CN103592984A - Method for decomposing and reconstructing current instantaneous sequence component of triangular connection current transformer - Google Patents

Abstract

The instantaneous order components of triangular connection current transformer electric current of the invention decompose and reconstructing method, including step S1, establish instantaneous order components decomposition algorithm model, make the full decoupled decomposition under synchronous coordinate system of fundamental wave and each harmonic positive sequence and negative sequence component; S2, m order harmonic components related coefficient is sought

With

S3, the compensation current reference value for reconstructing each chain link current controller of triangle, obtain the reference current of each chain link of triangle, and are used for triangle current transformer current follow-up control. It has the advantage that the positive sequence and negative sequence component for efficiently extracting fundamental wave and individual harmonic current, and is accurately realized load current and each chain current of triangle is quickly and accurately decomposed and reconstructed. For triangular connection current transformer provide accurately compensation current reference value, neatly can compensate for deficiency that number by selected harmonic, avoid traditional sequence decomposition algorithm it is computationally intensive and only consideration fundametal compoment.

Description

Triangle connects the instantaneous order component of current transformer electric current and decomposes and reconstructing method

Technical field

The invention belongs to Three-phase Power Systems Semiconductor Converting Technology field, the Current Decomposition and the reconstructing method that relate to current transformer, be specifically related to a kind of instantaneous decomposition and restructing algorithm that connects the current first harmonics of current transformer and the positive-sequence component of each harmonic and negative sequence component for triangle.

Background technology

In electric system, during the operation of electronic power conversion device, current imbalance and distortion are unavoidable.Especially at current transformers such as active filter, wind-powered electricity generation and photovoltaics, participate in the occasion of operation, three-phase current unbalance phenomenon is more common.Conventionally need to adopt the asymmetric and distortion situation of 3-phase power converter compensation three-phase current.Especially, in the large capacity of middle pressure uses, what adopt due to most of occasion is three-phase three-wire system (connecting without the center line) mode of connection, so 3-phase power converter often adopts triangle connected mode in this occasion.

For desirable electric system, due to voltage/current three-phase symmetrical, only contain positive-sequence component.When load unbalanced or the system failure, cause three-phase voltage/electric current asymmetric, thereby produce positive-sequence component and negative sequence component simultaneously.Therefore, in the time of can be by detection system fault or dissymmetrical load, positive-sequence component and negative sequence component be controlled system, guarantee the normal operation of system.Again due under triangle connected mode, between the line current of threephase load electric current and each chain link of triangle, there is the phase shift relation of 30 degree, therefore, to the extraction of the positive sequence of load current and negative sequence component minute walk-off angle internal reference electric current become triangle be connected current transformer control in unusual the key link, and the accuracy of the positive and negative sequence component of separated first-harmonic and each harmonic component and rapidity will directly affect dynamic responding speed and control accuracy that system is controlled.

Common positive-negative sequence component decomposition method mainly adopts the positive and negative order decomposition algorithm of low-pass filter or trapper that three-phase current is passed through to coordinate transform to forward synchronous coordinate system and reversion synchronous coordinate system at present, then under two coordinate systems, add low-pass filter or trapper respectively, conventionally through a few tens of milliseconds, realize again the decomposition of the first-harmonic positive-negative sequence component of voltage/current.Other conventional methods comprise that 1/4 power frequency period time-delay method of employing carries out positive and negative order separation in addition, the value after the instantaneous value of voltage/current and T/4 time delay is superposeed and subtract each other and obtain first-harmonic positive and negative sequence component, 10ms consuming time left and right.These methods have only realized the positive-negative sequence of fundametal compoment decomposes, and does not consider the impact of harmonic wave on first-harmonic positive-negative sequence component detection, does not consider the positive sequence of each harmonic and the decomposition of negative sequence component, and computation process is comparatively complicated simultaneously, and calculated amount is large.Therefore, these algorithms are all difficult to be applied to the occasion that overtone order is uncertain and response speed is had relatively high expectations of electric current.

Summary of the invention

The present invention, in order to solve the deficiencies such as existing triangle connection current transformer positive-negative sequence component separation algorithm computation process complexity and calculated amount are large, proposes a kind of triangle and connects the instantaneous order component decomposition of current transformer electric current and reconstructing method.

Concrete technical scheme of the present invention is: triangle connects the instantaneous order component of current transformer electric current and decomposes and reconstructing method, it is characterized in that, comprises the steps:

S1, set up and take the instantaneous order component decomposition algorithm model based on first-harmonic and each harmonic that three-phase current is benchmark, make the full decoupled decomposition under synchronous coordinate system of first-harmonic and each harmonic positive sequence and negative sequence component;

S2, ask for m order harmonic components related coefficient

with

wherein m is integer, and m>0 and m<0 represent respectively | the positive sequence of m| subharmonic and negative sequence component;

S3, utilize the offset current reference value of each chain link current controller of phase relation reconstruct triangle of harmonic component related coefficient, software phase-lock loop output, load current and each chain current of triangle, obtain the reference current of each chain link of triangle, and control for triangle unsteady flow device current tracking.

Further, the method for setting up the instantaneous order component decomposition algorithm model of first-harmonic and each harmonic in step S1 is:

S11, use following mode to express threephase load electric current:

Wherein m is integer, and m>0 and m<0 represent respectively | the positive sequence of m| subharmonic and negative sequence component, ω ₀for first-harmonic angular frequency, with

be respectively the initial phase angle of fundamental positive sequence and m component of degree n n,

the amplitude that represents fundamental positive sequence,

represent | the amplitude of m| subharmonic positive sequence or negative sequence component, the relation between itself and three-phase software phase-lock loop angle is respectively:

θ wherein ₁with Δ θ ₁represent voltage on line side fundamental phase angle and evaluated error that three-phase software phase-lock loop is estimated, the phase estimation error of m component of degree n n is Δ θ _m;

S12, three-phase current is transformed under dq synchronous coordinate system to d axle and q shaft current through Park conversion

with

be respectively:

θ wherein _pLL=ω ₀t+ θ ₁for system voltage genlock angle;

S13, arrange formula (3) and (4) and by i _ld ^{+ 1}and i _lq ^{+ 1}be rewritten as:

Wherein N is higher harmonic current number of times;

S14, two coefficients that order is relevant to m order harmonic components

with

be respectively:

When m=-6k-1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k-1)}=a_{-6k-1}^{\left(1\right)}\sin \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k-1}^{\left(2\right)}\cos \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(8\right)$

$i_{{\mathrm{Lq}}^{+1}(-6k-1)}=a_{-6k-1}^{\left(2\right)}\sin \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k-1}^{\left(1\right)}\cos \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(9\right)$

When m=6k-1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k-1)}={-a}_{6k-1}^{\left(1\right)}\sin \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k-1}^{\left(2\right)}\cos \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(10\right)$

$i_{{\mathrm{Lq}}^{+1}(6k-1)}=a_{6k-1}^{\left(2\right)}\sin \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k-1}^{\left(1\right)}\cos \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(11\right)$

When m=6k+1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(6k+1)}=-a_{6k+1}^{\left(1\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k+1}^{\left(2\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(12\right)$

$i_{{\mathrm{Lq}}^{+1}(6k+1)}=a_{6k+1}^{\left(2\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k+1}^{\left(1\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(13\right)$

When m=-6k+1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k+1)}=a_{-6k+1}^{\left(1\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k+1}^{\left(2\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(14\right)$

$i_{{\mathrm{Lq}}^{+1}(-6k+1)}=-a_{-6k+1}^{\left(2\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k+1}^{\left(1\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(15\right)$

S15, according to adaptive filter algorithm, formula (5) and (6) are transformed to following matrix form:

${\hat{Y}}_d=W_{d}X,{\hat{Y}}_q=W_{q}X---\left(16\right)$

W wherein _dand W _qfor weight vector coefficient, X is input vector, that is:

$\begin{array}{c}{W}_d=[w_{d0}^{\left(1\right)},w_{d2}^{\left(1\right)},w_{d2}^{\left(2\right)},w_{d4}^{\left(1\right)},w_{d4}^{\left(2\right)},...]\\ =[a_1^{\left(2\right)},a_{-1}^{\left(1\right)},a_{-1}^{\left(2\right)},{-a}_5^{\left(1\right)},a_5^{\left(2\right)},(-a_7^{\left(1\right)}+a_{-5}^{\left(1\right)}),(a_7^{\left(2\right)}+a_{-5}^{\left(2\right)}),\\ ...a_{-6k-1}^{\left(1\right)},a_{-6k-1}^{\left(2\right)},-a_{6k-1}^{\left(1\right)},a_{6k-1}^{\left(2\right)},(-a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}),(a_{6k+1}^{\left(2\right)}+a_{-6k+1}^{\left(2\right)}),...]\end{array}---\left(17\right)$

$\begin{array}{c}{W}_q=[w_{q0}^{\left(1\right)},w_{q2}^{\left(1\right)},w_{q2}^{\left(2\right)},w_{q4}^{\left(1\right)},w_{q4}^{\left(2\right)},...]\\ =[a_1^{\left(1\right)},{-a}_{-1}^{\left(2\right)},a_{-1}^{\left(1\right)},a_5^{\left(2\right)},a_5^{\left(1\right)},(a_7^{\left(2\right)}-a_{-5}^{\left(2\right)}),(a_7^{\left(1\right)}+a_{-5}^{\left(1\right)}),\\ ...{-a}_{-6k-1}^{\left(2\right)},a_{-6k-1}^{\left(1\right)},a_{6k-1}^{\left(2\right)},a_{6k-1}^{\left(1\right)},(a_{6k+1}^{\left(2\right)}-a_{-6k+1}^{\left(2\right)}),(a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}),...]\end{array}---\left(18\right)$

X=[1,sin[2(ω ₀t+θ ₁)],cos[2(ω ₀t+θ ₁)],...] ^T （19）

Further, in step S2, ask for m order harmonic components related coefficient

with

concrete grammar be:

M order harmonic components related coefficient

with

with weight vector coefficient W _dand W _qbetween there is following relation:

$\begin{array}{c}{a}_{-6k-1}^{\left(1\right)}=w_{d(6k-4)}^{\left(1\right)},a_{-6k-1}^{\left(2\right)}=w_{d(6k-4)}^{\left(2\right)}\\ a_{6k-1}^{\left(1\right)}=-w_{d(6k-2)}^{\left(1\right)},a_{6k-1}^{\left(2\right)}=w_{d(6k-2)}^{\left(2\right)}\end{array}---\left(20\right)$

$\left\{\begin{array}{c}{w}_{d\left(6k\right)}^{\left(1\right)}=-a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)},w_{d\left(6k\right)}^{\left(2\right)}=a_{6k+1}^{\left(2\right)}+a_{-6k+1}^{\left(2\right)}\\ w_{q\left(6k\right)}^{\left(1\right)}=a_{6k+1}^{\left(2\right)}-a_{-6k+1}^{\left(2\right)},w_{q\left(6k\right)}^{\left(2\right)}=a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}\end{array}\right.---\left(21\right)$

Draw the related coefficient of ± (6k ± 1) order harmonic components with

for:

$\left\{\begin{array}{c}{a}_{6k+1}^{\left(1\right)}=\frac{1}{2}[w_{q\left(6k\right)}^{\left(2\right)}-w_{d\left(6k\right)}^{\left(1\right)}],a_{6k+1}^{\left(2\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(2\right)}+w_{q\left(6k\right)}^{\left(1\right)}]\\ a_{-6k+1}^{\left(1\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(1\right)}+w_{q\left(6k\right)}^{\left(2\right)}],a_{-6k+1}^{\left(2\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(2\right)}-w_{q\left(6k\right)}^{\left(1\right)}]\end{array}\right.---\left(22\right)$

Further, described step S3 comprises the following steps:

S31, obtain the m order harmonic components of load A phase current, it carried out to trigonometric function and divide and solve:

According to the amplitude of load current and the intrinsic amplitude of each chain current of triangle and each chain current of the known triangle of phase propetry, it is load phase current magnitude

phase place leading phase electric current 30 degree;

S32, obtain the reference offset current that the m order harmonic components of AB chain link load current forms in each chain link of triangle for:

$\begin{array}{c}{i}_{\mathrm{ab},m}^{\mathrm{ref}}=1/\sqrt{3}{I}_L^m\cos (m{\mathrm{\&omega;}}_{0}t+{\mathrm{\&phi;}}^m+\mathrm{\&pi;}/6)\\ =1/\sqrt{3}{a}_m^{\left(2\right)}\cos (m({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)+\mathrm{\&pi;}/6)-1/\sqrt{3}{a}_m^{\left(1\right)}\sin (m({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)+\mathrm{\&pi;}/6)\end{array}---\left(24\right)$

S33, according to the method for step S31 and S32, carry out the reconstruct of all the other subharmonic current reference values, the offset current reference value of obtaining chain link AB, BC, CA is:

Wherein x=ab, bc, ca,

represent that respectively each time of chain link AB, BC, CA is with reference to the phase differential of offset current and corresponding number of times load current, wherein

S33, according to the positive sequence of the first-harmonic of reconstruct and individual harmonic current and negative sequence component, obtain the offset current reference value of each chain link of triangle, and control for the current tracking of current transformer.

Beneficial effect of the present invention: the instantaneous order component of electric current that connects current transformer for triangle of the present invention decomposes and restructing algorithm can extract positive sequence and the negative sequence component of first-harmonic and individual harmonic current effectively, and realize exactly load current and each chain current of triangle decomposes and reconstruct quickly and accurately.For triangle connects current transformer, provide offset current reference value accurately, selected harmonic compensation number of times, has avoided traditional order decomposition algorithm calculated amount to consider greatly and only the deficiency of fundametal compoment neatly.By introducing adaptive filter algorithm, adopt recursion iterative process, calculate quickly and accurately the weight coefficient of each component of voltage, overcome conventional lowpass filter and be subject to cutoff frequency to affect large shortcoming; The weight coefficient of asking for by iterative process obtains the related coefficient of each harmonic component, and individual harmonic current is carried out to selective extraction, is not subject to the impact of current harmonics number of times uncertainty and time variation.This algorithm utilizes load current and intrinsic phase place and the amplitude characteristic of each chain current of triangle, by decomposing the first-harmonic of load current and the positive sequence of each harmonic, negative sequence component, reconstruct exactly the reference value of each chain link individual harmonic current of triangle, for triangle connects current transformer, provide offset current reference value accurately.This algorithm is when the method for operation that load current is asymmetric and distortion is serious, all can decompose exactly and the offset current reference value that reconstructs triangle and be connected current transformer, stability is strong, overtone order is had to good adaptivity, for triangle connects current transformer, providing current tracking control signal accurately, is to realize important foundation and the key link that triangle connects the idle and harmonic compensation of current transformer.

Accompanying drawing explanation

Fig. 1 is the instantaneous order component decomposition of method of the present invention and the control flow chart of each chain link offset current reference value of reconstruct triangle;

Fig. 2 is the dynamic effect picture of method of the present invention under balanced load and load current sudden change;

Fig. 3 is the first-harmonic of method of the present invention under balanced load and load current sudden change and the dynamic effect picture of the decomposition of each harmonic order component and electric current reconstructing;

Fig. 4 is the dynamic effect picture of method of the present invention under asymmetrically placed load and load current sudden change;

Fig. 5 is the first-harmonic of method of the present invention under asymmetrically placed load and load current sudden change and the dynamic effect picture of the decomposition of each harmonic order component and electric current reconstructing.

Embodiment

Below in conjunction with accompanying drawing, embodiments of the invention are elaborated: the present embodiment is implemented take technical solution of the present invention under prerequisite, provided detailed embodiment and concrete operating process, but protection scope of the present invention is not limited to following embodiment.

As shown in Figure 1, the triangle of the present embodiment connects the instantaneous order component of current transformer electric current and decomposes and reconstructing method, comprises the steps:

S1, set up and take the instantaneous order component decomposition algorithm model based on first-harmonic and each harmonic that three-phase current is benchmark, make the full decoupled decomposition under synchronous coordinate system of first-harmonic and each harmonic positive sequence and negative sequence component.The described method of setting up the instantaneous order component decomposition algorithm model of first-harmonic and each harmonic is:

S11, due to the present embodiment for problem be that the instantaneous order component of electric current decomposes and reconstruct, therefore suppose that threephase load electric current forms (because the non-threephase load electric current forming by this kind of mode is not in problem scope solved by the invention) by first-harmonic and each harmonic component, uses following mode to express threephase load electric current:

Wherein m is integer, and m>0 and m<0 represent respectively | the positive sequence of m| subharmonic and negative sequence component, ω ₀for first-harmonic angular frequency,

with be respectively the initial phase angle of fundamental positive sequence and m component of degree n n,

the amplitude that represents fundamental positive sequence,

represent | the amplitude of m| subharmonic positive sequence or negative sequence component, the relation between itself and three-phase software phase-lock loop angle is respectively:

θ wherein ₁with Δ θ ₁represent voltage on line side fundamental phase angle and evaluated error that three-phase software phase-lock loop is estimated, the phase estimation error of m component of degree n n is Δ θ _m;

S12, three-phase current is transformed under dq synchronous coordinate system to d axle and q shaft current through Park conversion

with be respectively:

Because the present embodiment does not relate to the specific algorithm of software phase-lock loop, and system voltage genlock angle can accurately obtain by prior art, therefore supposes that software phase-lock loop has obtained the angle of system voltage genlock accurately, wherein θ _pLL=ω ₀t+ θ ₁for system voltage genlock angle;

S13, arrange formula (3) and (4) and by i _ld ^{+ 1}and i _lq ^{+ 1}be rewritten as:

Wherein N is higher harmonic current number of times; In view of the general characteristic of nonlinear-load, the characteristic harmonics of three-phase current is mainly 5,7,11,13 integers such as 6k ± 1 such as grade, i.e. m=± (6k ± 1).Therefore in step S14, make two coefficients relevant to m order harmonic components with

be respectively:

When m=-6k-1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k-1)}=a_{-6k-1}^{\left(1\right)}\sin \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k-1}^{\left(2\right)}\cos \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(8\right)$

$i_{{\mathrm{Lq}}^{+1}(-6k-1)}={-a}_{-6k-1}^{\left(2\right)}\sin \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k-1}^{\left(1\right)}\cos \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(9\right)$

When m=6k-1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k-1)}={-a}_{6k-1}^{\left(1\right)}\sin \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k-1}^{\left(2\right)}\cos \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(10\right)$

$i_{{\mathrm{Lq}}^{+1}(6k-1)}=a_{6k-1}^{\left(2\right)}\sin \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k-1}^{\left(1\right)}\cos \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(11\right)$

When m=6k+1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(6k+1)}=-a_{6k+1}^{\left(1\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k+1}^{\left(2\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(12\right)$

$i_{{\mathrm{Lq}}^{+1}(6k+1)}=a_{6k+1}^{\left(2\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k+1}^{\left(1\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(13\right)$

When m=-6k+1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k+1)}=a_{-6k+1}^{\left(1\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k+1}^{\left(2\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(14\right)$

$i_{{\mathrm{Lq}}^{+1}(-6k+1)}=-a_{-6k+1}^{\left(2\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k+1}^{\left(1\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(15\right)$

S15, according to adaptive filter algorithm, formula (5) and (6) are transformed to following matrix form:

${\hat{Y}}_d=W_{d}X,{\hat{Y}}_q=W_{q}X---\left(16\right)$

W wherein _dand W _qfor weight vector coefficient, X is input vector, that is:

$\begin{array}{c}{W}_d=[w_{d0}^{\left(1\right)},w_{d2}^{\left(1\right)},w_{d2}^{\left(2\right)},w_{d4}^{\left(1\right)},w_{d4}^{\left(2\right)},...]\\ =[a_1^{\left(2\right)},a_{-1}^{\left(1\right)},a_{-1}^{\left(2\right)},{-a}_5^{\left(1\right)},a_5^{\left(2\right)},(-a_7^{\left(1\right)}+a_{-5}^{\left(1\right)}),(a_7^{\left(2\right)}+a_{-5}^{\left(2\right)}),\\ ...a_{-6k-1}^{\left(1\right)},a_{-6k-1}^{\left(2\right)},-a_{6k-1}^{\left(1\right)},a_{6k-1}^{\left(2\right)},(-a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}),(a_{6k+1}^{\left(2\right)}+a_{-6k+1}^{\left(2\right)}),...]\end{array}---\left(17\right)$

$\begin{array}{c}{W}_q=[w_{q0}^{\left(1\right)},w_{q2}^{\left(1\right)},w_{q2}^{\left(2\right)},w_{q4}^{\left(1\right)},w_{q4}^{\left(2\right)},...]\\ =[a_1^{\left(1\right)},{-a}_{-1}^{\left(2\right)},a_{-1}^{\left(1\right)},a_5^{\left(2\right)},a_5^{\left(1\right)},(a_7^{\left(2\right)}-a_{-5}^{\left(2\right)}),(a_7^{\left(1\right)}+a_{-5}^{\left(1\right)}),\\ ...{-a}_{-6k-1}^{\left(2\right)},a_{-6k-1}^{\left(1\right)},a_{6k-1}^{\left(2\right)},a_{6k-1}^{\left(1\right)},(a_{6k+1}^{\left(2\right)}-a_{-6k+1}^{\left(2\right)}),(a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}),...]\end{array}---\left(18\right)$

X=[1,sin[2(ω ₀t+θ ₁)],cos[2(ω ₀t+θ ₁)],...] ^T （19）

By adaptive filter algorithm, ask for the process of weight vector coefficient and mentioned in the prior art, so the detailed process here can be with reference to prior art.

S2, ask for m order harmonic components related coefficient

with

wherein m is integer, and m>0 and m<0 represent respectively | the positive sequence of m| subharmonic and negative sequence component.In this step, ask for m order harmonic components related coefficient

with

concrete grammar be

M order harmonic components related coefficient

with

with weight vector coefficient W _dand W _qbetween there is following relation:

$\begin{array}{c}{a}_{-6k-1}^{\left(1\right)}=w_{d(6k-4)}^{\left(1\right)},a_{-6k-1}^{\left(2\right)}=w_{d(6k-4)}^{\left(2\right)}\\ a_{6k-1}^{\left(1\right)}=-w_{d(6k-2)}^{\left(1\right)},a_{6k-1}^{\left(2\right)}=w_{d(6k-2)}^{\left(2\right)}\end{array}---\left(20\right)$

$\left\{\begin{array}{c}{w}_{d\left(6k\right)}^{\left(1\right)}=-a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)},w_{d\left(6k\right)}^{\left(2\right)}=a_{6k+1}^{\left(2\right)}+a_{-6k+1}^{\left(2\right)}\\ w_{q\left(6k\right)}^{\left(1\right)}=a_{6k+1}^{\left(2\right)}-a_{-6k+1}^{\left(2\right)},w_{q\left(6k\right)}^{\left(2\right)}=a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}\end{array}\right.---\left(21\right)$

Draw the related coefficient of ± (6k ± 1) order harmonic components with

for:

$\left\{\begin{array}{c}{a}_{6k+1}^{\left(1\right)}=\frac{1}{2}[w_{q\left(6k\right)}^{\left(2\right)}-w_{d\left(6k\right)}^{\left(1\right)}],a_{6k+1}^{\left(2\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(2\right)}+w_{q\left(6k\right)}^{\left(1\right)}]\\ a_{-6k+1}^{\left(1\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(1\right)}+w_{q\left(6k\right)}^{\left(2\right)}],a_{-6k+1}^{\left(2\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(2\right)}-w_{q\left(6k\right)}^{\left(1\right)}]\end{array}\right.---\left(22\right)$

Above-mentioned relation can be learnt by above-mentioned derivation, therefore directly quote at this.Formula (22) shows, obtain exactly parameter

with

actual value, must be at the weight vector coefficient W that guarantees step S1 _dand W _qobtaining Accurate prerequisite under.

S3, utilize the offset current reference value of each chain link current controller of phase relation reconstruct triangle of harmonic component related coefficient, software phase-lock loop output, load current and each chain current of triangle, obtain the reference current of each chain link of triangle, and control for triangle unsteady flow device current tracking.

Step S3 is in order to obtain the reference current of each chain link of triangle, need to utilize the harmonic wave related coefficient of S2 acquisition and the output of software phase-lock loop, and in conjunction with the phase relation of load current and each chain current of triangle, reconstruct the offset current reference value of each chain link controller of triangle.Specifically comprise the following steps:

S31, obtain the m order harmonic components of load A phase current, it carried out to trigonometric function and divide and solve:

According to the amplitude of load current and the intrinsic amplitude of each chain current of triangle and each chain current of the known triangle of phase propetry, it is load phase current magnitude

phase place leading phase electric current 30 degree;

S32, obtain the reference offset current that the m order harmonic components of AB chain link load current forms in each chain link of triangle

for:

$\begin{array}{c}{i}_{\mathrm{ab},m}^{\mathrm{ref}}=1/\sqrt{3}{I}_L^m\cos (m{\mathrm{\&omega;}}_{0}t+{\mathrm{\&phi;}}^m+\mathrm{\&pi;}/6)\\ =1/\sqrt{3}{a}_m^{\left(2\right)}\cos (m({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)+\mathrm{\&pi;}/6)-1/\sqrt{3}{a}_m^{\left(1\right)}\sin (m({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)+\mathrm{\&pi;}/6)\end{array}---\left(24\right)$

S33, according to the method for step S31 and S32, carry out the reconstruct of all the other subharmonic current reference values, the offset current reference value of obtaining chain link AB, BC, CA is:

Wherein x=ab, bc, ca,

represent that respectively each time of chain link AB, BC, CA is with reference to the phase differential of offset current and corresponding number of times load current, wherein

S33, according to the positive sequence of the first-harmonic of reconstruct and individual harmonic current and negative sequence component, obtain the offset current reference value of each chain link of triangle, and control for the current tracking of current transformer.

What deserves to be explained is, certain phase that above-mentioned steps S31 to S33 specifically obtains or certain chain link can be other phase or chain links.

The course of work of above-described embodiment method is for the flow process by described in Fig. 1 is first by threephase load current i _la, i _lb, i _lcby park, convert and obtain dq component

with

according to adaptive filter algorithm, realize order component respectively and decompose, and by iterative computation weight coefficient Wd and Wq, utilize the related coefficient of formula (22) the calculating each harmonic in summary of the invention

with

acquisition system three-phase voltage V _sa, V _sb, V _sc, based on software phase-lock loop PLL, obtain θ _pLL=ω ₀t+ θ ₁, then utilize formula (24) and (25) to reconstruct the reference offset current of each chain link of triangle unsteady flow device

for the current tracking of each chain link, control current reference signal is accurately provided.

In order to verify for triangle, connect the instantaneous order component decomposition of electric current of current transformer and the feasibility of restructing algorithm, first-harmonic when Fig. 2～Fig. 5 decomposes instantaneous order component under two kinds of load behaviors and the simulation waveform of each harmonic and reconstruct harmonic wave, the operating mode 1 of the corresponding load symmetry of Fig. 2 and Fig. 3 and current break wherein, the operating mode 2 of the asymmetric and current break of the corresponding load of Fig. 4 and Fig. 5.Operating mode 1 time, system voltage effective value is 10kV, and load is three-phase diode rectifier bridge, DC side parallel RL branch road (L=2mH), after t=0.04s load changing, R switches to 25 Ω by 50 Ω, and AC is by the series reactance access electrical network of 2mH, mainly containing 5,7 characteristic harmonics; Operating mode 2 times, except accessing in system the load of operating mode 1, simultaneously also load-side AB mutually between single-phase diode rectifier bridge in parallel, its DC side parallel RL branch road (L=2mH, R switches to 50 Ω by 100 Ω when t=0.04s).

In Fig. 2 and Fig. 4, v _sa, v _sb, v _scand i _la, i _lb, i _lcbe respectively three-phase system voltage and load current,

offset current reference value for triangle AB, BC, each chain link of CA; In Fig. 3 and Fig. 5, i _ld ^{+ 1}and i _lq ^{+ 1}be respectively the dq axle component of load current after park conversion,

with

for the evaluated error of adaptive filter algorithm in order component decomposable process,

with

for the fundamental positive sequence coefficient extracting,

with

for first-harmonic negative sequence phasor,

with

be 5 negative sequence phasors,

with

be 7 positive sequence phasors,

for each chain link fundamental positive sequence reference value of reconstruct,

for the first-harmonic negative phase-sequence reference value of reconstruct,

for 5 subharmonic negative phase-sequence reference values of reconstruct,

the 7 subharmonic positive sequence reference values for reconstruct.

After introduction based on service condition and basic parameter in above-mentioned Fig. 2～Fig. 5, respectively the dynamic effect under operating mode 1 and operating mode 2 is described in detail below.

Fig. 2 and Fig. 3 have provided the dynamic effect picture when 1 time compensation balanced load of operating mode and load current sudden change.Be not difficult to find out, due to the existence of harmonic component, i _ld ^{+ 1}and i _lq ^{+ 1}the even-order harmonic that superposeed on the basis of direct current component, evaluated error after load changing

with

through the dynamic process of 20ms, get back to zero left and right, corresponding each harmonic coefficient estimates new steady-state value in 20ms.Due to load current three-phase symmetrical, there is not negative sequence component, first-harmonic negative sequence phasor

with

by zero, get back to zero after 20ms, transient state estimates that peak value is 50A, corresponding reconstruct first-harmonic negative phase-sequence

after 20ms, get back to zero; by 20A, be transitioned into 70A,

by-be transitioned into-60A of 20A,

waveform as shown in the figure; In like manner,

by 10A, be transitioned into 35A,

by 5A, be transitioned into 0A,

be about 20ms the transit time of getting back to steady-state value when load changing, finally according to the reconstruct component sum of each harmonic, forms the offset current reference value of each chain link of triangle as shown in Figure 2

three-phase offset current full symmetric.

Correspondingly, Fig. 4 and Fig. 5 have provided the dynamic effect picture when 2 times compensation asymmetric loads of operating mode and load current sudden change.Be not difficult to find out, due to the existence of first-harmonic negative sequence component in load current, i _ld ^{+ 1}and i _lq ^{+ 1}constituent different from Fig. 2, and with

after load changing, through 20ms, get back to zero left and right; First-harmonic negative sequence phasor

by 30A, be transitioned into 90A,

by 15A, be transitioned into 50A, reconstruct component

three-phase symmetrical, after sudden change, amplitude is transitioned into 103A by 33A through 10ms, finally according to the reconstruct component of each harmonic, forms the complete asymmetric offset current reference value of three-phase

According to the simulation result under above-mentioned two kinds of operating modes, show, the instantaneous order component decomposition of electric current and the restructing algorithm that connects current transformer for triangle in this paper, can under the operating modes such as load current is symmetrical and asymmetric, all can extract quickly and accurately first-harmonic positive-negative sequence in load current and the related coefficient of each harmonic component, and can reconstruct according to the inherent characteristic in load current and each chain link current amplitude of triangle and phase place the compensate component of individual harmonic current, finally for connecting current transformer, triangle provides offset current reference value accurately, the control strategy research that connects current transformer for next step three-phase triangle performs place mat.

What in the present invention, propose connects the instantaneous order component decomposition of electric current of current transformer and first-harmonic and the each harmonic component that restructing algorithm can extract electric current exactly for triangle, can be widely used in harmonic current and detect, its accuracy of detection, rapidity and operand all more traditional order decomposition algorithm are significantly improved; Simultaneously, the present invention proposes wink, Time Series Decomposition Method was owing to having adopted adaptive filter algorithm, can select neatly to extract according to the distribution situation of harmonic wave the harmonic compensation current reference value of corresponding number of times, more flexible, and be not subject to the impact of load current uncertainty and time variation, overcome the shortcoming that algorithm dynamic responding speed and precision can not be taken into account in the past, for current controller provides offset current reference value accurately, can be widely used in the control occasion that triangle connects current transformer.

The foregoing is only the specific embodiment of the present invention, one skilled in the art will appreciate that in the disclosed technical scope of the present invention, can carry out various modifications, replacement and change to the present invention.Therefore the present invention should not limited by above-mentioned example, and should limit with the protection domain of claims.

Claims (4)

1. triangle connects the instantaneous order component decomposition of current transformer electric current and reconstructing method, it is characterized in that, comprises the steps:

S1, set up and take the instantaneous order component decomposition algorithm model based on first-harmonic and each harmonic that three-phase current is benchmark, make the full decoupled decomposition under synchronous coordinate system of first-harmonic and each harmonic positive sequence and negative sequence component;

S2, ask for m order harmonic components related coefficient

with

wherein m is integer, and m>0 and m<0 represent respectively | the positive sequence of m| subharmonic and negative sequence component;

S3, utilize the offset current reference value of each chain link current controller of phase relation reconstruct triangle of harmonic component related coefficient, software phase-lock loop output, load current and each chain current of triangle, obtain the reference current of each chain link of triangle, and control for triangle unsteady flow device current tracking.

2. method according to claim 1, is characterized in that, the method for setting up the instantaneous order component decomposition algorithm model of first-harmonic and each harmonic in step S1 is:

S11, use following mode to express threephase load electric current:

Wherein m is integer, and m>0 and m<0 represent respectively | the positive sequence of m| subharmonic and negative sequence component, ω ₀for first-harmonic angular frequency,

with

be respectively the initial phase angle of fundamental positive sequence and m component of degree n n,

the amplitude that represents fundamental positive sequence,

represent | the amplitude of m| subharmonic positive sequence or negative sequence component, the relation between itself and three-phase software phase-lock loop angle is respectively:

θ wherein ₁with Δ θ ₁represent voltage on line side fundamental phase angle and evaluated error that three-phase software phase-lock loop is estimated, the phase estimation error of m component of degree n n is Δ θ _m;

S12, three-phase current is transformed under dq synchronous coordinate system to d axle and q shaft current through Park conversion

with

be respectively:

θ wherein _pLL=ω ₀t+ θ ₁for system voltage genlock angle;

S13, arrange formula (3) and (4) and by i _ld ^{+ 1}and i _lq ^{+ 1}be rewritten as:

wherein N is higher harmonic current number of times;

S14, two coefficients that order is relevant to m order harmonic components

with be respectively:

When m=-6k-1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k-1)}=a_{-6k-1}^{\left(1\right)}\sin \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k-1}^{\left(2\right)}\cos \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(8\right)$

$i_{{\mathrm{Lq}}^{+1}(-6k-1)}={-a}_{-6k-1}^{\left(2\right)}\sin \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k-1}^{\left(1\right)}\cos \left[(6k+2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(9\right)$

When m=6k-1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k-1)}={-a}_{6k-1}^{\left(1\right)}\sin \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k-1}^{\left(2\right)}\cos \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(10\right)$

$i_{{\mathrm{Lq}}^{+1}(6k-1)}=a_{6k-1}^{\left(2\right)}\sin \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k-1}^{\left(1\right)}\cos \left[(6k-2)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(11\right)$

When m=6k+1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(6k+1)}=-a_{6k+1}^{\left(1\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k+1}^{\left(2\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(12\right)$

$i_{{\mathrm{Lq}}^{+1}(6k+1)}=a_{6k+1}^{\left(2\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{6k+1}^{\left(1\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(13\right)$

When m=-6k+1, formula (5) and (6) are rewritten as:

$i_{{\mathrm{Ld}}^{+1}(-6k+1)}=a_{-6k+1}^{\left(1\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k+1}^{\left(2\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(14\right)$

$i_{{\mathrm{Lq}}^{+1}(-6k+1)}=-a_{-6k+1}^{\left(2\right)}\sin \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]+a_{-6k+1}^{\left(1\right)}\cos \left[\left(6k\right)({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)\right]---\left(15\right)$

S15, according to adaptive filter algorithm, formula (5) and (6) are transformed to following matrix form:

${\hat{Y}}_d=W_{d}X,{\hat{Y}}_q=W_{q}X---\left(16\right)$

W wherein _dand W _qfor weight vector coefficient, X is input vector, that is:

$\begin{array}{c}{W}_d=[w_{d0}^{\left(1\right)},w_{d2}^{\left(1\right)},w_{d2}^{\left(2\right)},w_{d4}^{\left(1\right)},w_{d4}^{\left(2\right)},...]\\ =[a_1^{\left(2\right)},a_{-1}^{\left(1\right)},a_{-1}^{\left(2\right)},{-a}_5^{\left(1\right)},a_5^{\left(2\right)},(-a_7^{\left(1\right)}+a_{-5}^{\left(1\right)}),(a_7^{\left(2\right)}+a_{-5}^{\left(2\right)}),\\ ...a_{-6k-1}^{\left(1\right)},a_{-6k-1}^{\left(2\right)},-a_{6k-1}^{\left(1\right)},a_{6k-1}^{\left(2\right)},(-a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}),(a_{6k+1}^{\left(2\right)}+a_{-6k+1}^{\left(2\right)}),...]\end{array}---\left(17\right)$

$\begin{array}{c}{W}_q=[w_{q0}^{\left(1\right)},w_{q2}^{\left(1\right)},w_{q2}^{\left(2\right)},w_{q4}^{\left(1\right)},w_{q4}^{\left(2\right)},...]\\ =[a_1^{\left(1\right)},{-a}_{-1}^{\left(2\right)},a_{-1}^{\left(1\right)},a_5^{\left(2\right)},a_5^{\left(1\right)},(a_7^{\left(2\right)}-a_{-5}^{\left(2\right)}),(a_7^{\left(1\right)}+a_{-5}^{\left(1\right)}),\\ ...{-a}_{-6k-1}^{\left(2\right)},a_{-6k-1}^{\left(1\right)},a_{6k-1}^{\left(2\right)},a_{6k-1}^{\left(1\right)},(a_{6k+1}^{\left(2\right)}-a_{-6k+1}^{\left(2\right)}),(a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}),...]\end{array}---\left(18\right)$

X=[1,sin[2(ω ₀t+θ ₁)],cos[2(ω ₀t+θ ₁)],...] ^T （19）。

3. method according to claim 1, is characterized in that, asks for m order harmonic components related coefficient in step S2

with

concrete grammar be:

M order harmonic components related coefficient

with

with weight vector coefficient W _dand W _qbetween there is following relation:

$\begin{array}{c}{a}_{-6k-1}^{\left(1\right)}=w_{d(6k-4)}^{\left(1\right)},a_{-6k-1}^{\left(2\right)}=w_{d(6k-4)}^{\left(2\right)}\\ a_{6k-1}^{\left(1\right)}=-w_{d(6k-2)}^{\left(1\right)},a_{6k-1}^{\left(2\right)}=w_{d(6k-2)}^{\left(2\right)}\end{array}---\left(20\right)$

$\left\{\begin{array}{c}{w}_{d\left(6k\right)}^{\left(1\right)}=-a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)},w_{d\left(6k\right)}^{\left(2\right)}=a_{6k+1}^{\left(2\right)}+a_{-6k+1}^{\left(2\right)}\\ w_{q\left(6k\right)}^{\left(1\right)}=a_{6k+1}^{\left(2\right)}-a_{-6k+1}^{\left(2\right)},w_{q\left(6k\right)}^{\left(2\right)}=a_{6k+1}^{\left(1\right)}+a_{-6k+1}^{\left(1\right)}\end{array}\right.---\left(21\right)$

Draw the related coefficient of ± (6k ± 1) order harmonic components

with

for:

$\left\{\begin{array}{c}{a}_{6k+1}^{\left(1\right)}=\frac{1}{2}[w_{q\left(6k\right)}^{\left(2\right)}-w_{d\left(6k\right)}^{\left(1\right)}],a_{6k+1}^{\left(2\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(2\right)}+w_{q\left(6k\right)}^{\left(1\right)}]\\ a_{-6k+1}^{\left(1\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(1\right)}+w_{q\left(6k\right)}^{\left(2\right)}],a_{-6k+1}^{\left(2\right)}=\frac{1}{2}[w_{d\left(6k\right)}^{\left(2\right)}-w_{q\left(6k\right)}^{\left(1\right)}]\end{array}\right.---\left(22\right).$

4. method according to claim 1, is characterized in that, described step S3 comprises the following steps:

S31, obtain the m order harmonic components of load A phase current, it carried out to trigonometric function and divide and solve:

According to the amplitude of load current and the intrinsic amplitude of each chain current of triangle and each chain current of the known triangle of phase propetry, it is load phase current magnitude

phase place leading phase electric current 30 degree;

S32, obtain the reference offset current that the m order harmonic components of AB chain link load current forms in each chain link of triangle

for:

$\begin{array}{c}{i}_{\mathrm{ab},m}^{\mathrm{ref}}=1/\sqrt{3}{I}_L^m\cos (m{\mathrm{\&omega;}}_{0}t+{\mathrm{\&phi;}}^m+\mathrm{\&pi;}/6)\\ =1/\sqrt{3}{a}_m^{\left(2\right)}\cos (m({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)+\mathrm{\&pi;}/6)-1/\sqrt{3}{a}_m^{\left(1\right)}\sin (m({\mathrm{\&omega;}}_{0}t+{\mathrm{\&theta;}}_1)+\mathrm{\&pi;}/6)\end{array}---\left(24\right)$

S33, according to the method for step S31 and S32, carry out the reconstruct of all the other subharmonic current reference values, the offset current reference value of obtaining chain link AB, BC, CA is:

Wherein x=ab, bc, ca,

represent that respectively each time of chain link AB, BC, CA is with reference to the phase differential of offset current and corresponding number of times load current, wherein

S33, according to the positive sequence of the first-harmonic of reconstruct and individual harmonic current and negative sequence component, obtain the offset current reference value of each chain link of triangle, and control for the current tracking of current transformer.

Images