CN106533206A - Neutral-point voltage balance control method of back-to-back three-level double-PWM converter - Google Patents

Abstract

The invention discloses a neutral-point voltage balance control method of a back-to-back three-level double-PWM converter. The voltage space vectors positioned in other sections are normalized to a first section for unified calculation through mathematic coordinate rotation transformation, so that distinguishing calculation treatment is not needed for each section, and the algorithmic complexity level is simplified. Meanwhile, the method sufficiently uses a neutral-point voltage control capacity of a grid-side converter or a machine-side converter, when the control capacity on one side is insufficient, the converter on the other side can be used for providing the control capacity, and the defect that the control action of neutral-point voltage is mutually weakened possibly caused by the fact that the neutral-point voltage is controlled by converters on two sides independently in a traditional method is avoided, so that the fluctuation of the neutral-point voltage of the back-to-back double-PWM converter is reduced, and effective balance control for the neutral-point voltage is realized.

Description

The neutral-point voltage balance method of the double pwm converters of back-to-back three level

Technical field

The present invention is relevant with three-level converter, particularly belongs to a kind of midpoint electricity of the double pwm converters of back-to-back three level Pressure balance control method.

Background technology

With the development of Power Electronic Technique, three-level converter is with its higher delivery efficiency and relatively low output current The advantages of percent harmonic distortion, obtain in fields such as big-power transducer, three-phase grid-connected inverter and three-phase uninterrupted power supplys It is widely applied.

At present, field and wind power generation field is driven to occur in that in heavy-duty motor back-to-back based on three-level topology The double pwm converters of three level, the changer include the pusher side conversion of the grid side converter for connecting electrical network and connection motor Device, and grid side converter and machine-side converter all have the ability of balance mid-point voltage.

At present, it is in disclosed technical literature to adopt an independent side converter to control mid-point voltage more, such as only with net side Changer is controlling mid-point voltage, although this method can realize the balance control of mid-point voltage, but as net side becomes The modulation degree of parallel operation is higher, therefore the ability of its balance mid-point voltage is than relatively limited.

Additionally, also one technology path is to comprehensively utilize the control ability of two side converter alignment voltages, though So grid side converter and machine-side converter are respectively adopted independent mid-point voltage control strategy and can strengthen centering point voltage Control ability, but due to specifically do not distinguish in each switch periods the voltage-controlled size of two side converter alignments and Direction, therefore the method has the control action of two side converter alignment voltages and is likely to occur overcompensation and cancels out each other Problem.

Referring to《The double pwm converter capacitor voltage balance Comprehensive Controls of three level》, Fan Bishuan, Tan Guanzheng, Fan Shaosheng, Electric Machines and Control, 2014,18 (1)：38-43, a kind of grid side converter of paper proposition and machine-side converter Electric capacity neutral point voltage balance integrated control strategy, analyzes the shadow of all redundancy small vector centering point voltages of three-level converter Ring, give voltage vector positioned at each sector when midpoint average current calculating formula, by each phase current of subsequent time With the prediction of capacitance voltage reference quantity, it is the minimum merit functions of mid-point voltage fluctuation to obtain a target, superfluous by adjusting Complement vector time distribution factor reaches the target of control mid-point voltage fluctuation.Although the method does not result in two side converters pair The overcompensation of mid-point voltage and undercompensation, but which distributes identical redundant vectors time factor to two side converters, still The control ability of two side converter alignment voltages is limited to a certain extent, and the method needs constantly iteration fortune Calculate, algorithm is realized more complicated.Therefore, in place of for the deficiencies in the prior art, this area is urgently proposed in a kind of algorithm Realize simply can effectively realizing the neutral-point voltage balance method of the double pwm converters of back-to-back three level.

The content of the invention

The technical problem to be solved in the present invention is to provide a kind of neutral point voltage balance of the double pwm converters of back-to-back three level Control method, can not only avoid two side converter independent control mid-point voltages produce control action mutually weaken lack Point, and two side converters can be overcome fully to control the problem of mid-point voltage.

In order to solve the above problems, the neutral point voltage balance control of the double pwm converters of back-to-back three level that the present invention is provided Method processed, comprises the following steps：

The voltage vector of the voltage vector and machine-side converter of grid side converter is converted respectively by step 1 using coordinate transform To two-phase rest frame, and rower change is entered according to DC voltage value to two voltage vectors；

Step 2, according to the value of grid side converter voltage vector and machine-side converter voltage vector in two-phase rest frame Judge the actual sector number that two voltage vectors are located；

Step 3, will mark the grid side converter voltage vector and machine-side converter voltage vector changed using transformation matrix of coordinates It is converted into the first sector；

According to the grid side converter voltage vector after rotation transformation and machine-side converter voltage vector, step 4, judges that rotation becomes The little sector triangle numbering in the first sector that two voltage vectors after changing are located；

Step 5, according to the different little sector triangle that grid side converter voltage vector is located with machine-side converter voltage vector, Three fundamental voltage space vectors and corresponding midpoint electric current of grid side converter and machine-side converter are determined respectively；

Step 6, according to the different little sector triangle that grid side converter voltage vector is located with machine-side converter voltage vector, The action time of three fundamental voltage space vectors that calculating is acted on successively；

Step 7, three be each input into according to the deviant of DC side mid-point voltage and grid side converter and machine-side converter Phase current values, calculate the neutral point voltage balance factor of grid side converter and machine-side converter side respectively；

Step 8, according to grid side converter and the machine-side converter each action time of three fundamental voltage space vectors and meter The neutral point voltage balance factor for obtaining, calculates the output pulse duration for obtaining that vector is located at different little sector trianglees respectively Than；

Step 9, according to grid side converter voltage vector and the actual sector number of machine-side converter voltage vector, respectively by net Side converter and machine-side converter map back grid side converter voltage vector in the calculated duty cycle information in the first sector The actual sector being each located with machine-side converter voltage vector, and duty cycle information is scaled corresponding fiducial value to produce Raw pwm pulse drives grid side converter and machine-side converter respectively.

The invention has benefit that：

1) space vector of voltage positioned at other sectors is normalized to the first sector by mathematical coordinates rotation transformation by the present invention Unified calculation, therefore need not carry out distinguishing calculating process for each sector, so as to simplify the complexity of algorithm；

2) present invention makes full use of the mid-point voltage control ability of grid side converter or machine-side converter, when wherein side Control ability can not be provided completely by another side converter, it is to avoid two side converter independent control midpoints electricity in traditional method The shortcoming that the mid-point voltage control action that pressure may be brought mutually weakens, so that reduce back-to-back double pwm converters Mid-point voltage fluctuates, and realizes the active balance control of centering point voltage.

Description of the drawings

Fig. 1 is the circuit diagram of the double pwm converters of back-to-back three level that the inventive method is suitable for；

Fig. 2 is the region division schematic diagram of space vector of voltage in the inventive method；

Fig. 3 is the first sector voltage vector relation in the inventive method and the division schematic diagram between delta；

Fig. 4 a to Fig. 4 f are that the fundamental voltage space vector effect in the inventive method in each delta in the first sector is shown It is intended to.

Fig. 5 is the algorithm flow schematic diagram in the inventive method.

Specific embodiment

The present invention is further detailed explanation with specific embodiment below in conjunction with the accompanying drawings.

The present invention is applied to the double pwm converters of back-to-back three level, and the changer includes grid side converter (hereinafter referred to as For GVSC) and machine-side converter (hereinafter referred to as MVSC), as shown in figure 1, neutral-point voltage balance method Comprise the following steps：

S1：Using Clarke coordinate transform respectively by the three-phase voltage vector of three-phase voltage vector MVSC of GVSC Two-phase rest frame is transformed to, and rower change is entered according to DC voltage value to two voltage vectors, be divided into：

S11：GVSC voltage vectors are transformed to into two-phase rest frame, and two-phase voltage is sweared according to DC voltage value Measure into rower and change, comprise the following steps that：

(1) voltage vector is transformed under two-phase rest frame using following formula；

In formula, u_ag、u_bgAnd u_cgRespectively component of the GVSC voltage vectors in abc coordinate systems；u_αg、u_βgRespectively For component of the GVSC voltage vectors in α β coordinate systems；

(2) amplitude limiting processing is carried out to two-phase voltage vector according to DC voltage value；

IfThen

IfThen

In formula, u_dcFor DC voltage value；u_αgL、u_βgLRespectively GVSC voltage vectors after amplitude limiting processing in α β Component in coordinate system；

(3) rower change is entered according to DC voltage value to two-phase voltage vector；

In formula, v_αg、v_βgThe respectively per unit value of component of the GVSC voltage vectors in α β coordinate systems.

S12：MVSC voltage vectors are transformed to into two-phase rest frame, and two-phase voltage is sweared according to DC voltage value Measure into rower and change, comprise the following steps that：

(1) voltage vector is transformed under two-phase rest frame using following formula；

In formula, u_am、u_bmAnd u_cmRespectively component of the MVSC voltage vectors in abc coordinate systems；u_αm、u_βmPoint Wei not component of the MVSC voltage vectors in α β coordinate systems；

(2) amplitude limiting processing is carried out to two-phase voltage vector according to DC voltage value；

IfThen

IfThen

In formula, u_dcFor DC voltage value；u_αmL、u_βmLRespectively MVSC voltage vectors after amplitude limiting processing in α β Component in coordinate system；

(3) rower change is entered according to DC voltage value to two-phase voltage vector；

In formula, v_αm、v_βmThe respectively per unit value of component of the MVSC voltage vectors in α β coordinate systems.

It should be noted that the related physical quantity of grid side converter GVSC contains subfix in symbol in subsequent step The related physical quantity of g, machine-side converter MVSC contains subfix m in symbol, illustrates hereby.

S2：Component value according to GVSC voltage vectors and MVSC voltage vectors in two-phase rest frame judges electricity Actual sector of the pressure vector in the regular hexagon formed by three level fundamental voltage space vectors shown in Fig. 2 number (according to Regular hexagon is divided into into 6 sector trianglees clockwise, is numbered respectively from 1 to 6), it is divided into：

S21：The reality that voltage vector is located is judged according to component value of the GVSC voltage vectors in two-phase rest frame Sector number, concrete grammar are as follows：

(1) in order to judge the sector number at GVSC voltage vectors place, three variable a are defined first, b, c are simultaneously utilized Equation below calculates the value of a, b and c；

(2) defined variable P, its value are calculated according to following formula：

P=sign (a)+2sign (b)+4sign (c)

In formula, sign () is defined as follows shown for seeking data symbol function：

(3) the actual sector N that voltage vector is located is judged according to the value of following table and P；

S22：The sector number that voltage vector is located is judged according to MVSC voltage vector values, concrete grammar is identical with S21.

S3：The GVSC voltage vectors and MVSC voltage vectors of marking change are converted into into first using transformation matrix of coordinates Sector, is divided into：

S31：The GVSC voltage vectors for marking change are converted into into the first sector using transformation matrix of coordinates, concrete grammar is such as Under：

The voltage vector unification of the first sector to the 6th sector is transformed to into the first sector using following coordinate transform formula：

In formula, v_αTg、v_βTgRespectively GVSC voltage vectors after coordinate transform the first sector component；T_N1For The transformation matrix of coordinates of N sectors to the first sector, N values are from 1 to 6；Wherein：

1) when voltage vector is located at the first sector (delta-shaped region surrounded by vector 000, PNN and PPN in Fig. 2) When, corresponding transformation matrix of coordinates is

2) when voltage vector is located at the second sector (delta-shaped region surrounded by vector 000, PPN and NPN in Fig. 2) When, corresponding transformation matrix of coordinates is

3) when voltage vector is located at the 3rd sector (delta-shaped region surrounded by vector 000, NPN and NPP in Fig. 2) When, corresponding transformation matrix of coordinates is

4) when voltage vector is located at the 4th sector (delta-shaped region surrounded by vector 000, NPP and NNP in Fig. 2) When, corresponding transformation matrix of coordinates is

5) when voltage vector is located at the 5th sector (delta-shaped region surrounded by vector 000, NNP and PNP in Fig. 2) When, corresponding transformation matrix of coordinates is

6) when voltage vector is located at the 6th sector (delta-shaped region surrounded by vector 000, PNP and PNN in Fig. 2) When, corresponding transformation matrix of coordinates is

S32：The MVSC voltage vectors changed will be marked using transformation matrix of coordinates and are converted into the first sector, concrete grammar with S31 is identical.

S4：First sector triangle is by U₀～U₆Be divided into 6 little sector gables, numbering be followed successively by A, B, C, D, E and F, wherein U₀Refer to zero vector 000, NNN and PPP (these three vectors coincidences, i.e. α β coordinate systems Round dot), U₁Refer to small vector 0NN and P00 (the two vectors overlap), U₂Refer to small vector PP0 and 00N (the two vectors overlap), U₃Refer to big vector PNN, U₄Refer to middle vector P0N, U₅Refer to big vector PPN, U₆For middle vector U₄Half, the zero vector, small vector, the big vector of middle vector be three-level converter The professional technique noun of space vector modulation；

As shown in figure 3, by U between A deltas₀、U₁And U₆Surround, by U between B deltas₀、U₆And U₂Institute Surround, by U between C deltas₁、U₄And U₆Surrounded, by U between D deltas₂、U₄And U₆Surrounded, E By U between delta₁、U₃And U₄Surrounded, by U between F deltas₂、U₄And U₅Surrounded；

Voltage vector after rotation transformation is judged according to the GVSC voltage vectors and MVSC voltage vectors after rotation transformation Number between the little sector delta that the first sector is located, be divided into：

S41：Judge the numbering between the delta that the GVSC voltage vectors after rotation transformation are located, be specifically divided into as follows Situation：

1) Rule of judgment that the GVSC voltage vectors after rotation transformation are located between A deltas is：

2) Rule of judgment that the GVSC voltage vectors after rotation transformation are located between B deltas is：

3) Rule of judgment that the GVSC voltage vectors after rotation transformation are located between C deltas is：

4) Rule of judgment that the GVSC voltage vectors after rotation transformation are located between D deltas is：

5) Rule of judgment that the GVSC voltage vectors after rotation transformation are located between E deltas is：

6) Rule of judgment that the GVSC voltage vectors after rotation transformation are located between F deltas is：

S42：Judge the numbering between the delta that the MVSC voltage vectors after rotation transformation are located, concrete grammar and S41 It is identical.

S5：GVSC is determined respectively between the different deltas according to GVSC voltage vectors and MVSC voltage vectors place With tri- fundamental voltage space vectors of MVSC and midpoint electric current, it is divided into：

S51：Determine tri- fundamental voltage space vectors of GVSC and midpoint electric current, comprise the following steps that：

(1) the three fundamental voltage space vectors for acting on successively are set and is respectively v_1g, v_2gAnd v_3g, then the six of the first sector Between individual little delta, corresponding three fundamental voltage space vectors are carried out according to the following table selection.

(2) set the corresponding midpoint electric current of three fundamental voltage space vectors for acting on successively and be respectively i_1g, i_2gAnd i_3g, then The corresponding midpoint electric current of three fundamental voltage space vectors between the first little delta in sector six is selected according to following table.

When (3) the 2nd～6 sectors transform to 1 sector, as three fundamental voltage space vectors of correspondence position occur Change, thus corresponding midpoint electric current is also differed, therefore the actual sector being located according to GVSC voltage vectors will be basic The corresponding midpoint electric current of space vector of voltage is also carried out corresponding conversion, and concrete change situation is as shown in the table.

In table, i_aNg, i_bNgAnd i_cNgIt is the corresponding three-phase current of GVSC voltage vectors positioned at n-th sector.

S52：Determine tri- fundamental voltage space vectors of GVSC and midpoint electric current, concrete grammar is identical with S51.

S6：Acted on according to the different triangle interval computations that GVSC voltage vectors and MVSC voltage vectors are located successively Three fundamental voltage space vectors action time, be divided into：

S61：Three fundamental voltage skies that the different triangle interval computations being located according to GVSC voltage vectors are acted on successively Between vector action time, be specifically divided into：

1) between A deltas, the effect schematic diagram of three fundamental voltage space vectors as shown in fig. 4 a, action time Respectively：

Wherein, t_1gFor the action time of vector 0NN；t_2gFor the action time of vector 00N；t_3gFor the work of vector 000 With time, T_sFor switch periods；

2) between B deltas, the effect schematic diagram of three fundamental voltage space vectors as shown in Figure 4 b, action time Respectively：

Wherein, t_1gFor the action time of vector 00N；t_2gFor the action time of vector 000；t_3gFor the work of vector P00 With time, T_sFor switch periods；

3) between C deltas, the effect schematic diagram of three fundamental voltage space vectors as illustrated in fig. 4 c, action time Respectively：

Wherein, t_1gFor the action time of vector 0NN；t_2gFor the action time of vector 00N；t_3gFor the work of vector P0N With time, T_sFor switch periods；

4) between D deltas, the effect schematic diagram of three fundamental voltage space vectors as shown in figure 4d, action time Respectively：

Wherein, t_1gFor the action time of vector 00N；t_2gFor the action time of vector P0N；t_3gFor the work of vector P00 With time, T_sFor switch periods；

5) between E deltas, the effect schematic diagram of three fundamental voltage space vectors as shown in fig 4e, action time Respectively：

Wherein, t_1gFor the action time of vector 0NN；t_2gFor the action time of vector PNN；t_3gFor vector P0N's Action time, T_sFor switch periods；

6) between F deltas, the effect schematic diagram of three fundamental voltage space vectors as shown in fig. 4f, action time Respectively：

Wherein, t_1gFor the action time of vector 00N；t_2gFor the action time of vector P0N；t_3gFor the work of vector PPN With time, T_sFor switch periods；

S62：Three fundamental voltage skies that the different triangle interval computations being located according to MVSC voltage vectors are acted on successively Between vector action time, concrete grammar is identical with S61.

It should be noted that the PWM carrier waves of GVSC and MVSC be taken as it is same.

S7：Calculated according to the deviant and GVSC and MVSC inputs three-phase electricity flow valuve of DC side mid-point voltage respectively The respective neutral point voltage balance factor in GVSC and MVSC sides, comprises the following steps that：

(1) assume to need the mid-point voltage for adjusting to be u_adj, then regulating time distribution factor k calculated according to the following formula：

In formula, u_adj=-u_N0=u_dc1-u_dc2, C_dcFor the size of dc-link capacitance, u_N0For mid-point voltage value, u_dc1 And u_dc2Respectively direct current bus bar holds magnitude of voltage and lower capacitance voltage value；

(2) two neutral point voltage balance factors k are set_gAnd k_mTo redistribute grid side converter and machine-side converter first The action time of fundamental voltage space vector；

Current midpoint voltage is u_N0, then the maximum balance factor computing formula of the neutral-point voltage balance of grid side converter For：

The maximum balance factor computing formula of the neutral-point voltage balance of machine-side converter is：

(3) constraint of the neutral point voltage balance factor is calculated

Take k_g=k₁, k_m=k₂, k is corrected according to following rule_gAnd k_mValue, it is specific as follows shown in：

If 1) abs (k₁)≤1, abs (k₂)≤1, then make k_m=0, k_gKeep constant；

If 2) abs (k₁)≤1, abs (k₂)>1, then make k_m=0, k_gKeep constant；

If 3) abs (k₁)>1, abs (k₂)≤1, then make k_g=0, k_mKeep constant；

If 4) abs (k₁)>1, abs (k₂)>1, then make k_g=sign (k_g), k_mCalculated by below equation：

Wherein u_adjGFor the mid-point voltage value that net side current transformer can be adjusted, its computing formula is as follows：

S8：According to GVSC and MVSC each the action time of three fundamental space vectors and calculated balance because Subvalue, calculates output pulse duty factor, is divided into：

S81：According to action time and the calculated balance factor value of tri- fundamental space vectors of GVSC, calculate defeated Go out pulse duty factor, specifically include：

1) three variable t are defined_ag, t_bgAnd t_cg, its value is respectively：

Define three variable d_a, d_bAnd d_cAs three dutyfactor values, it is respectively according to the following formula

2) according between the different deltas at GVSC voltage vectors place, select the effective duty cycle d of three-phase bridge arm_Ag、 d_BgAnd d_Cg, concrete condition is as shown in the table.

S82：According to action time and the calculated balance factor value of tri- fundamental space vectors of MVSC, calculate defeated Go out pulse duty factor, concrete grammar is identical with S81.

S9：To be calculated in the first sector according to the actual sector number of GVSC voltage vectors and MVSC voltage vectors Duty cycle information map back the original sector at respective place, and dutycycle is scaled into corresponding fiducial value, to produce Pwm pulse drives GVSC and MVSC respectively, is divided into：

S91：Tabled look-up in the calculated duty cycle information in the first sector according to the actual sector number of GVSC voltage vectors The original sector at GVSC places is mapped back, and dutycycle is scaled into corresponding fiducial value, to produce pwm pulse point GVSC is not driven, is specifically included：

(1) actual sector being located according to GVSC voltage vectors, determines two groups of duty cycle informations of three-phase bridge arm；

If A, C and E that the GVSC voltage vectors after rotation transformation are located at the first sector is interval, bridge arm dutycycle Value is as shown in the table；

If B, D and F that the GVSC voltage vectors after rotation transformation are located at the first sector is interval, bridge arm dutycycle Value is as shown in the table；

(2) value of tri- bridge arm six modulating waves of correspondence of GVSC is determined using following formula, it is specific as follows shown：

In formula, CP_A1g、CP_A2gThe respectively value of two modulating waves of A phases；CP_B1g、CP_B2gRespectively B phases two The value of modulating wave；CP_C1g、CP_C2gThe respectively value of two modulating waves of C phases；Peak values of the CNT for triangular carrier.

S92：Tabled look-up in the calculated duty cycle information in the first sector according to the actual sector number of MVSC voltage vectors The original sector at MVSC places is mapped back, and dutycycle is scaled into corresponding fiducial value, to produce pwm pulse point MVSC is not driven, concrete grammar is identical with S91.

The schematic flow sheet of the method for the invention is as shown in Figure 5.

This method avoid what grid side converter and machine-side converter independent control mid-point voltage in traditional method may bring The shortcoming that mid-point voltage control mutually weakens, can make full use of the mid-point voltage control of grid side converter and machine-side converter Ability, so as to reduce the fluctuation of mid-point voltage, realizes the balance control of mid-point voltage.

The present invention is described in detail above by specific embodiment, above-described embodiment is only the preferable of the present invention Embodiment, the invention is not limited in above-mentioned embodiment.Without departing from the principles of the present invention, the skill of this area Equivalent replacement and improvement that art personnel make, are regarded as in the technology category protected of the invention.

Claims (10)

1. a kind of neutral-point voltage balance method of the double pwm converter of back-to-back three level, it is characterised in that bag Include following steps：

The voltage vector of the voltage vector and machine-side converter of grid side converter is converted respectively by step 1 using coordinate transform To two-phase rest frame, and rower change is entered according to DC voltage value to two voltage vectors；

Step 2, according to the value of grid side converter voltage vector and machine-side converter voltage vector in two-phase rest frame Judge the actual sector number that two voltage vectors are located；

Step 3, will mark the grid side converter voltage vector and machine-side converter voltage vector changed using transformation matrix of coordinates It is converted into the first sector；

According to the grid side converter voltage vector after rotation transformation and machine-side converter voltage vector, step 4, judges that rotation becomes The little sector triangle numbering in the first sector that two voltage vectors after changing are located；

Step 5, according to the different little sector triangle that grid side converter voltage vector is located with machine-side converter voltage vector, Three fundamental voltage space vectors and corresponding midpoint electric current of grid side converter and machine-side converter are determined respectively；

Step 6, according to the different little sector triangle that grid side converter voltage vector is located with machine-side converter voltage vector, The action time of three fundamental voltage space vectors that calculating is acted on successively；

Step 7, three be each input into according to the deviant of DC side mid-point voltage and grid side converter and machine-side converter Phase current values, calculate the neutral point voltage balance factor of grid side converter and machine-side converter side respectively；

Step 8, according to grid side converter and the machine-side converter each action time of three fundamental voltage space vectors and meter The neutral point voltage balance factor for obtaining, calculates the output pulse duration for obtaining that vector is located at different little sector trianglees respectively Than；

Step 9, according to grid side converter voltage vector and the actual sector number of machine-side converter voltage vector, respectively by net Side converter and machine-side converter map back grid side converter voltage vector in the calculated duty cycle information in the first sector The actual sector being each located with machine-side converter voltage vector, and duty cycle information is scaled corresponding fiducial value to produce Raw pwm pulse drives grid side converter and machine-side converter respectively.

2. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, in step 1, using Clarke transform by three-phase voltage vector in two-phase rest frame, And rower change is entered according to DC voltage value to voltage vector, concretely comprise the following steps：

1) voltage vector is transformed under two-phase rest frame using formula (1)；

$\left[\begin{array}{c}{u}_{\mathrm{\α}}\\ u_{\mathrm{\β}}\end{array}\right]=\frac{2}{3}\left[\begin{array}{ccc}1& -\frac{1}{2}& -\frac{1}{2}\\ 0& \frac{\sqrt{3}}{2}& -\frac{\sqrt{3}}{2}\end{array}\right]\left[\begin{array}{c}{u}_a\\ u_b\\ u_c\end{array}\right]---\left(1\right)$

In above formula, u_a、u_bAnd u_cRespectively component of the voltage vector in abc coordinate systems, u_α、u_βRespectively voltage is sweared Component of the amount in α β coordinate systems；

2) amplitude limiting processing is carried out to two-phase voltage vector, wherein,

If $\sqrt{u_{\mathrm{\α}}^2+u_{\mathrm{\β}}^2}>\frac{u_{dc}}{\sqrt{3}},$ Then $\left\{\begin{array}{c}{u}_{\mathrm{\α}L}=\frac{u_{\mathrm{\α}}}{\sqrt{u_{\mathrm{\α}}^2+u_{\mathrm{\β}}^2}}\·\frac{u_{dc}}{\sqrt{3}}\\ u_{\mathrm{\β}L}=\frac{u_{\mathrm{\β}}}{\sqrt{u_{\mathrm{\α}}^2+u_{\mathrm{\β}}^2}}\·\frac{u_{dc}}{\sqrt{3}}\end{array}\right.$

If $\sqrt{u_{\mathrm{\α}}^2+u_{\mathrm{\β}}^2}\≤\frac{u_{dc}}{\sqrt{3}},$ Then $\left\{\begin{array}{c}{u}_{\mathrm{\α}L}=u_{\mathrm{\α}}\\ u_{\mathrm{\β}L}=u_{\mathrm{\β}}\end{array}\right.$

In above formula, u_dcFor DC voltage value, u_αL、u_βLRespectively voltage vector after amplitude limiting processing in α β coordinate systems In component；

3) enter rower change to two-phase voltage vector using formula (2)；

$\left\{\begin{array}{c}{v}_{\mathrm{\α}}=u_{\mathrm{\α}L}.\frac{\sqrt{3}}{u_{dc}}\\ v_{\mathrm{\β}}=u_{\mathrm{\β}L}\·\frac{\sqrt{3}}{u_{dc}}\end{array}\right.---\left(2\right)$

In above formula, v_α、v_βThe respectively per unit value of component of the voltage vector in α β coordinate systems.

3. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, judging in step 2 that the method for the voltage vector place sector number of grid side converter or machine-side converter is as follows：

1) defined variable a, b, c, its component u with voltage vector in α β coordinate systems_α、u_βBetween relation be

$\left[\begin{array}{c}a\\ b\\ c\end{array}\right]=\left[\begin{array}{cc}0& 1\\ \frac{\sqrt{3}}{2}& -\frac{1}{2}\\ -\frac{\sqrt{3}}{2}& -\frac{1}{2}\end{array}\right]\left[\begin{array}{c}{u}_{\mathrm{\α}}\\ u_{\mathrm{\β}}\end{array}\right]$

2) defined variable P, which with the relation between variable a, b, c is

P=sign (a)+2sign (b)+4sign (c)

In above formula, sign () is defined as seeking data symbol function $sign\left(u\right)=\left\{\begin{array}{cc}1& u\&GreaterEqual;0\\ 0& u<0\end{array}\right.;$

3) the sector number N that voltage vector is located is judged according to the value of variable P, the relation of wherein P and N is

P 1 2 3 4 5 6 N 2 6 1 4 3 5

Wherein, the span of N is 1～6.

4. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, in step 3, voltage vector is converted into into the first sector using coordinate transform formula (3)；

$\left[\begin{array}{c}{v}_{\mathrm{\&alpha;}T}\\ v_{\mathrm{\&beta;}T}\end{array}\right]=T_{N1}\left[\begin{array}{c}{v}_{\mathrm{\&alpha;}}\\ v_{\mathrm{\&beta;}}\end{array}\right]---\left(3\right)$

In above formula, v_αT、v_βTRespectively voltage vector after coordinate transform the first sector component, T_N1For N The transformation matrix of coordinates of sector to the first sector, the value of N is 1～6, wherein：

When voltage vector is located at the first sector, corresponding transformation matrix of coordinates is $T_{11}=\left[\begin{array}{cc}1& 0\\ 0& 1\end{array}\right];$

When voltage vector is located at the second sector, corresponding transformation matrix of coordinates is $T_{21}=\left[\begin{array}{cc}-\frac{1}{2}& \frac{\sqrt{3}}{2}\\ \frac{\sqrt{3}}{2}& \frac{1}{2}\end{array}\right];$

When voltage vector is located at three sectors, corresponding transformation matrix of coordinates is $T_{31}=\left[\begin{array}{cc}-\frac{1}{2}& \frac{\sqrt{3}}{2}\\ -\frac{\sqrt{3}}{2}& -\frac{1}{2}\end{array}\right];$

When voltage vector is located at four sectors, corresponding transformation matrix of coordinates is $T_{41}=\left[\begin{array}{cc}-\frac{1}{2}& -\frac{\sqrt{3}}{2}\\ -\frac{\sqrt{3}}{2}& \frac{1}{2}\end{array}\right];$

When voltage vector is located at five sectors, corresponding transformation matrix of coordinates is $T_{51}=\left[\begin{array}{cc}-\frac{1}{2}& -\frac{\sqrt{3}}{2}\\ \frac{\sqrt{3}}{2}& -\frac{1}{2}\end{array}\right];$

When voltage vector is located at six sectors, corresponding transformation matrix of coordinates is $T_{61}=\left[\begin{array}{cc}1& 0\\ 0& -1\end{array}\right].$

5. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, first sector is divided into the little sector trianglees of A～F six, by U between A deltas₀、U₁ And U₆Surround, by U between B deltas₀、U₆And U₂Surrounded, by U between C deltas₁、U₄And U₆Enclosed Into by U between D deltas₂、U₄And U₆Surrounded, by U between E deltas₁、U₃And U₄Surrounded, F tri- Angular interval is by U₂、U₄And U₅Surrounded, wherein U₀For zero vector 000, NNN and PPP, U₁For small vector 0NN And P00, U₂For small vector PP0 and 00N, U₃For big vector PNN, U₄For middle vector P0N, U₅For big vector PPN, U₆For middle vector U₄Half；In step 4：

Voltage vector after rotation transformation be located at A deltas between Rule of judgment be $\left\{\begin{array}{c}{v}_{\mathrm{\&beta;}T}\&le;\frac{\sqrt{3}}{3}{v}_{\mathrm{\&alpha;}T}\\ v_{\mathrm{\&beta;}T}\&le;-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+1\end{array}\right.;$

Voltage vector after rotation transformation be located at B deltas between Rule of judgment be $\left\{\begin{array}{c}{v}_{\mathrm{\&beta;}T}\&le;\sqrt{3}{v}_{\mathrm{\&alpha;}T}\\ v_{\mathrm{\&beta;}T}>\frac{\sqrt{3}}{3}{v}_{\mathrm{\&alpha;}T}\\ v_{\mathrm{\&beta;}T}\&le;-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+1\end{array}\right.;$

Voltage vector after rotation transformation be located at C deltas between Rule of judgment be $\left\{\begin{array}{c}{v}_{\mathrm{\&beta;}T}\&le;\frac{\sqrt{3}}{3}{v}_{\mathrm{\&alpha;}T}\\ v_{\mathrm{\&beta;}T}>-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+1\\ v_{\mathrm{\&beta;}T}>\sqrt{3}{v}_{\mathrm{\&alpha;}T}-1\\ v_{\mathrm{\&beta;}T}\&le;\frac{1}{2}\end{array}\right.;$

Voltage vector after rotation transformation be located at D deltas between Rule of judgment be $\left\{\begin{array}{c}{v}_{\mathrm{\&beta;}T}>\frac{\sqrt{3}}{3}{v}_{\mathrm{\&alpha;}T}\\ v_{\mathrm{\&beta;}T}>-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+1\\ v_{\mathrm{\&beta;}T}>\sqrt{3}{v}_{\mathrm{\&alpha;}T}-1\\ v_{\mathrm{\&beta;}T}\&le;\frac{1}{2}\end{array}\right.;$

Voltage vector after rotation transformation be located at E deltas between Rule of judgment be $\left\{\begin{array}{c}{v}_{\mathrm{\&beta;}T}\&le;\sqrt{3}{v}_{\mathrm{\&alpha;}T}-1\\ v_{\mathrm{\&beta;}T}\&le;-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+2\end{array}\right.;$

Voltage vector after rotation transformation be located at F deltas between Rule of judgment be $\left\{\begin{array}{c}{v}_{\mathrm{\&beta;}T}>\frac{1}{2}\\ v_{\mathrm{\&beta;}T}\&le;\sqrt{3}{v}_{\mathrm{\&alpha;}T}\\ v_{\mathrm{\&beta;}T}\&le;-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+2\end{array}\right..$

6. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, three fundamental voltage space vectors and midpoint electric current of grid side converter or machine-side converter in steps of 5 Determination method it is as follows：

1) the three fundamental voltage space vectors for acting on successively are set and is respectively v₁, v₂And v₃, then the first sector six is little by three The corresponding three fundamental voltage space vectors in angular interval are determined according to following table；

A B C D E F v₁ 0NN 00N 0NN 00N 0NN 00N v₂ 00N 000 00N P0N PNN P0N v₃ 000 P00 P0N P00 P0N PPN

2) set the corresponding midpoint electric current of three fundamental voltage space vectors for acting on successively and be respectively i₁, i₂And i₃, then first Between the little delta in sector six, the corresponding midpoint electric current of three fundamental voltage space vectors is determined according to following table；

A B C D E F i₁ i_a -i_c i_a -i_c i_a -i_c i₂ -i_c 0 -i_c i_b 0 i_b i₃ 0 -i_a i_b -i_a i_b 0

3) fundamental voltage space vector corresponding midpoint electric current is entered by the actual sector being located according to voltage vector according to following table Row is corresponding to be converted；

1st sector 2nd sector 3rd sector 4th sector 5th sector 6th sector i_a i_a1 i_b2 i_b3 i_c4 i_c5 i_a6 i_b i_b1 i_a2 i_c3 i_b4 i_a5 i_c6 i_c i_c1 i_c2 i_a3 i_a4 i_b5 i_b6

In upper table, i_aN, i_bNAnd i_cNFor the corresponding three-phase current in N sectors.

7. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, in step 6 three fundamental voltage space vectors that grid side converter or machine-side converter are acted on successively work It is as follows with time computational methods：

Voltage vector after rotation transformation is located between A deltas, the action time point of three fundamental voltage space vectors It is not $\left\{\begin{array}{c}{t}_1=(\sqrt{3}{v}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T})T_s\\ t_2=2v_{\mathrm{\&beta;}T}{T}_s\\ t_3=(1-\sqrt{3}{v}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T})T_s\end{array}\right.;$ Wherein, t₁For the action time of vector 0NN；t₂For vector 00N effect when Between；t₃For the action time of vector 000, T_sFor switch periods；

Voltage vector after rotation transformation is located between B deltas, the action time point of three fundamental voltage space vectors It is not $\left\{\begin{array}{c}{t}_1=2v_{\mathrm{\&beta;}T}{T}_s\\ t_2=(1-\sqrt{3}{v}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T})T_s\\ t_3=(\sqrt{3}{v}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T})T_s\end{array}\right.;$ Wherein, t₁For the action time of vector 00N；t₂For vector 000 effect when Between；t₃For the action time of vector P00；

Voltage vector after rotation transformation is located between C deltas, the action time point of three fundamental voltage space vectors It is not $\left\{\begin{array}{c}{t}_1=(1-2v_{\mathrm{\&beta;}T})T_s\\ t_2=(1-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+v_{\mathrm{\&beta;}T})T_s\\ t_3=(-1+\sqrt{3}{v}_{\mathrm{\&alpha;}T}+v_{\mathrm{\&beta;}T})T_s\end{array}\right.;$ Wherein, t₁For the action time of vector 0NN；t₂For the effect of vector 00N Time；t₃For the action time of vector P0N；

Voltage vector after rotation transformation is located between D deltas, the action time point of three fundamental voltage space vectors It is not $\left\{\begin{array}{c}{t}_1=(1-\sqrt{3}{v}_{\mathrm{\&alpha;}T}+v_{\mathrm{\&beta;}T})T_s\\ t_2=(-1+\sqrt{3}{v}_{\mathrm{\&alpha;}T}+v_{\mathrm{\&beta;}T})T_s\\ t_3=(1-2v_{\mathrm{\&beta;}T})T_s\end{array}\right.;$ Wherein, t₁For the action time of vector 00N；t₂For the effect of vector P0N Time；t₃For the action time of vector P00；

Voltage vector after rotation transformation is located between E deltas, the action time point of three fundamental voltage space vectors It is not $\left\{\begin{array}{c}{t}_1=(2-\sqrt{3}{v}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T})T_s\\ t_2=(\sqrt{3}{v}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T}-1)T_s\\ t_3=2v_{\mathrm{\&beta;}T}{T}_s\end{array}\right.;$ Wherein, t₁For the action time of vector 0NN；t₂For the effect of vector PNN Time；t₃For the action time of vector P0N；

Voltage vector after rotation transformation is located between F deltas, the action time point of three fundamental voltage space vectors It is not $\left\{\begin{array}{c}{t}_1=(2-\sqrt{3}{v}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T})T_s\\ t_2=(\sqrt{3}{u}_{\mathrm{\&alpha;}T}-v_{\mathrm{\&beta;}T})T_s\\ t_3=(2v_{\mathrm{\&beta;}T}-1)T_s\end{array}\right.;$ Wherein, t₁For the action time of vector 00N；t₂For vector P0N effect when Between；t₃For the action time of vector PPN.

8. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, in step 7, the neutral point voltage balance factor computational methods of grid side converter or machine-side converter are as follows：

1) mid-point voltage for needing to adjust is set as u_adj, regulating time distribution factor k, wherein C is calculated according to formula (4)_dc For the size of dc-link capacitance；

$k=\frac{2C_{dc}{u}_{adj}-i_2{t}_2-i_3{t}_3}{i_1{t}_1}---\left(4\right)$

Wherein, u_adj=-u_N0=u_dc1-u_dc2, u_N0For mid-point voltage value, u_dc1And u_dc2The respectively upper electricity of dc bus Hold magnitude of voltage and lower capacitance voltage value；

2) the maximum balance factor k of the neutral-point voltage balance of grid side converter is calculated according to formula (5)₁, according to formula (6) the maximum balance factor k of the neutral-point voltage balance of computer-side changer₂；

$k_1=\frac{-2C_{dc}{u}_{NO}-i_{2g}{t}_{2g}-i_{3g}{t}_{3g}}{i_{1g}{t}_{1g}}---\left(5\right)$

$k_2=\frac{-2C_{dc}{u}_{NO}-i_{2m}{t}_{2m}-i_{3m}{t}_{3m}}{i_{1m}{t}_{1m}}---\left(6\right)$

Wherein, i_1g、i_2g、i_3gThe corresponding midpoint of three fundamental voltage space vectors that respectively grid side converter is acted on successively Electric current, t_1g、t_2g、t_3gRespectively action time of three fundamental voltage space vectors that grid side converter is acted on successively, i_1m、 i_2m、i_3mThe corresponding midpoint electric current of three fundamental voltage space vectors that respectively machine-side converter is acted on successively, t_1m、t_2m、 t_3mRespectively action time of three fundamental voltage space vectors that machine-side converter is acted on successively；

3) two neutral point voltage balance factors k are set_gAnd k_mRedistribute grid side converter and machine-side converter first is basic The action time of space vector of voltage, take k_g=k₁And k_m=k₂, neutral point voltage balance factor k_gAnd k_mModification method such as Under：

(1) if abs is (k₁)≤1, abs (k₂)≤1, then make k_m=0, k_gKeep constant；

(2) if abs is (k₁)≤1, abs (k₂)>1, then make k_m=0, k_gKeep constant；

(3) if abs is (k₁)>1, abs (k₂)≤1, then make k_g=0, k_mKeep constant；

(4) if abs is (k₁)>1, abs (k₂)>1, then make k_g=sign (k_g), k_mCalculated according to formula (7)；

$k_m=\frac{-2C_{dc}(u_{NO}-u_{adjG})-i_{2m}{t}_{2m}-i_{3m}{t}_{3m}}{i_{1m}{t}_{1m}}---\left(7\right)$

Wherein u_adjGFor the mid-point voltage value that net side current transformer can be adjusted, its computing formula is $u_{adjG}=\frac{1}{2C_{dc}}\&lsqb;i_{1g}{t}_{1g}\&CenterDot;sign\left(k_g\right)+i_{2g}{t}_{2g}+i_{3g}{t}_{3g}\&rsqb;.$

9. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, in step 8, the computational methods for exporting pulse duty factor are：

1) three variable t are defined_a, t_bAnd t_c, and calculate according to formula (8)；

Define three variable d_a、d_bAnd d_cAs three dutyfactor values, calculate according to formula (9)；

$\left\{\begin{array}{c}{d}_{ag}=1-\frac{2\&CenterDot;t_{ag}}{T_s}\\ d_{bg}=1-\frac{2\&CenterDot;t_{bg}}{T_s}\\ d_{cg}=1-\frac{2\&CenterDot;t_{cg}}{T_s}\end{array}\right.---\left(9\right)$

In above formula, k_iRepresent neutral point voltage balance factor k of grid side converter_gOr the neutral point voltage balance of machine-side converter because Sub- k_m；

2) according between the different deltas at the voltage vector place after rotation transformation, three-phase bridge arm is selected according to following table Effective duty cycle；

A B C D E F d_A d_c d_b d_b d_a d_a d_a d_B d_a d_cg d_a d_c d_b d_b d_C d_b d_a d_c d_b d_c d_c

Wherein, d_A、d_BAnd d_CFor the effective duty cycle of three-phase bridge arm.

10. the neutral-point voltage balance method of the double pwm converter of back-to-back three level according to claim 1, Characterized in that, the concrete grammar of step 9 includes：

1) actual sector being located according to grid side converter voltage vector or machine-side converter voltage vector, three-phase bridge arm Two groups of duty cycle information conversions are as follows：

When the voltage vector after rotation transformation be located at A deltas between or C deltas between or E deltas between when, Two dutycycles of three-phase bridge arm of former sector that voltage vector is located are mapped back according to following table value；

1st sector 2nd sector 3rd sector 4th sector 5th sector 6th sector D_a1 d_A 0 0 0 0 d_A D_a2 1 d_B d_C d_C d_B 1 D_b1 0 d_A d_A 0 0 0 D_b2 d_B 1 1 d_B d_C d_C D_c1 0 0 0 d_A d_A 0 D_c2 d_C d_C d_B 1 1 d_B

When the voltage vector after rotation transformation be located at B deltas between or D deltas between or F deltas between when, Two dutycycles of three-phase bridge arm of former sector that voltage vector is located are mapped back according to following table value；

1st sector 2nd sector 3rd sector 4th sector 5th sector 6th sector D_a1 d_A d_B 0 0 d_B d_A D_a2 1 1 d_C d_C 1 1 D_b1 d_B d_A d_A d_B 0 0 D_b2 1 1 1 1 d_C d_C D_c1 0 0 d_B d_A d_A d_B D_c2 d_C d_C 1 1 1 1

2) value of corresponding six modulating waves of three bridge arms is determined according to formula (10)

$\left\{\begin{array}{c}{\mathrm{CP}}_{A1}=CNT\&CenterDot;D_{a1}\\ {\mathrm{CP}}_{A2}=CNT\&CenterDot;D_{a2}\\ {\mathrm{CP}}_{B1}=CNT\&CenterDot;D_{b1}\\ {\mathrm{CP}}_{B2}=CNT\&CenterDot;D_{b2}\\ {\mathrm{CP}}_{C1}=CNT\&CenterDot;D_{c1}\\ {\mathrm{CP}}_{C2}=CNT\&CenterDot;D_{c2}\end{array}\right.---\left(10\right)$

In above formula, CP_A1、CP_A2The respectively value of two modulating waves of A phases；CP_B1、CP_B2Respectively B phases two are adjusted The value of ripple processed；CP_C1、CP_C2The respectively value of two modulating waves of C phases；Peak values of the CNT for triangular carrier.