CN105162333A - DAB-BDC modulation method based on high-frequency alternating-current buck-boost principle - Google Patents

Abstract

The invention discloses a DAB-BDC modulation method based on a high-frequency alternating-current buck-boost principle. An energy buffer inductance current is designed to be in a current critical continuous mode (BCM) or interrupted mode (DCM), constant-frequency control is adopted, and a switching device in a BDC is ensured to realize zero voltage switching (ZVS) and zero current switching (ZCS). An algorithm realizing optimal current stress multivariable solutions of a device is provided, a current sensor is omitted in the control process, and the dynamic performance of the system is directly improved. Compared with the prior art, the method provided by the invention has good performance.

Description

A kind of DAB-BDC modulator approach based on high-frequency ac buck principle

Technical field

The present invention relates to a kind of DAB-BDC modulator approach based on high-frequency ac buck principle, achieve the algorithm of switching device optimal current stress multivariable solution, and eliminate current sensor in the controlling, directly enhance the dynamic property of system, belong to the control field of converters.

Background technology

Two-way DC/DC converter (BidirectionalDC/DCconverter, BDC) application scenarios such as Aero-Space, generation of electricity by new energy, AC and DC micro-capacitance sensor are widely used in, cost, volume, the weight of system is saved, so be subject to increasing research because of its bidirectional power flow kinetic force.

BDC divides non-isolation type and isolated form two class, and because the input of BDC, output voltage fluctuation range are comparatively large, and both electric pressures generally differ comparatively large, and therefore non-isolated BDC is difficult to realize high efficiency by single-stage power conversion.And isolated form BDC can utilize high frequency transformer to mate input, output voltage grade, make conversion efficiency higher.Usual isolated form BDC can be divided into voltage-type BDC and current mode BDC, and current mode BDC breaker in middle Guan Yi causes due to voltage spikes, must increase clamp circuit and realize due to voltage spikes suppression, which increase the complexity of circuit; And voltage-type BDC DC side has capacitor-clamped, voltage stress is less, is therefore widely used.Classical isolated DC transducer can obtain voltage-type BDC after circuit modification, general circuit is structure symmetrically, there is inverse-excitation type BDC, Forward-flyback mixing BDC, half-bridge reversible transducer and full-bridge reversible transducer etc., wherein, two active bridge (DualActivebridge, DAB) BDC is owing to can realize zero voltage switch (ZVS) easily, and phase shifting control can realize to and fro flow of power, and is subject to extensive use.Transformer primary secondary ac square-wave voltage phase place, pulsewidth is all adjustable, achieves more achievement in research in recent years for reduction feedback power, device current peak value and total losses aspect.

In prior art, comparatively advanced is the concept expanding phase shift, be intended to the loss reducing backflow power and generation thereof, and achieve good effect, but there are three independently control variables in BDC, i.e. former limit bridge duty ratio d1, phase shifting angle φ between secondary bridge duty ratio d2 and former secondary, not proposing a kind of best practice realizing these three variablees, first fixing the duty ratio of a bridge when realizing, therefore the optimal control of unrealized backflow power; Some scholar is minimum for purpose of design to realize switching tube peak current, proposes the control strategy of dual phase shift, but the duty ratio pre-setting former secondary full-bridge is equal, still can optimize further; Some scholar converter processing power of giving chapter and verse realizes variable frequency control stage by stage, and to ensure the efficiency comparatively optimized, control by stages realizes more difficult on the one hand, adopts the triangle of electric current to control on the other hand, makes device current stress larger; All kinds of loss models of complete BDC are established in prior art, the size of above-mentioned d1, d2 and φ is determined according to the design principle that loss is minimum, be really control program optimum at present, but its operational effect depends on the accuracy of loss model, and the foundation of loss model is very complicated.In addition, in DAB-BDC, the acquisition of ZVS is relevant to load, during underloading, is difficult to realize Sofe Switch.

The control strategy of tradition DAB-BDC comprises an output voltage outer ring, a current inner loop, because current inner loop feedback quantity is inductive current, a usual increase low pass filter obtains more level and smooth mean value and carries out closed loop feedback, but low pass filter can have influence on the dynamic characteristic of system.Therefore inductive current PREDICTIVE CONTROL is widely studied, but the method is very strong according to lazyness to the precision in current measurement moment, and slightly error can affect control effects.List of references

[1]A. A.V′azquez，D.G.Lamar,etal.DifferentPurposeDesignStrategiesandTechniquestoImprovethePerformanceofaDualActiveBridgeWithPhase-ShiftControl[J].IEEETrans.onPowerElectronics，2015，30(2)：790–804.[2]B.Zhao，Q.Song，W.Liu.EfficiencyCharacterizationandOptimizationofIsolatedBidirectionalDC-DCConverterBasedonDual-Phase-ShiftControlforDCDistributionApplication[J].IEEETrans.onPow.Electron，2013，28(4)：1711-1727.

Summary of the invention

Goal of the invention: for existing methodical deficiency, a kind of DAB-BDC modulation strategy based on high-frequency ac buck principle is proposed herein, by the design of energy snubber inductive current in electric current critical continuous conduction mode (BCM) or discontinuous mode (DCM), and adopt constant frequency to control, ensure that the switching device in BDC realizes ZVS and ZCS.Propose a kind of algorithm realizing device optimal current stress multivariable solution, and eliminate current sensor in the controlling, directly enhance the dynamic property of system.

Technical scheme: a kind of DAB-BDC modulator approach based on high-frequency ac buck principle, in DAB-BDC, input, output voltage are U _in, U _o, then buffer inductance both sides voltage is

$u_L={\mathrm{AU}}_{in}-B\frac{U_o}{n}---\left(1\right)$

In formula, A, B are respectively the state of inductance both sides current potential, and n is transformer voltage ratio;

The control strategy of DAB-BDC is as follows:

The positive and negative periodic symmetry of high frequency, every half period comprises 4 mode, and the modulation ratio of 4 mode is respectively d ₁, d ₂, d ₃, d ₄; In positive half period, the value that under 4 mode, the value of A is respectively 1,1,0,0, B is respectively 0,1,1,0; In negative half-cycle, the value that under 4 mode, the value of A is respectively-1 ,-1,0,0, B is respectively 0 ,-1 ,-1,0;

According to the situation of input voltage, output voltage and power output, above-mentioned all 4 mode in half switch periods, may not be comprised.

Make T _sfor switch periods, y ₁, y ₂be respectively first mode finish time and second mode finish time buffer inductance current i _lvalue; According to the time acted on buffer inductance and voltage, determine

$y_1=\frac{U_{in}{d}_1{T}_S}{2L}---\left(2\right)$

$y_2=\frac{{\mathrm{nU}}_{in}{T}_S{d}_1+({\mathrm{nU}}_{in}-U_o)T_S{d}_2}{2nL}---\left(3\right)$

In formula, L is the inductance value of buffer inductance, equal with slippage according to inductive current ascending amount, obtains

$d_3=\frac{{\mathrm{nU}}_{in}{d}_1+({\mathrm{nU}}_{in}-U_o)d_2}{U_o}---\left(4\right)$

Primary side current of transformer i _l, only have the electric current of second mode and the 3rd mode time period to flow to load-side, the mean value of electric current after transformer conversion of these two mode equals load current,

$d_2^2({\mathrm{nU}}_{in}-U_o)+d_2(2{\mathrm{nd}}_1{U}_{in}+{\mathrm{nd}}_3{U}_{in}-d_3{U}_o)+{\mathrm{nd}}_1{d}_3{U}_{in}-4n^2{\mathrm{LI}}_o/T_S=0---\left(5\right)$

(4) are substituted into (5),

$d_2^2\left({\mathrm{nU}}_{in}\right({\mathrm{nU}}_{in}-U_o\left)\right)+d_2\left(2n^2{U}_{in}^2{d}_1\right)+n^2{U}_{in}^2{d}_1^2-4n^2{\mathrm{LU}}_o{I}_o/T_S=0---\left(6\right)$

Solve

$d_2=\frac{x_2{d}_1+\sqrt{x_3{d}_1^2+x_4}}{x_1}---\left(7\right)$

d ₃＝x ₅d ₁+x ₆d ₂(8)

Wherein

x ₁＝U _in(nU _in-U _o)； $x_2=-{\mathrm{nU}}_{in}^2;x_3={\mathrm{nU}}_{in}^3{U}_o;$

$x_4=\frac{4{\mathrm{nU}}_{in}{U}_o{\mathrm{LI}}_o({\mathrm{nU}}_{in}-U_o)}{T_S};x_5=\frac{{\mathrm{nU}}_{in}}{U_o};x_6=\frac{({\mathrm{nU}}_{in}-U_o)}{U_o}$

Only when second mode and the 3rd mode, inductive current i _ljust load-side can be flow to.Therefore, if d ₂+ d ₃value comparatively large, then when same load current, electric current is just comparatively level and smooth, and the current effective value that switching device bears is also less; Make y=d ₂+ d ₃, then d when asking y to obtain maximum ₁, d ₂, d ₃value, therefore just require below the maximum of y; By the expression formula of (7), (8) substitution y,

$y=x_5{d}_1+(1+x_6)\frac{x_2{d}_1+\sqrt{x_3{d}_1^2+x_4}}{x_1}---\left(9\right)$

D is asked to (9) ₁derivative,

$\frac{dy}{d\left(d_1\right)}=x_5+\frac{x_2(1+x_6)}{x_1}+\frac{x_3(1+x_6)d_1}{x_1\sqrt{x_3{d}_1^2+x_4}}=\frac{x_2}{x_1}+\frac{x_3(1+x_6)d_1}{x_1\sqrt{x_3{d}_1^2+x_4}}---\left(10\right)$

(10) formula is made to equal 0, when obtaining that in one section of interval, y obtains extreme value, d ₁the value of correspondence be d _1y

$d_{1y}=\sqrt{\frac{x_4{x}_2^2}{{(x_3+x_3{x}_6)}^2-x_3{x}_2^2}}---\left(11\right)$

According to nU _inwith U _orelation, and the size solved, has three kinds of situations

I.nU _in≥U _o

Now, except x ₂outside <0, x ₁, x ₃~ x ₆all be greater than zero, the amount therefore in formula (7) inside radical sign is greater than zero automatically, and will ensure d ₂be greater than zero, d ₁must meet

$d_{1x1}=0\&le;d_1\&le;\sqrt{\frac{x_4}{(x_2^2-x_3)}}=d_{1z1}---\left(12\right)$

By d _1x1, d _1y, d _1z1substitution formula (9), getting amount corresponding to its maximum is d ₁value, then try to achieve d according to (7), (8) respectively ₂, d ₃value.

II.nU _in＜U _o

Now, except x ₁, x ₂, x ₄, x ₆all be less than 0, x ₃, x ₅be greater than zero, therefore except d will be ensured ₂be greater than outside zero, also will ensure that the amount in formula (7) inside radical sign is greater than zero, must meet

$d_{1x2}=\sqrt{\frac{-x_4}{x_3}}\&le;d_1\&le;\sqrt{\frac{-x_4}{(x_3-x_2^2)}}=d_{1z2}---\left(13\right)$

By d _1x2, d _1y, d _1z2substitution formula (9), getting amount corresponding to its maximum is d ₁value, then try to achieve d according to (7), (8) respectively ₂, d ₃value.

Also has a kind of possibility, when substitution data calculate d ₁+ d ₂+ d ₃during >1, illustrate that needing to become switching frequency could realize, for realizing constant frequency, must by d ₁+ d ₂+ d ₃value be limited in 1, and using this restriction as solving d ₁condition;

III.d ₁+d ₂+d ₃＝1

Now

d ₃＝1-d ₁-d ₂(14)

In the case, formula (4) still meets, and (4) are substituted into (14),

$d_2=\frac{U_o-({\mathrm{nU}}_{in}+U_o)d_1}{{\mathrm{nU}}_{in}}---\left(15\right)$

Ensure d ₂>0, then meet

$d_1\&le;\frac{U_o}{{\mathrm{nU}}_{in}+U_o}---\left(16\right)$

In the case, the relation of formula (6) still meets, and (15) are substituted into (6),

$(n^2{U}_{in}^2+{\mathrm{nU}}_{in}|u_G|+{U_o}^2)d_1^2-2{U_o}^2{d}_1+{U_o}^2-{\mathrm{nu}}_{in}{U_o}^2+\frac{4n^3{U}_{in}{\mathrm{Li}}^{*}}{T_S}=0---\left(17\right)$

According to the constraints of (16),

$d_1=\frac{-b-\sqrt{b^2-4ac}}{2a}---\left(18\right)$

Wherein

$a=n^2{U}_{in}^2+{\mathrm{nu}}_{in}{U}_o+{U_o}^2,b=-2{U_o}^2,c={U_o}^2-{\mathrm{nU}}_{in}{U}_o+\frac{4n^3{U}_{in}{\mathrm{LI}}_o}{T_S}$

Then d is tried to achieve according to (14), (15) ₂with d ₃;

According to above-mentioned three kinds of situations about calculating, the d obtained under finally choosing a kind of situation ₁, d ₂with d ₃as the modulation ratio of final front 3 mode, if d ₁+ d ₂+ d ₃=1, then d ₄=0; If d ₁+ d ₂+ d ₃<1, then d ₄=1-(d ₁+ d ₂+ d ₃), and by d ₁, d ₂, d ₃with d ₄as source signal, directly realized the drive singal of switching tube by modulation strategy.

Accompanying drawing explanation

Fig. 1 is DAB-BDC circuit structure diagram;

Fig. 2 is high-frequency ac ascending, descending pressure modulation strategy, and wherein (a) is high-frequency ac step-down control mode, and (b) is high-frequency ac buck control mode, and (c) is high-frequency ac boosting rectifier control mode;

Fig. 3 is unified high-frequency ac buck mode;

Fig. 4 is that multivariable solves flow chart;

Fig. 5 is the interchange buck control strategy of DAB-BDC;

Fig. 6 is that current effective value is with input voltage change curve;

Fig. 7 is DPS control strategy;

Fig. 8 is that the current effective value of DPS control is with input voltage change curve;

Fig. 9 controls to compare with current stress under this paper institute extracting method for DPS.

Embodiment

Below in conjunction with specific embodiment, illustrate the present invention further, these embodiments should be understood only be not used in for illustration of the present invention and limit the scope of the invention, after having read the present invention, the amendment of those skilled in the art to the various equivalent form of value of the present invention has all fallen within the application's claims limited range.

As shown in Figure 1, the topological structure of DAB-BDC, comprises the low-pressure side full-bridge be made up of switching tube S1-S4, the high-pressure side full-bridge that S5-S8 is formed, C _r1-C _r8for the equivalent output capacitance of corresponding switching tube, L is energy snubber inductance, comprises series inductance and transformer leakage inductance sum, U _inand U _obe respectively the magnitude of voltage of low pressure and high pressure, C ₁and C ₂be respectively low-pressure side and on high-tension side filter capacitor, i _l, i _sbe respectively transformer primary secondary current, u _l1, u _sbe respectively low-pressure side, high-pressure side full-bridge circuit AC voltage, u _l2for voltage u _sconversion is to the value on former limit, and T is high frequency transformer, and n is transformer voltage ratio, R _lfor load, I _ofor load current.

In the active bridge of traditional double, voltage u _l1, u _sbe all the high-frequency ac voltage of positive and negative width 50%, control u _l1, u _sbetween phase shifting angle just can realize flow of power, but this control method is only at U _in=U _ohigher efficiency can be obtained near/n, its reason be the method control under converter at U _inwith U _owhen/n difference is larger, there is larger feedback power.For overcoming this problem, in the positive half cycle of switch or negative half period, if the energy be stored in inductance can be discharged completely, i.e. positive half cycle and before negative half period start time, the electric current of inductance L is 0.

Input, output voltage U in DAB-BDC _in, U _ofluctuation very large, for raising the efficiency, the no-load voltage ratio n of general Design of Transformer to mate input, output voltage, therefore, usual voltage u _l1and u _l2magnitude relation be unfixed, for ensureing that the electric current of inductance L is 0 before half switch periods terminates, can by the control strategy of buck, buck/boost, boost circuit of classics when discontinuous current.Suppose that in buck, buck/boost, boost tri-kinds of circuit, input, output voltage are all U _inand U _o/ n, then corresponding inductance both end voltage is

u _L＝AU _in-BU _o/n(1)

In formula, A, B are respectively the state of inductance both sides current potential, and its value is as shown in table 1, has three mode in a switch periods respectively, and corresponding current rises, declines and remain zero respectively.The method is expanded to the high-frequency ac in DAB-BDC, then the voltage u in corresponding diagram 1 _l1, u _l2, and current i _lwaveform as shown in Figure 2, respectively corresponding high-frequency ac step-down, buck and boosting.D in figure _bu, D _bb, D _bowhen being respectively step-down, buck time and boosting time modulation ratio.

The input of table 1 DC converter inductance, output voltage state

Preset voltage threshold, according to voltage u _l1and u _l2magnitude relation determination circuit working in that control mode shown in Fig. 2, so how realize ZVS and ZCS of device, but also there is obvious shortcoming in this control strategy:

1). relative to the situation of current continuity state, current i _lpeak value comparatively large, the current stress causing device to bear is larger;

2). control aspect discontinuous, easily causes system unstable when switching.

For ZVS and the ZCS characteristic of retaining means, and be the current stress of suppression device, must ensure that inductance works in DCM or BCM on the one hand, the control strategy that unified control high-frequency ac buck principle is corresponding must be excavated on the other hand.In DAB-BDC, inductance both end voltage still meets formula (1), and the present invention proposes control strategy as shown in table 2.

The high-frequency ac buck control strategy that table 2 is unified

The positive and negative periodic symmetry of high frequency shown in Fig. 2, every half period comprises 4 mode, and modulation ratio is respectively d ₁, d ₂, d ₃, d ₄.Explain with positive half cycle, the first two mode in table 2 is identical with the boost in table 1, and modulation ratio is d ₂, d ₃two mode identical with the buck in table 1, two state common modulation are than being d ₂mode, namely table 2 put forward the Hybrid mode mode that control strategy essence is buck, boost, as long as change the time ratio shared by them, just can easily achieve boosting or step-down, and controling parameters consecutive variations, without saltus step.The 3rd mode finish time, current i _lvanishing, remains device ZCS feature.According to the situation of input, output voltage and power output, one or two mode shown in table 2 in half switch periods, may not be comprised.

According to control strategy shown in table 2, a switch periods internal inductance both sides voltage u _l1, u _l2and inductive current i _lschematic diagram as shown in Figure 3.In figure, d ₁, d ₂, d ₃, d ₄distinguish the modulation ratio of 4 mode in the half period in corresponding table 2, T _sfor switch periods, y ₁, y ₂be respectively first mode finish time and second mode finish time current i _lvalue.

According to the time acted on inductance and voltage, determine

$y_1=\frac{U_{in}{d}_1{T}_S}{2L}---\left(2\right)$

$y_2=\frac{{\mathrm{nU}}_{in}{T}_S{d}_1+({\mathrm{nU}}_{in}-U_o)T_S{d}_2}{2nL}---\left(3\right)$

In Fig. 3, equal with slippage according to inductive current ascending amount, obtain

$d_3=\frac{{\mathrm{nU}}_{in}{d}_1+({\mathrm{nU}}_{in}-U_o)d_2}{U_o}---\left(4\right)$

Current i in Fig. 3 _l, only have the electric current of second mode and the 3rd mode time period can flow to load-side, the dash area namely in figure, therefore the mean value of electric current after transformer conversion of these two mode equals load current,

$d_2^2({\mathrm{nU}}_{in}-U_o)+d_2(2{\mathrm{nd}}_1{U}_{in}+{\mathrm{nd}}_3{U}_{in}-d_3{U}_o)+{\mathrm{nd}}_1{d}_3{U}_{in}-4n^2{\mathrm{LI}}_o/T_S=0---\left(5\right)$

(4) are substituted into (5),

$d_2^2\left({\mathrm{nU}}_{in}\right({\mathrm{nU}}_{in}-U_o\left)\right)+d_2\left(2n^2{U}_{in}^2{d}_1\right)+n^2{U}_{in}^2{d}_1^2-4n^2{\mathrm{LU}}_o{I}_o/T_S=0---\left(6\right)$

Solve

$d_2=\frac{x_2{d}_1+\sqrt{x_3{d}_1^2+x_4}}{x_1}---\left(7\right)$

d ₃＝x ₅d ₁+x ₆d ₂(8)

Wherein

x ₁＝U _in(nU _in-U _o) $x_2=-{\mathrm{nU}}_{in}^2{x}_3={\mathrm{nU}}_{in}^3{U}_o{x}_5=\frac{{\mathrm{nU}}_{in}}{U_o}$

$x_4=\frac{4{\mathrm{nU}}_{in}{U}_o{\mathrm{LI}}_o({\mathrm{nU}}_{in}-U_o)}{T_S}{x}_6=\frac{({\mathrm{nU}}_{in}-U_o)}{U_o}$

As can be seen from Figure 3, only when second mode and the 3rd mode, inductive current i _ljust load-side can be flow to.Therefore, if d ₂+ d ₃value comparatively large, then when same load current, electric current is just comparatively level and smooth, and the current effective value that switching device bears is also less.Make y=d ₂+ d ₃, then d when asking y to obtain maximum ₁, d ₂, d ₃value, therefore just require below the maximum of y.(7), (8) are substituted into,

$y=x_5{d}_1+(1+x_6)\frac{x_2{d}_1+\sqrt{x_3{d}_1^2+x_4}}{x_1}---\left(9\right)$

D is asked to (9) ₁derivative,

$\frac{dy}{d\left(d_1\right)}=x_5+\frac{x_2(1+x_6)}{x_1}+\frac{x_3(1+x_6)d_1}{x_1\sqrt{x_3{d}_1^2+x_4}}=\frac{x_2}{x_1}+\frac{x_3(1+x_6)d_1}{x_1\sqrt{x_3{d}_1^2+x_4}}---\left(10\right)$

Make (10) formula equal 0, obtain the d that in one section of interval, extreme value is corresponding ₁value be d _1y

$d_{1y}=\sqrt{\frac{x_4{x}_2^2}{{(x_3+x_3{x}_6)}^2-x_3{x}_2^2}}---\left(11\right)$

According to nU _inwith U _orelation, and the size solved, has three kinds of situations

I.nU _in≥U _o

Now, except x ₂outside <0, x ₁, x ₃~ x ₆all be greater than zero, the amount therefore in formula (7) inside radical sign is greater than zero automatically, and will ensure d ₂be greater than zero, d ₁must meet

$d_{1x1}=0\&le;d_1\&le;\sqrt{\frac{x_4}{(x_2^2-x_3)}}=d_{1z1}---\left(12\right)$

By d _1x1, d _1y, d _1z1substitution formula (9), getting amount corresponding to its maximum is d ₁value, then try to achieve d according to (7), (8) respectively ₂, d ₃value.

II.nU _in＜U _o

Now, except x ₁, x ₂, x ₄, x ₆all be less than 0, x ₃, x ₅be greater than zero, therefore except d will be ensured ₂be greater than outside zero, also will ensure that the amount in formula (7) inside radical sign is greater than zero, must meet

$d_{1x2}=\sqrt{\frac{-x_4}{x_3}}\&le;d_1\&le;\sqrt{\frac{-x_4}{(x_3-x_2^2)}}=d_{1z2}---\left(13\right)$

By d _1x2, d _1y, d _1z2substitution formula (9), getting amount corresponding to y maximum is d ₁value, then try to achieve d according to (7), (8) respectively ₂, d ₃value.

Also has a kind of possibility, when substitution data calculate d ₁+ d ₂+ d ₃during >1, illustrate that needing to become switching frequency could realize, for realizing constant frequency, must by d ₁+ d ₂+ d ₃value be limited in 1, and using this restriction as solving d ₁condition;

III.d ₁+d ₂+d ₃＝1

Now

d ₃＝1-d ₁-d ₂(14)

In the case, formula (4) still meets, and (4) are substituted into (14),

$d_2=\frac{U_o-({\mathrm{nU}}_{in}+U_o)d_1}{{\mathrm{nU}}_{in}}---\left(15\right)$

Ensure d ₂>0, then meet

$d_1\&le;\frac{U_o}{{\mathrm{nU}}_{in}+U_o}---\left(16\right)$

In the case, the relation of formula (6) still meets, and (15) are substituted into (6),

$(n^2{U}_{in}^2+{\mathrm{nU}}_{in}|U_G|+{U_o}^2)d_1^2-2{U_o}^2{d}_1+{U_o}^2-{\mathrm{nU}}_{in}{U_o}^2+\frac{4n^3{U}_{in}{\mathrm{Li}}^{*}}{T_S}=0---\left(17\right)$

According to the constraints of (16),

$d_1=\frac{-b-\sqrt{b^2-4ac}}{2a}---\left(18\right)$

Wherein

$a=n^2{U}_{in}^2+{\mathrm{nU}}_{in}{U}_o+{U_o}^2,b=-2{U_o}^2,c={U_o}^2-{\mathrm{nU}}_{in}{U}_o+\frac{4n^3{U}_{in}{\mathrm{LI}}_o}{T_S}$

Then d is tried to achieve according to (14), (15) ₂with d ₃;

According to above-mentioned three kinds of situations about calculating, the d obtained under finally choosing a kind of situation ₁, d ₂with d ₃as the modulation ratio of final front 3 mode, if d ₁+ d ₂+ d ₃=1, then d ₄=0; If d ₁+ d ₂+ d ₃<1, then d ₄=1-(d ₁+ d ₂+ d ₃), and by d ₁, d ₂, d ₃with d ₄as source signal, directly realized the drive singal of switching tube by modulation strategy.

Flow chart shown in Fig. 4 clearly describes above-mentioned solution procedure.Notice that load current reference current I* substitutes by the calculating in flow chart.Clear process shown in Fig. 4, can be realized by DSP easily.

Obtaining d ₁, d ₂, d ₃with d ₄, the drive singal of switching tube is directly realized by modulation strategy.Under the boost mode that the present invention adopts, control block diagram as shown in Figure 5.During decompression mode work, be all generally that adopt identical control block diagram, only during constant current charge, the benchmark of electric current loop is a constant to device chargings such as such as storage batterys, and in Figure 5, reference current I _o* be amplitude.Can find out, compared to traditional control method, institute of the present invention extracting method saves current sensor, and electric current loop does not need current feedback amount, directly modulation ratio is obtained by inventing the Mathematical Modeling set up, no longer need to set up more accurate feedback quantity by transducer and current prediction method, which greatly enhances the dynamic property of system.

Parameters design

In DAB-BDC shown in Fig. 1, have two crucial parameters to need to be designed, one is the no-load voltage ratio n of transformer, and another is buffer inductance L.Because DAB-BDC is when input voltage, conversion are equal to the output voltage on former limit, the current stress of device is minimum, the fluctuation range of input, output voltage can be considered, when the median of voltage fluctuation, make input voltage, conversion equal to the output voltage on former limit.

DAB-BDC mono-is typically applied as in direct-current grid and carries out discharge and recharge to energy storage battery, and typical input, output voltage parameter are: U _inscope is at 42-56V, U _oscope is at 360-400V, and the present embodiment designs peak power output 500W.Be described by means of the design of this parameter to inductance L.

The voltage placed in the middle of input, output voltage is respectively 49V and 380V, therefore, first determines transformer voltage ratio n=380/49=7.75.And the design of inductance L is without there being multiple method, document [1] just proposes two kinds of methods for designing, and a kind of angle from reducing reactive power designs, and the method needs inductance value should be little as far as possible; The another kind of scope from increasing device ZVS designs, and this method needs inductance value large as much as possible when ensureing rated output power, two kinds of methods respectively have pluses and minuses.The present embodiment is from inductive current stress (linear with switching tube electric current) minimum design inductance value.

Y according to formula (2), (3) ₁, y ₂value, row to write in Fig. 2 the function of electric current within 3 stages in half period, try to achieve inductive current i according to institute's array function _leffective value I _lfor

$I_L=\sqrt{\frac{y_1^2(d_1+d_2)+y_2^2(d_2+d_3)+y_1{y}_2{d}_2}{3}}---\left(19\right)$

Calculate I _lthree sections of modulation ratio d must be known ₁, d ₂, d ₃, but calculate d ₁, d ₂, d ₃process can not solve by complete, a continuous print expression formula, under the prerequisite not affecting computational accuracy, the present embodiment adopts matlab/simulink to construct Simulation Calculation shown in flow chart, obtains the d in different input voltage, different induction value situation ₁, d ₂, d ₃value, as shown in table 3, as can be seen from Table 3: in not all situation, all experience the mode of 4 shown in Fig. 3, in, underloading situation low at input voltage, only have front 3 mode, corresponding with the AC boosting situation shown in Fig. 2 (c); In, underloading situation high at input voltage, only there are rear 3 mode, the ac buck situation namely shown in corresponding diagram 2 (a).But in most of the cases, d ₁+ d ₂+ d ₃=1, this contributes to the current stress reducing switching tube.Modulation ratio data in corresponding table 3 and formula (19), obtain in different capacity, different induction situation, I _lwith the change curve of input voltage, as shown in Figure 6.Can find out, in whole input voltage range, during underloading, larger inductance (10 μ H inductance in such as figure) can make I _lless; But along with the change of power is large, the inductive current effective value I that larger inductance produces _lincrease very fast, and much larger than situation during other inductance value; Under equal-wattage, too small inductance (2.5 μ H inductance in such as figure) changes acutely in whole input voltage range, and I _lvalue also very large.Therefore inductance value should be chosen moderate (5 μ H in such as figure and 7.5 μ H), and on the one hand, in whole change range of input voltage, during equal-wattage, device current stress can not change too violent; On the other hand, in underloading in rated load ranges, when the current stress relative maximum inductance value of device and extra low inductance value, be all in comparatively Optimal State.Consider above-mentioned situation, choose L=6 μ H.

The high-frequency ac buck control strategy that table 3 is unified

It should be noted that: in the design of document [1], the scope of device ZVS and the suppression of reactive power are competing, and larger device ZVS area requirement inductance value is the bigger the better; But for suppressing reactive power, inductance value is the smaller the better, and this is conflict.But in the inventive method, choose moderate inductance value and can realize current stress optimization on the one hand, on the other hand, moderate inductance also can ensure wider range of device ZVS.

With comparing of two phase shifting control strategy current stress

In the improvement control strategy of numerous DAB-BDC, two-track phase (the Dual-Phase-Shift that document [2] proposes, DPS) control strategy is more representative, the control schematic diagram proposed as shown in Figure 7, first the modulation ratio arranging transformer forward and backward level full-bridge circuit is equal, the D namely in figure ₁realize; On this basis, then high-pressure side, low-pressure side full-bridge are carried out phase shift, the D namely in figure ₂realize.

The parameter of converter is same as described above, the method solving controlled quentity controlled variable utilizing document [2] to provide, and obtains one group of (D ₁, D ₂) data, as shown in table 4, according to these data, when obtaining different induction, inductive current effective value with input voltage change curve, as Fig. 8, in from underloading to full-load range, during different induction value, size of current is inconsistent, and during underloading, inductance value can realize less current stress more greatly; Full load, situation is just in time contrary.

Consider comparatively moderate inductance value, finally get L=10 μ H.

Table 4DPS control strategy

A. current stress compares

For directly being compared with document [2] institute extracting method by institute of the present invention extracting method, Fig. 9 gives the current stress of two kinds of control methods under different capacity.Can find out, at full load, the present invention puies forward high-frequency ac buck and has clear superiority; When underloading, the current stress of institute's extracting method of the present invention and DPS control method is in staggered situation.Generally, institute's extracting method current stress of the present invention slightly advantage.

B. to the excavation of control object potentiality

During DPS shown in Fig. 7 controls, the AC voltage width of high-pressure side, low-pressure side full-bridge circuit is designed to equal, therefore, if increase the pulse-width modulation of high-pressure side full-bridge, switching device current stress in converter will be optimized further, but calculate the complexity of the Mathematical Modeling of phase shift time by increase at double, need the situation of consideration also just more, calculate three phase shift times almost impossible.And the present invention's carry is based in high-frequency ac buck control method, there is 4 control objects, is respectively the d shown in Fig. 2 ₁, d ₂, d ₃, d ₄, according to bound variable, find out (the d wherein had most ₁, d ₂, d ₃, d ₄) combination.Should say in the excavation to control object potentiality, the present invention carry control strategy based on high-frequency ac buck due to DPS control strategy.

C. Sofe Switch ability compares

Load large to a certain extent time, in the DAB-BDC that DPS controls, all devices can realize ZVS, but when underloading, the energy shortage stored in inductance is to take the electric charge stored in two switch junctions electric capacity away, therefore, during underloading, DPS method can not realize the Sofe Switch of device ^[1].Inductive current design in DCM or BCM state, therefore, when beginning and the inductive current vanishing of half period, can be ensured that device realizes ZCS switch, and no matter be in underloading or full load condition by institute of the present invention extracting method.This feature can the effective efficiency of Lifting Transform device when underloading.

Claims (1)

1., based on a DAB-BDC modulator approach for high-frequency ac buck principle, in DAB-BDC, input, output voltage are U _in, U _o, then buffer inductance both sides voltage is

$u_L={\mathrm{AU}}_{in}-B\frac{U_o}{n}---\left(1\right)$

In formula, A, B are respectively the state of inductance both sides current potential, and n is transformer voltage ratio;

It is characterized in that, the control strategy of DAB-BDC is as follows:

The positive and negative periodic symmetry of high frequency, every half period comprises 4 mode, and the modulation ratio of 4 mode is respectively d ₁, d ₂, d ₃, d ₄; In positive half period, the value that under 4 mode, the value of A is respectively 1,1,0,0, B is respectively 0,1,1,0; In negative half-cycle, the value that under 4 mode, the value of A is respectively-1 ,-1,0,0, B is respectively 0 ,-1 ,-1,0;

According to the situation of input voltage, output voltage and power output, above-mentioned all 4 mode in half switch periods, may not be comprised.

Make T _sfor switch periods, y ₁, y ₂be respectively first mode finish time and second mode finish time buffer inductance current i _lvalue; According to the time acted on buffer inductance and voltage, determine

$y_1=\frac{U_{in}{d}_1{T}_S}{2L}---\left(2\right)$

$y_2=\frac{{\mathrm{nU}}_{in}{T}_S{d}_1+({\mathrm{nU}}_{in}-U_o)T_S{d}_2}{2nL}---\left(3\right)$

In formula, L is the inductance value of buffer inductance, equal with slippage according to inductive current ascending amount, obtains

$d_3=\frac{{\mathrm{nU}}_{in}{d}_1+({\mathrm{nU}}_{in}-U_o)d_2}{U_o}---\left(4\right)$

Primary side current of transformer i _l, only have the electric current of second mode and the 3rd mode time period to flow to load-side, the mean value of electric current after transformer conversion of these two mode equals load current,

$d_2^2({\mathrm{nU}}_{in}-U_o)+d_2(2{\mathrm{nd}}_1{U}_{in}+{\mathrm{nd}}_3{U}_{in}-d_3{U}_o)+{\mathrm{nd}}_1{d}_3{U}_{in}-4n^2{\mathrm{LI}}_o/T_S=0---\left(5\right)$

(4) are substituted into (5),

$d_2^2\left({\mathrm{nU}}_{in}\right({\mathrm{nU}}_{in}-U_o\left)\right)+d_2\left(2n^2{U}_{in}^2{d}_1\right)+n^2{U}_{in}^2{d}_1^2-4n^2{\mathrm{LU}}_o{I}_o/T_S=0---\left(6\right)$

Solve

$d_2=\frac{x_2{d}_1+\sqrt{x_3{d}_1^2+x_4}}{x_1}---\left(7\right)$

d ₃＝x ₅d ₁+x ₆d ₂(8)

Wherein

$\begin{array}{ccc}{x}_1=U_{in}({\mathrm{nU}}_{in}-U_o);& x_2=-{\mathrm{nU}}_{in}^2;& x_3={\mathrm{nU}}_{in}^3{U}_o\end{array};$

$\begin{array}{ccc}{x}_4=\frac{4{\mathrm{nU}}_{in}{U}_o{\mathrm{LI}}_o({\mathrm{nU}}_{in}-U_o)}{T_S};& x_5=\frac{{\mathrm{nU}}_{in}}{U_o};& x_6=\frac{({\mathrm{nU}}_{in}-U_o)}{U_o}\end{array}$

Only when second mode and the 3rd mode, inductive current i _ljust load-side can be flow to.Therefore, if d ₂+ d ₃value comparatively large, then when same load current, electric current is just comparatively level and smooth, and the current effective value that switching device bears is also less; Make y=d ₂+ d ₃, then d when asking y to obtain maximum ₁, d ₂, d ₃value, therefore just require below the maximum of y; By the expression formula of (7), (8) substitution y,

$y=x_5{d}_1+(1+x_6)\frac{x_2{d}_1+\sqrt{x_3{d}_1^2+x_4}}{x_1}---\left(9\right)$

D is asked to (9) ₁derivative,

$\frac{dy}{d\left(d_1\right)}=x_5+\frac{x_2(1+x_6)}{x_1}+\frac{x_3(1+x_6)d_1}{x_1\sqrt{x_3{d}_1^2+x_4}}=\frac{x_2}{x_1}+\frac{x_3(1+x_6)d_1}{x_1\sqrt{x_3{d}_1^2+x_4}}---\left(10\right)$

(10) formula is made to equal 0, when obtaining that in one section of interval, y obtains extreme value, d ₁the value of correspondence be d _1y

$d_{1y}=\sqrt{\frac{x_4{x}_2^2}{{(x_3+x_3{x}_6)}^2-x_3{x}_2^2}}---\left(11\right)$

According to nU _inwith U _orelation, and the size solved, has three kinds of situations

I.nU _in≥U _o

Now, except x ₂outside <0, x ₁, x ₃~ x ₆all be greater than zero, the amount therefore in formula (7) inside radical sign is greater than zero automatically, and will ensure d ₂be greater than zero, d ₁must meet

$d_{1x1}=0\&le;d_1\&le;\sqrt{\frac{x_4}{(x_2^2-x_3)}}=d_{1z1}---\left(12\right)$

By d _1x1, d _1y, d _1z1substitution formula (9), getting amount corresponding to its maximum is d ₁value, then try to achieve d according to (7), (8) respectively ₂, d ₃value.

II.nU _in＜U _o

Now, except x ₁, x ₂, x ₄, x ₆all be less than 0, x ₃, x ₅be greater than zero, therefore except d will be ensured ₂be greater than outside zero, also will ensure that the amount in formula (7) inside radical sign is greater than zero, must meet

$d_{1x2}=\sqrt{\frac{-x_4}{x_3}}\&le;d_1\&le;\sqrt{\frac{-x_4}{(x_3-x_2^2)}}=d_{1z2}---\left(13\right)$

By d _1x2, d _1y, d _1z2substitution formula (9), getting amount corresponding to its maximum is d ₁value, then try to achieve d according to (7), (8) respectively ₂, d ₃value.

Also has a kind of possibility, when substitution data calculate d ₁+ d ₂+ d ₃during >1, illustrate that needing to become switching frequency could realize, for realizing constant frequency, must by d ₁+ d ₂+ d ₃value be limited in 1, and using this restriction as solving d ₁condition;

III.d ₁+d ₂+d ₃＝1

Now

d ₃＝1-d ₁-d ₂(14)

In the case, formula (4) still meets, and (4) are substituted into (14),

$d_2=\frac{U_o-({\mathrm{nU}}_{in}+U_o)d_1}{{\mathrm{nU}}_{in}}---\left(15\right)$

Ensure d ₂>0, then meet

$d_1\&le;\frac{U_o}{{\mathrm{nU}}_{in}+U_o}---\left(16\right)$

In the case, the relation of formula (6) still meets, and (15) are substituted into (6),

$(n^2{U}_{in}^2+{\mathrm{nU}}_{in}|u_G|+{U_o}^2)d_1^2-2{U_o}^2{d}_1+{U_o}^2-{\mathrm{nU}}_{in}{U_o}^2+\frac{4n^3{U}_{in}{\mathrm{Li}}^{*}}{T_S}=0---\left(17\right)$

According to the constraints of (16),

$d_1=\frac{-b-\sqrt{b^2-4ac}}{2a}---\left(18\right)$

Wherein

$\begin{array}{ccc}a=n^2{U}_{in}^2+{\mathrm{nU}}_{in}{U}_o+{U_o}^2,& b=-2{U_o}^2,& c={U_o}^2-{\mathrm{nU}}_{in}{U}_o+\frac{4n^3{U}_{in}{\mathrm{LI}}_o}{T_S}\end{array}$

Then d is tried to achieve according to (14), (15) ₂with d ₃;

According to above-mentioned three kinds of situations about calculating, the d obtained under finally choosing a kind of situation ₁, d ₂with d ₃as the modulation ratio of final front 3 mode, if d ₁+ d ₂+ d ₃=1, then d ₄=0; If d ₁+ d ₂+ d ₃<1, then d ₄=1-(d ₁+ d ₂+ d ₃), and by d ₁, d ₂, d ₃with d ₄as source signal, directly realized the drive singal of switching tube by modulation strategy.