts damping schemes such as capacitance current feedback, capacitance voltage feedback, bridge arm current closed loop and the like. According to the double-closed-loop control scheme, the system oscillation is inhibited through the inner ring, so that the system stability is improved; however, the inner and outer ring designs in each scheme are coupled, which affects the control performance of the system; and these solutions require additional sensors, increasing the system hardware cost.

The scholars propose to reform the controlled object through state feedback, so as to simplify the control scheme of the design of the control system regulator and ensure the control performance of the grid-connected current. However, the realization of state feedback needs to detect a plurality of state quantities of the system, and an additional sensor is added. Therefore, researchers begin to research and design a traditional state observer to observe state feedback variables to replace an active damping link and an additional sensor added for state feedback, but the method does not consider the influence of characteristic matrix parameter change of the state observer and input deviation of the state observer on an observation effect, and the control performance of the system cannot be guaranteed.

Disclosure of Invention

The invention aims to overcome the defect that the traditional state observer does not consider the characteristic matrix parameter change of the state observer and the influence of the input deviation of the state observer on the observation effect, and provides a grid-connected inverter control method based on the state observer and the state observer thereof.

In order to achieve the purpose, the invention adopts the following technical scheme:

a grid-connected inverter control method based on a state observer is provided, and the topological structure of the grid-connected inverter related to the control method comprises a direct current source U_dcDC side filter capacitor C_dcThree-phase full-bridge inverter circuit, LCL filter and DC side filter capacitor C_dcIs connected in parallel to a direct current source U_dcTwo ends of (1), a direct current source U_dcThe two power output ends of the three-phase full-bridge inverter circuit are respectively connected with two input ends of the three-phase full-bridge inverter circuit, the three-phase output end of the three-phase full-bridge inverter circuit is correspondingly connected with the three-phase input end of the LCL filter, and the three-phase output end of the LCL filter is respectively connected with the three-phase power grid e_a、e_b、e_cConnecting; what is needed isThe LCL filter is composed of an inverter side inductor L₁Network side filter inductor L₂And a filter capacitor C₁Composition is carried out; the control method comprises the following steps:

step one, detecting bridge arm current i by utilizing current sensor of bridge arm current_L1(ii) a Detection of a network voltage e by means of a network voltage sensor_gPerforming phase locking to obtain a phase angle theta;

step two, according to the bridge arm voltage u_iAnd the network voltage e_gAnd obtaining an observation state quantity through a state observer based on an internal model: observing bridge arm current

Observing the voltage of the capacitor

And observing the net side current

Thirdly, obtaining a state quantity feedback signal X from the observation state quantity in the second step through a state feedback matrix K_fdThe method comprises the following steps:

firstly, a state feedback matrix K, K = [ K ] is obtained through an expected pole allocation method₁，k₂，k₃]，k₁、k₂、k₃All are state feedback coefficients;

secondly, observing the observation state quantity in the step two, namely observing the bridge arm current

Observing the voltage of the capacitor

Observing net side currentAre respectively multiplied by state feedback coefficients k₁、k₂、k₃Obtaining the state quantity feedback signal $X_{\mathrm{fd}}=[{\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">k</figure-callout>}}_1\&times;{\hat{i}}_{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">L</figure-callout>}1};{\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">k</figure-callout>}}_2\&times;{\hat{u}}_C;{\mathrm{<figure-callout\; id="3"\; label="k"\; filenames="HDA00003478918900024.png"\; state="\{\{state\}\}">k</figure-callout>}}_3\&times;{\hat{i}}_{L2}];$

Step four, observing the current of the network sideObtaining a network side current active component i through abc-dq coordinate transformation_{d_fdb}Net side current reactive component i_{q_fdb}Respectively associated with the network-side current to give an active signal i_{d_fref}Given a reactive signal i_{q_ref}Making difference to obtain net side current error signal e_d、e_qNamely:

e_d＝i_{d_fref}－i_{d_fdb}，

e_q＝i_{q_ref}－i_{q_fdb}；

step five, the network side current error signal e obtained in the step four is subjected to comparison through a PI controller_d、e_qClosed-loop processing is carried out, and then the inverse transformation of dq-abc coordinates is carried out to obtain wave-emitting voltage u_i1；

Sixthly, the wave-generating voltage u is measured_i1With said state quantity feedback signal X_fdSVPWM control signal u of three-phase full-bridge inverter circuit can be generated by superposition_i。

Further, the state observer based on the internal model in the second step is implemented by the following steps:

selecting bridge arm current i_L1Capacitor voltage u_CNet side current i_L2As a system state quantity, bridge arm voltage u is set_iAnd the network voltage e_gAs an input quantity, a discrete state space equation of the grid-connected inverter system is obtained through the following formula:

$\left\{\begin{array}{c}x(k+1)=A_{d}x\left(k\right)+B_{d}u\left(k\right)\\ y\left(k\right)=C_{d1}x\left(k\right)+D_{d}u\left(k\right)\end{array}---\left(1\right)\right.$

in the formula (1), A_dFor the grid-connected inverter system characteristic matrix, B_dFor grid-connected inverter system input matrix, C_d1For the grid-connected inverter system output matrix, D_dDirectly transmitting a matrix for a grid-connected inverter system;

x (k) represents the system state quantity: bridge arm current i_L1Capacitor voltage u_CAnd net side current i_L2；

x (k +1) is the system state quantity delayed by one beat;

u (k) is input quantity of the grid-connected inverter system;

y (k) is the output quantity of the grid-connected inverter system;

secondly, according to the discrete state space equation of the grid-connected inverter system, the change of the characteristic matrix parameters of the state observer is respectively considered

And the state observer input deviation delta u (k) is obtained by utilizing the following formula through a dual principle to obtain a state observer state space equation based on an internal model:

$\left\{\begin{array}{c}\hat{x}(k+1)={\hat{A}}_d\hat{x}\left(k\right)+{\hat{B}}_d{u}^{\&prime;}\left(k\right)+G(1+{\mathrm{\&phi;}}_e^{-1}\left(z\right))[y\left(k\right)-\hat{y}\left(k\right)]\\ \hat{y}\left(k\right)={\hat{C}}_{d1}\hat{x}\left(k\right)+{\hat{D}}_{d}u\left(k\right)\end{array}\right.---\left(2\right)$

in the formula (2), the reaction mixture is,

in the form of a state observer feature matrix,

in order to input the matrix to the state observer,

in order to output the matrix for the state observer,

directly transmitting the matrix for a state observer;

represents the observed state quantity: observing bridge arm currentObserving the voltage of the capacitor

And observing the net side current

The state quantity is an observation state quantity delayed by one beat;

u' (k) is the input quantity of the state observer, and G is the feedback matrix of the state observer;

y (k) is the grid-connected inverter system output,

outputting an error for the state observer;

is an internal mold term.

State observer characteristic matrixAn observation error Deltax obtained by the above formulas (1) and (2) when the parameter is changed₁Represented by the formula:

in the formula (3), the reaction mixture is,

being characteristic of state observersThe amount of change in the matrix is,

to take into account the state observer feature matrix

The internal model item set when the parameter changes is output error according to the state observer

Setting as a first-order integral link or a proportional resonance link;

part (a) in the formula (3) reaches a value of 0 by an expected pole allocation method, and the formula (3) passes through the internal model of part (b)

Is set so that the sum of both the part (b) and the part (c) reaches a value of 0, and finally the observation error Δ x₁A value of 0 is reached.

(iv) when the state observer input amount u' (k) is deviated, the observation error Δ x obtained by the expressions (1) and (2)₂Represented by the formula:

in the formula (4), Δ u (k) is a change amount of the state observer input amount u (k),

in order to take account of internal model terms set when the state observer input u' (k) is biased, the error is output according to the state observer

Setting as a first-order integral link or a proportional resonance link;

the part (d) in the formula (4) reaches a value of 0 by an expected pole allocation method, and the formula (4) passes through the internal model of the part (e)

Is set so that the sum of both the part (e) and the part (f) reaches a value of 0, and finally the observation error Δ x₂A value of 0 is reached.

According to the invention, the influence of the characteristic matrix parameter change of the state observer and the input deviation of the state observer on the observation effect are considered, the accurate observation state quantity is obtained through the state observer based on the internal model, the controlled object is transformed through state feedback, and the current loop control of the grid-connected inverter is carried out by adopting the current on the side of the observation network, so that good dynamic and steady-state performances are obtained.

Drawings

Fig. 1 is a structure diagram of a grid-connected inverter control based on a state observer according to the present invention.

Fig. 2 is a structural diagram of a conventional state observer.

Fig. 3 is a structural diagram of a state observer based on an internal model.

Fig. 4a and 4b are comparison simulation graphs of dynamic observation effects of bridge arm currents of a traditional state observer and a state observer based on an internal model in a static coordinate system respectively.

Fig. 5a and 5b are comparison simulation graphs of dynamic observation effects of a traditional state observer and a state observer based on an internal model on bridge arm currents in a rotating coordinate system respectively.

Fig. 6a and 6b are comparison simulation graphs of bridge arm current steady-state observation effects of a traditional state observer and a state observer based on an internal model in a static coordinate system respectively.

Fig. 7a and 7b are experimental waveforms of dynamic comparison of bridge arm currents of a conventional state observer and a state observer based on an internal model in a static coordinate system respectively.

Fig. 8a and 8b are waveforms of steady-state comparison experiments of bridge arm currents of a conventional state observer and a state observer based on an internal model in a static coordinate system respectively.

Fig. 9a and 9b are respectively waveforms of an active current comparison experiment of a bridge arm in a rotating coordinate system by a traditional state observer and a state observer based on an internal model.

Fig. 10a and 10b are experimental waveforms of capacitance-voltage dynamic comparison of a conventional state observer and an internal model-based state observer in a static coordinate system, respectively.

Fig. 11a and 11b are respectively the dynamic experimental waveforms of the current on the net side based on the feedback of the observed state of the conventional state observer and the state observer based on the internal model.

Detailed Description

In order to describe the present invention more specifically, the state observer-based grid-connected inverter control method and the state observer thereof according to the present invention will be described in detail with reference to the accompanying drawings and the detailed description.

In the present embodiment, as shown in fig. 1, a three-phase LCL grid-connected inverter converts a 260V dc power into a 220V ac power, and inputs the ac power to a power grid after being filtered by an LCL filter.

The LCL filter parameters are: bridge arm filter inductance L₁=1mH, parasitic resistance r of bridge arm filter inductor₁=0.001 Ω, filter capacitance C₁=20 μ F, net side filter inductance L₂=0.5mH, parasitic resistance r of filter inductor on network side₂=0.001 Ω, grid impedance L_gThe variation range is 0-1.5 mH.

A grid-connected inverter control method based on a state observer comprises the following steps:

1. current sensor H using bridge arm current_L1a、H_L1bDetecting bridge arm current i_L1(ii) a Using grid voltage sensors H_ea、H_ebDetecting network voltagee_a、e_b、e_cPerforming phase locking to obtain a phase angle theta;

2. according to bridge arm voltage u_iAnd the network voltage e_gThe following observed state quantities are obtained by a state observer based on an internal model: observing bridge arm current

Observing the voltage of the capacitor

And observing the net side current

The present example is as follows:

selecting a state variable: bridge arm current i_L1Capacitor voltage u_CNet side current i_L2Is marked as x ═ i_L1，u_C，i_L2]^T(ii) a Selecting input quantity: output voltage u of inverter bridge arm_iAnd the network electromotive force e_gAnd is recorded as u = [ u ]_i，e_g]^T，

Establishing a state space model of the LCL grid-connected inverter system:

$\left\{\begin{array}{c}\stackrel{\&CenterDot;}{x}=\mathrm{Ax}+\mathrm{Bu}\\ y=\mathrm{Cx}+\mathrm{Du}\end{array}\right.$

in the above formula, A is a grid-connected inverter system characteristic matrix, $A=\left[\begin{array}{ccc}-\frac{{\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">r</figure-callout>}}_1+R_p}{L_1}& -\frac{1}{L_1}& \frac{{\mathrm{<figure-callout\; id="1"\; label="R\; p\; L"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">R</figure-callout>}}_p}{L_{}}\\ \frac{1}{C}& 0& -\frac{1}{C}\\ \frac{{\mathrm{<figure-callout\; id="2"\; label="R\; p\; L"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">R</figure-callout>}}_p}{L_{}^{}}& \frac{1}{{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2^{\&prime;}}& -\frac{{\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">r</figure-callout>}}_2+{\mathrm{<figure-callout\; id="2"\; label="R\; P\; L"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">R</figure-callout>}}_P}{L_{}^{}}\end{array}\right];$

wherein R is_pIs a filter capacitor C₁The damping resistors are connected in series;

L₂' is a network side filter inductor L₂Equivalent inductance L with power grid impedance_gSumming;

b is the grid-connected inverter system input matrix, $B=\left[\begin{array}{ccc}& .& \\ {\mathrm{<figure-callout\; id="1"\; label="B"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">B</figure-callout>}}_1& .& {\mathrm{<figure-callout\; id="2"\; label="B"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">B</figure-callout>}}_2\\ & .& \end{array}\right]=\left[\begin{array}{cc}1/{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">L</figure-callout>}}_1& 0\\ 0& 0\\ 0& -1/{{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2}^{\&prime;}\end{array}\right];$

c is output matrix of grid-connected inverter system, and when selecting bridge arm current i_L1As an output, C₁＝[1，0，0](ii) a When the net side current i is selected_L2As an output, C₂＝[0，0，1]；

And D is a direct transmission matrix of the grid-connected inverter system, and D is [0, 0 ].

For the convenience of digital control, the sampling frequency f is chosen_sDiscretization method with zero order keeper at 10 kHz: C_d1＝C₁，D_ddiscretizing the continuous state equation to obtain a discrete state space equation of the LCL grid-connected inverter:

$\left\{\begin{array}{c}x(k+1)=A_{d}x\left(k\right)+B_{d}u\left(k\right)\\ y\left(k\right)=C_{d1}x\left(k\right)+D_{d}u\left(k\right)\end{array}\right.$

in the above formula, C_d1＝[1，0，0]Namely, selecting bridge arm current as output;

x (k) is a discretized state quantity x (k) ═ i_L1(k)，u_C(k)，i_L2(k)]^TWherein: i.e. i_L1(k) For discrete domain bridge arm current, u_C(k) For discrete domain capacitor voltage, i_L2(k) Is a discrete domain network side current;

u (k) is a discrete domain input, u (k) is [ u (k) ]_i(k)，e_g(k)]^TWherein: u. of_i(k) For discrete domain inverter leg output voltage, e_g(k) Is the grid electromotive force of the discrete domain.

According to the LCL grid-connected inverter discrete state space model, on the basis of a traditional state observer (as shown in fig. 2), the state observer based on an internal model is obtained by considering the characteristic matrix parameter change of the state observer and the input change of the observer, and the state space expression of the state observer is as follows:

$\left\{\begin{array}{c}\hat{x}(k+1)={\hat{A}}_d\hat{x}\left(k\right)+{\hat{B}}_d{u}^{\&prime;}\left(k\right)+G(1+{\mathrm{\&phi;}}_e^{-1}\left(z\right))[y\left(k\right)-\hat{y}\left(k\right)]\\ \hat{y}\left(k\right)={\hat{C}}_{d1}\hat{x}\left(k\right)+{\hat{D}}_{d}u\left(k\right)\end{array}\right.$

wherein,

in the form of a state observer feature matrix,

in order to input the matrix to the state observer,

in order to output the matrix for the state observer,for the state observer to transmit the matrix directly,

is the variation of the characteristic matrix of the state observer;

represents the observed state quantity: observing bridge arm current

Observing the voltage of the capacitorAnd observing the net side current

u (k) is the input quantity of the state observer, and Δ u (k) is the variation of the input quantity u (k) of the state observer;

g is a feedback matrix of the state observer;

the output quantity of the state observer is y (k), and the output quantity of the grid-connected inverter is y (k);

an internal model term is represented.

According to the attached figure 3, the internal model-based state observer outputs the grid-connected inverter y (k), namely the bridge-arm side filter inductive current i of the grid-connected inverter, on the basis of an LCL discrete state space equation_L1And the output of the state observer

I.e. observing bridge arm currentMaking a difference to obtain an output error e (k); because the expression form of the output error is different in different coordinate systems, the analysis is convenient and the output error is consideredThe general form of the output error, denoted as E;

and solving a characteristic polynomial according to the output error E:

φ_e(z)＝det(zI-E)

in the formula, phi_e(z) a characteristic polynomial representing the output error E, z being a discrete transform operator;

taking the output error E into account as the input of the internal model, phi is expressed by the following formula_e(z) an unstable pole polynomial:

φ(z)＝z^m+a_m-1z^m-1+a_m-2z^m-2+…+a₁z+a₀

wherein phi (z) is the first polynomial of z, a_iRepresents a polynomial coefficient, said a_i∈[0，m-1]M is the order of discrete transform factors;

internal mold phi^-1(z) is achieved by the following equation of state:

$\left\{\begin{array}{c}{x}_c(k+1)=G_c{x}_c\left(k\right)+b_{c}e\left(k\right)\\ y_c\left(k\right)=x_c\left(k\right)\end{array}\right.$

in the formula, G_cAn equation feature matrix is implemented for the internal model, $G_c=\left[\begin{array}{cccccc}0& 1& 0& 0& ...& 0\\ 0& 0& 1& 0& ...& 0\\ .& & & & & .\\ .& & & & & .\\ .& & & & & .\\ 0& ...& & & 0& 1\\ -a_0& -a_1& ...& & & -a_{m-1}\\ & & & & & \end{array}\right],$

b_cimplementing the input matrix for the internal model, b_c＝[0,0…0,1]^T；

x_c(k) Is an internal model state quantity, x_c(k +1) is an internal model state delayed by one beat,e (k) is the actual output error, expressed as $e\left(k\right)=y\left(k\right)-\hat{y}\left(k\right);$

In the three-phase rotating coordinate system, the error and the state quantity of the output quantity are constant, and the internal modelMay be provided as a first order integral regulator in general; in a static coordinate system, the error of the output quantity and the state quantity are sinusoidal quantities, and the internal model can be set as a resonance regulator.

Referring to FIG. 3, the output error e (k) is compared with the output term x passing through the internal model_c(k) And adding, namely obtaining accurate observation state quantity through a state observation matrix G: observing bridge arm currentObserving the voltage of the capacitor

Observing net side current

In FIG. 3, I_gFor net side active current, Zoh is a zero order keeper, 1/(sL)₁+r₁) As a transfer function of bridge arm inductance, 1/(sC)₁) Is a filter capacitance transfer function, 1/(sL)₂'+r₂) Is the net side filter inductance transfer function.

3. According to the observation state quantity in the step 2, a state quantity feedback signal X is obtained through a state feedback matrix K_fd. Introducing the state feedback to the systemAre reconfigured so that the poles of the system are distributed within the z-domain unit circle, thereby increasing the damping of the control system.

In this embodiment, the state feedback is obtained separately from the internal model-based state observer in step 2 according to a separation principle.

The discrete state space expression based on state feedback is

x(k+1)＝(A_d-B_dK)x(k)+B_du(k)

Including damping resistors R according to capacitive branch_pSystem feature matrix A of_d，A_dSeries resistor R of filter capacitor_p=0.5 Ω, and a system characteristic polynomial including passive damping is obtained; according to the system characteristic matrix A after state feedback_d-B_dK, using the corresponding term coefficients to be equal, to obtain a state feedback matrix K, i.e.

det(zI-A_d)＝det(zI-(A_d-B_dK))

Wherein z is a discrete transform operator, I is a 3 × 3 unit matrix, det (·) represents a characteristic polynomial, and K is a state feedback matrix, and is denoted as K = [ K ]₁，k₂，k₃]，k₁、k₂、k₃Referred to as state feedback coefficients;

the obtained state feedback coefficient is as follows: k is a radical of₁=2.8108609696657359180517147320431，

k₂=-0.42233855315199843186755694448278，

k₃=-2.8075596495029625451005704996724；

And (3) the observation state quantity in the step 2: observing bridge arm current

Observing the voltage of the capacitor

Observing net side current

Are respectively multiplied by state feedback coefficients k₁、k₂、k₃Obtaining a state quantity feedback signal $X_{\mathrm{fd}}=[{\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">k</figure-callout>}}_1\&times;{\hat{i}}_{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA00003478918900012.png,HDA00003478918900013.png"\; state="\{\{state\}\}">L</figure-callout>}1};{\mathrm{<figure-callout\; id="2"\; label="k"\; filenames="HDA00003478918900021.png,HDA00003478918900022.png"\; state="\{\{state\}\}">k</figure-callout>}}_2\&times;{\hat{u}}_C;{\mathrm{<figure-callout\; id="3"\; label="k"\; filenames="HDA00003478918900024.png"\; state="\{\{state\}\}">k</figure-callout>}}_3\&times;{\hat{i}}_{L2}].$

4. Will observe the net side current

Obtaining the active component i of the network side current through coordinate transformation_{d_fdb}Net side current reactive component i_{q_fdb}Respectively associated with the network-side current to give an active signal i_{d_fref}Given a reactive signal i_{q_ref}Making difference to obtain net side current error signal e_d、e_q；

5. Through PI controller to network side current error signal e_d、e_qPerforming closed loop processing, wherein PI controller takes k_p=15，k_i=1800, and obtaining wave-emitting voltage u through coordinate inverse transformation_i1。

6. Will u_i1And the state feedback signal X_fdSuperposition generated inverseSVPWM control signal u of bridge-changing switch tube, namely three-phase full-bridge inverter circuit_iAnd further controlling the grid-connected inverter grid-connected current.

Fig. 4a to 6b are comparison simulation waveforms of observed effect dynamics and steady state when the impedance of the network side changes in the conventional state observer and the state observer based on the internal model. Fig. 4a and 4b are comparison simulation graphs of dynamic observation effects of bridge arm currents in a static coordinate system, wherein a label '1' represents a bridge arm current actual measurement waveform, and a label '2' represents a bridge arm current observation waveform; 5a, 5b are comparison simulation graphs of dynamic observation effect of bridge arm current of a rotating coordinate system, wherein a label '1' represents an actual measurement waveform of bridge arm active current, and a label '2' represents an observation waveform of bridge arm active current; the reference numeral '3' represents the actual measurement waveform of the reactive current of the bridge arm, and the reference numeral '4' represents the observation waveform of the reactive current of the bridge arm.

Fig. 7a to 10b are experimental waveforms for comparing observed effect dynamics and steady state when the impedance of the network side changes based on the conventional state observer and the internal model state observer. It can be seen from the simulation diagram and the experimental diagram that the state observer based on the internal model is superior to the traditional state observer in both dynamic effect and steady state effect, and the observed state quantity is used for state feedback and further used for grid-connected current control, and the grid-side current dynamic waveform is good.

The invention can be applied to various grid-connected inverter control structures, reduces the cost of a control system and simultaneously improves the reliability of the control system. The method has universality and is not limited by a hardware structure and a control structure.

Claims (7)

1. A grid-connected inverter control method based on a state observer is provided, and the topological structure of the grid-connected inverter related to the control method comprises a direct current source U_dcDC side filter capacitor C_dcThree-phase full-bridge inverter circuit, LCL filter, and DC side filter capacitor C_dcIs connected in parallel to a direct current source U_dcTwo ends of (1), a direct current source U_dcThe two power output ends of the three-phase full-bridge inverter circuit are respectively connected with the two input ends of the three-phase full-bridge inverter circuit, the three-phase output ends of the three-phase full-bridge inverter circuit are correspondingly connected with the three-phase input end of the LCL filter one by one, and the LCL filterThe three-phase output end is respectively connected with a three-phase power grid e_a、e_b、e_cConnecting; the LCL filter is composed of an inverter side inductor L₁Network side filter inductor L₂And a filter capacitor C₁Composition is carried out; the method is characterized by comprising the following steps:

step one, detecting bridge arm current i by utilizing current sensor of bridge arm current_L1(ii) a Detection of a network voltage e by means of a network voltage sensor_gPerforming phase locking to obtain a phase angle theta;

step two, according to the bridge arm voltage u_iAnd the network voltage e_gAnd obtaining an observation state quantity through a state observer based on an internal model: observing bridge arm current

Observing the voltage of the capacitor

And observing the net side current

Thirdly, obtaining a state quantity feedback signal X from the observation state quantity obtained in the second step through a state feedback matrix K_fd；

Step four, observing the current of the network side

Obtaining a network side current active component i through abc-dq coordinate transformation_{d_fdb}Net side current reactive component i_{q_fdb}Respectively associated with the network-side current to give an active signal i_{d_fref}Given a reactive signal i_{q_ref}Making difference to obtain net side current error signal e_d、e_qNamely:

e_d＝i_{d_fref}－i_{d_fdb}，

e_q＝i_{q_ref}－i_{q_fdb}；

step five, the network side current error signal e obtained in the step four is subjected to comparison through a PI controller_d、e_qClosed-loop processing is carried out, then abc-dq coordinate is inversely transformed, and wave-transmitting voltage u is obtained_i1；

Sixthly, the wave-generating voltage u is measured_i1With said state quantity feedback signal X_fdSVPWM control signal u of three-phase full-bridge inverter circuit can be generated by superposition_i。

2. The grid-connected inverter control method based on the state observer as claimed in claim 1, wherein the state quantity feedback signal X is obtained through the state feedback matrix K in the third step_fdThe steps are as follows:

firstly, a state feedback matrix K, K = [ K ] is obtained through an expected pole allocation method₁，k₂，k₃]，k₁、k₂、k₃All are state feedback coefficients;

secondly, observing the state quantity, namely observing the bridge arm current, in the step two

Observing the voltage of the capacitorObserving net side current

Are respectively multiplied by state feedback coefficients k₁、k₂、k₃That is, the state quantity feedback signal is obtained

3. The grid-connected inverter control method based on the state observer according to claim 1, wherein the state observer based on the internal model in the second step is realized by the following steps:

selecting bridge arm current i_L1Capacitor voltage u_CGrid side electricityStream i_L2As a system state quantity, bridge arm voltage u is set_iAnd the network voltage e_gAs an input quantity, a discrete state space equation of the grid-connected inverter system is obtained through the following formula:

in the formula (1), A_dFor the grid-connected inverter system characteristic matrix, B_dFor grid-connected inverter system input matrix, C_d1For the grid-connected inverter system output matrix, D_dDirectly transmitting a matrix for a grid-connected inverter system;

x (k) represents the system state quantity: bridge arm current i_L1Capacitor voltage u_CAnd net side current i_L2；

x (k +1) is the system state quantity delayed by one beat;

u (k) is input quantity of the grid-connected inverter system;

y (k) is the output quantity of the grid-connected inverter system;

secondly, according to the discrete state space equation of the grid-connected inverter system, the change of the characteristic matrix parameters of the state observer is respectively considered

And the state observer input deviation delta u (k) is obtained by utilizing the following formula through a dual principle to obtain a state observer state space equation based on an internal model:

in the formula (2), the reaction mixture is,

in the form of a state observer feature matrix,

in order to input the matrix to the state observer,

in order to output the matrix for the state observer,

directly transmitting the matrix for a state observer;

represents the observed state quantity: observing bridge arm current

Observing the voltage of the capacitor

And observing the net side current

The state quantity is an observation state quantity delayed by one beat;

u' (k) is the input quantity of the state observer, and G is the feedback matrix of the state observer;

y (k) is the grid-connected inverter system output,

outputting an error for the state observer;

is an internal mold term.

4. The grid-connected inverter control method based on the state observer according to claim 3, wherein: the characteristic matrix of the state observer

An observation error Deltax obtained by the above formulas (1) and (2) when the parameter is changed₁Represented by the formula:

in the formula (3), the reaction mixture is,

in order to be the state observer feature matrix variation,to take into account the state observer feature matrix

An internal model item is set when the parameter is changed,

part (a) in the formula (3) reaches a value of 0 by an expected pole allocation method, and the formula (3) passes through the internal model of part (b)

Is set so that the sum of both the part (b) and the part (c) reaches a value of 0, and finally the observation error Δ x₁A value of 0 is reached.

5. The grid-connected inverter control method based on the state observer and the state observer thereof according to claim 3, wherein: when the input quantity u' (k) of the state observer is deviated, the following formula (1) andobservation error Δ x obtained by equation (2)₂Represented by the formula:

in the formula (4), Δ u (k) is a change amount of the state observer input amount u (k),

to account for the internal model terms set when the state observer input u' (k) is biased,

the part (d) in the formula (4) reaches a value of 0 by an expected pole allocation method, and the formula (4) passes through the internal model of the part (e)Is set so that the sum of both the part (e) and the part (f) reaches a value of 0, and finally the observation error Δ x₂A value of 0 is reached.

6. The grid-connected inverter control method based on the state observer according to claim 4, wherein: the internal mold term in the formula (3)

According to the output error of the state observer

The first-order integral element or the proportional resonance element is set.

7. The grid-connected inverter control method based on the state observer according to claim 5, wherein: the internal mold term in the formula (4)

According to the output error of the state observer

The first-order integral element or the proportional resonance element is set.

Images