CN105762833A - Dead zone compensation method for grid-connected inverter - Google Patents

Abstract

The invention relates to a dead zone compensation method for a grid-connected inverter, and the method comprises the steps: carrying out the coordinate transformation of three-phase output currents IA, IB and IC collected by the grid-connected inverter in each switching period Ts, and enabling the SVPWM (space vector pulse width modulation) sector judgment and the current vector position to be unified; obtaining a compensated voltage vector according to the position of a sector where the current vector is located; correcting a given vector of voltage through employing the compensated voltage vector, achieving the direct correction of the voltage vector on the stationary coordinate axes alpha and beta, and directly enabling the corrected voltage vector to serve as a reference instruction for the SVPWM. The method directly corrects the voltage instruction of the SVPWM, and directly carries out compensation in a stationary coordinate system of the voltage vector, thereby avoiding complex calculation, enabling the programming to be easy, and greatly reducing the operation time of a CPU.

Description

For the dead-zone compensation method in combining inverter

Technical field

The present invention relates to a kind of electric and electronic technical field, particularly to a kind of for the dead-zone compensation method in combining inverter.

Background technology

In combining inverter, mostly adopt voltage-type pulse width modulation inverter, in order to avoid the directly conducting of the switching component up and down of same brachium pontis, it is necessary to add Dead Time.But so can cause voltage waveform distortion and harmonic wave that inverter exports, particularly when low-voltage small area analysis, therefore the performance improvement of combining inverter will be very helpful by effective dead-zone compensation method, be required for carrying out dead area compensation in most of inverter.General dead area compensation needs two necessary conditions: the interpolation of accurately detection and the correct compensation vector of current zero-crossing point.Wherein major problem is that of the detection of current polarity, owing to the reason of sampling causes the detection of zero crossing inaccurate, speed when switching noise and electric current pass through zero point during low frequency bring difficulty all can to the detection of zero crossing；Secondly when calculating corresponding dead area compensation vector, most calculating concentrates on in the optimization of CPU counter, and what seldom have method directly simplicity provides a kind of general compensation method.

Application number be 200610144322 the patent dead-zone compensation method of output " a kind of space vector pulse width modulation " compensate PWM pulsewidth in each sector of SVPWM modulation strategy according to phase current polarity, the problem that the method exists Current polarity detection, simultaneously because be according to phase current polarity, the PWM pulsewidth of each sector is compensated, have 36 kinds of compensation situations, therefore more complicated, and occupy substantial amounts of cpu resource.Application number is the patent " dead-zone compensation method of converter " of 00122378, the rotating coordinate system that space voltage vector is directed obtains two shaft currents, the azimuth of resultant current vector is obtained after low-pass filtering, then three-phase phase current polarity is determined with this, but due to the problem using low pass filter the method that current vector angle estimation can be brought delayed, simultaneously when low frequency it is difficult to ensure that current vector angle estimation precision, and the method is more complicated occupies substantial amounts of cpu resource.Application number be 201010620219.3 patent " a kind of for the dead-zone compensation method in voltage space vector pulse width modulation technology " have employed the method that electric current directly detects, on SVPWM, direct compensation method can cause the zero crossing of detection inaccurate greatly and take great cpu resource in compensation process.

Summary of the invention

The present invention be directed to the problem that traditional space vector pulse width modulation takies a large amount of cpu resource, propose a kind of for the dead-zone compensation method in combining inverter, hardware circuit that need not be extra on the basis of traditional space vector pulse width modulation SVPWM, effectively make use of the resource of CPU, the sector of SVPWM is judged, and the polarity with electric current judges to have unified, decrease the time of judgement, on this basis, give one directly do not change Counter Value, the method that the voltage of compensation is applied directly to SVPWM input, is effectively reduced amount of calculation.

The technical scheme is that a kind of for the dead-zone compensation method in combining inverter, the three-phase collected in each for combining inverter switch periods is exported electric current I_A、I_B、I_CIn a switch periods Ts, the current expression based on static coordinate conversion is obtained by static coordinate conversion, rest frame is turned clockwise 30 ° again and carry out coordinate transform, each component of electric current is transferred in the coordinate system of 30 ° of turning clockwise, the sector of space vector pulse width modulation SVPWM is judged unified with current phasor position；Sector position according to current phasor place, the voltage vector being compensated；Utilize the voltage vector compensated that the given vector of voltage is modified, it is achieved directly correction to voltage vector on static coordinate α and β axle, directly will after correction voltage vector to SVPWM as reference instruction.

It is described that in a switch periods Ts, to the three-phase output electric current gathered, to perform twice at coordinate transform formula as follows:

$\left\{\begin{array}{c}\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]=K\left[\begin{array}{ccc}1& -\frac{1}{2}& -\frac{1}{2}\\ 0& \frac{\sqrt{3}}{2}& -\frac{\sqrt{3}}{2}\end{array}\right]\left[\begin{array}{c}{i}_A\\ i_B\\ i_C\end{array}\right]\\ \left[\begin{array}{c}{i_{\mathrm{\α}}}^{\′}\\ {i_{\mathrm{\β}}}^{\′}\end{array}\right]=\left[\begin{array}{cc}\cos \left(\frac{\mathrm{\π}}{6}\right)& -\sin \left(\frac{\mathrm{\π}}{6}\right)\\ \cos \left(\frac{\mathrm{\π}}{6}\right)& \sin \left(\frac{\mathrm{\π}}{6}\right)\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]\end{array}\right.$

Wherein $\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]$ For static coordinate conversion current expression, $\left[\begin{array}{c}{i_{\mathrm{\α}}}^{\′}\\ {i_{\mathrm{\β}}}^{\′}\end{array}\right]$ Turn clockwise the current expression after 30 ° for static coordinate, be chosen as according to the mode conversion COEFFICIENT K of constant power conversion or equivalentsOr

The voltage vector formula of described compensation is as follows:

$\left\{\begin{array}{c}{U}_{\mathrm{\α}comp}=\frac{4*Ud*Td}{3T}\×cos(\mathrm{\π}/3)\\ U_{\mathrm{\β}comp}=\frac{4*Ud*Td}{3T}\×sin(\mathrm{\π}/3)\end{array}\right.$

Wherein Ud is the voltage swing of dc bus, and Td is Dead Time, and T is switch periods, and the sector then passing through current polarity judges that the size of the voltage vector of compensation is applied directly to voltage to be given.

The beneficial effects of the present invention is: the present invention is for the dead-zone compensation method in combining inverter, directly the voltage instruction of space vector pulse width modulation is modified, directly compensate in the rest frame of voltage vector, thus avoiding the calculating of complexity, it is easily programmed, greatly reduces the operation time of CPU.

Accompanying drawing explanation

Fig. 1 is analog systems block diagram of the present invention；

Fig. 2 is twice coordinate transform of the present invention and current polarity judgement figure；

Fig. 3 is judgement figure in sector of the present invention and current polarity judgement figure；

Fig. 4 is that the calculating of the voltage vector that the present invention compensates is with reference to figure；

Fig. 5 a is the simulated current oscillogram that the present invention has added dead area compensation；

Fig. 5 b is the simulated current frequency spectrum block diagram that the present invention has added dead area compensation；

Fig. 5 c is the current waveform figure that the present invention is not added with dead area compensation emulation；

Fig. 5 d is the current waveform frequency spectrum block diagram that the present invention is not added with dead area compensation emulation；

Fig. 6 is the oscillogram being not added with dead area compensation during the present invention tests；

Fig. 7 is the oscillogram adding dead area compensation during the present invention tests.

Detailed description of the invention

For the dead-zone compensation method in combining inverter, it it is the one dead-zone compensation method that is used in space vector pulse width modulation (SVPWM), it is as follows that the method realizes step: analog systems block diagram as shown in Figure 1, and the three-phase collected in each switch periods is exported electric current I_A、I_B、I_C, obtain the current expression based on static coordinate conversion by static coordinate conversion, it is possible to reduce the impact of sampling zero crossing.Then passing through turns clockwise rest frame 30 ° carries out coordinate transform, is transferred to by each component of electric current in the coordinate system of 30 ° of turning clockwise, and thus judges to unite with it by the sector of SVPWM, decreases operand.Then the voltage vector of corresponding compensation is obtained by the positional information of electric current.Changing SVPWM action time with traditional direct compensation cycle rate counter different is, the voltage instruction of SVPWM can directly be modified by the method, directly compensate in the rest frame of voltage vector, thus avoiding the calculating of complexity, it is easily programmed, greatly reduces the operation time of CPU.

The method that the dead area compensation of the present invention will be described with the example of concrete combining inverter below.Fig. 1 is analog systems block diagram.Simulated implementation platform herein is Shanghai University Of Electric Power's information and engineering college's piconet simulation experiment platform project, this inverter is the model prototype actual electric pressure made after reducing in the way of 1:100, wherein, the voltage that net is surveyed is 3.8V, grid-connected voltage is 25V, filter inductance be sized to 1.7mH, switch periods is 12.8KHz, the TMSF320F2812 that CPU is TI company used, the Dead Time being manually set is 3.3us.Here current sensor is the current Hall sensor of Lyme, and its model is HO-NSMSP33.

The first step, the result of sampling is obtained by the result that three-phase current is sampled by Hall current sensor within an interrupt cycle of EV intervalometer, then pass through twice coordinate transform, coordinate is transformed to as shown in Figure 2, α and β axle be in rest frame horizontally and vertically, twice coordinate transform can reduce the judgement of current zero-crossing point and be forbidden, and the sector of SVPWM judging, the polarity with electric current judges to want to combine.The formula of its reference is:

$\left\{\begin{array}{c}\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]=K\left[\begin{array}{ccc}1& -\frac{1}{2}& -\frac{1}{2}\\ 0& \frac{\sqrt{3}}{2}& -\frac{\sqrt{3}}{2}\end{array}\right]\left[\begin{array}{c}{i}_A\\ i_B\\ i_C\end{array}\right]\\ \left[\begin{array}{c}{i_{\mathrm{\α}}}^{\′}\\ {i_{\mathrm{\β}}}^{\′}\end{array}\right]=\left[\begin{array}{cc}\cos \left(\frac{\mathrm{\π}}{6}\right)& -\sin \left(\frac{\mathrm{\π}}{6}\right)\\ \cos \left(\frac{\mathrm{\π}}{6}\right)& \sin \left(\frac{\mathrm{\π}}{6}\right)\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]\end{array}\right.$

Wherein $\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]$ For static coordinate conversion current expression, $\left[\begin{array}{c}{i_{\mathrm{\α}}}^{\′}\\ {i_{\mathrm{\β}}}^{\′}\end{array}\right]$ Turning clockwise the current expression after 30 ° for static coordinate, the mode conversion COEFFICIENT K here according to constant power conversion or equivalents is chosen asOr

Second step: by the sector position at the position judgment place of the current phasor after twice coordinate transform, as it is shown on figure 3, after the method for coordinate transform detects the polarity of electric current accurately, the sector position according to current phasor place, in a switch periods Ts, the applied voltage vector being compensated.

Formula such as its voltage vector of Fig. 4 is as follows, vector compensation reference by location table 1,

$\left\{\begin{array}{c}{U}_{\mathrm{\α}comp}=\frac{4*Ud*Td}{3T}\×cos(\mathrm{\π}/3)\\ U_{\mathrm{\β}comp}=\frac{4*Ud*Td}{3T}\×sin(\mathrm{\π}/3)\end{array}\right.$

Ud is the voltage swing of dc bus, Td is Dead Time, and T is switch periods, owing to utilizing above-mentioned changes in coordinates to be united by the sector judgment formula of the judgment formula of current polarity Yu SVPWM, so have only to the polarity according to electric current, the voltage vector in compensation formula.

Table 1

3rd step: utilize the vector compensated that the given vector of voltage is modified, it is achieved direct correction on α and β axle to voltage vector, it is achieved accurate voltage compensation.The dead band vector time of compensation is given to SVPWM as reference instruction directly as the vector signal of voltage,

$\left\{\begin{array}{c}{U_{\mathrm{\α}}}^{\′}=U_{\mathrm{\α}}+U_{\mathrm{\α}comp}\\ {U_{\mathrm{\β}}}^{\′}=U_{\mathrm{\β}}+U_{\mathrm{\β}comp}\end{array}\right.$

Wherein U_αAnd U_βFor voltage vector given in static coordinate, U_α' and U_β' for voltage vector revised in static coordinate.

Fig. 5 is the experimental result finally obtained, and wherein Fig. 5 c is the current waveform being not added with dead area compensation emulation, and Fig. 5 a is the simulated current waveform having added dead area compensation.Fig. 5 b is the frequency spectrum block diagram of Fig. 5 a, and Fig. 5 d is the frequency spectrum block diagram of 5c.Fig. 6 is the voltage and current waveform being not added with dead area compensation in experiment, and Fig. 7 is the voltage and current waveform plus dead area compensation.Can significantly find out that current waveform becomes better along with addition dead area compensation quality.

This shows that the present invention effectively improves the waveform quality of electric current, reduce current harmonics, improve code quality.

Claims (3)

1. the dead-zone compensation method being used in combining inverter, it is characterised in that the three-phase collected in each for combining inverter switch periods is exported electric current I_A、I_B、I_CIn a switch periods Ts, the current expression based on static coordinate conversion is obtained by static coordinate conversion, rest frame is turned clockwise 30 ° again and carry out coordinate transform, each component of electric current is transferred in the coordinate system of 30 ° of turning clockwise, the sector of space vector pulse width modulation SVPWM is judged unified with current phasor position；Sector position according to current phasor place, the voltage vector being compensated；Utilize the voltage vector compensated that the given vector of voltage is modified, it is achieved directly correction to voltage vector on static coordinate α and β axle, directly will after correction voltage vector to SVPWM as reference instruction.

2. according to claim 1 for the dead-zone compensation method in combining inverter, it is characterised in that described in a switch periods Ts, the three-phase output electric current gathered to be performed twice at coordinate transform formula as follows:

$\left\{\begin{array}{c}\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]=K\left[\begin{array}{ccc}1& -\frac{1}{2}& -\frac{1}{2}\\ 0& \frac{\sqrt{3}}{2}& -\frac{\sqrt{3}}{2}\end{array}\right]\left[\begin{array}{c}{i}_A\\ i_B\\ i_C\end{array}\right]\\ \left[\begin{array}{c}{i_{\mathrm{\α}}}^{\′}\\ {i_{\mathrm{\β}}}^{\′}\end{array}\right]=\left[\begin{array}{cc}cos\left(\frac{\mathrm{\π}}{6}\right)& -\sin \left(\frac{\mathrm{\π}}{6}\right)\\ cos\left(\frac{\mathrm{\π}}{6}\right)& \sin \left(\frac{\mathrm{\π}}{6}\right)\end{array}\right]\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]\end{array}\right.$

Wherein $\left[\begin{array}{c}{i}_{\mathrm{\α}}\\ i_{\mathrm{\β}}\end{array}\right]$ For static coordinate conversion current expression, $\left[\begin{array}{c}{i_{\mathrm{\α}}}^{\′}\\ {i_{\mathrm{\β}}}^{\′}\end{array}\right]$ Turn clockwise the current expression after 30 ° for static coordinate, be chosen as according to the mode conversion COEFFICIENT K of constant power conversion or equivalentsOr

3. according to claim 1 for the dead-zone compensation method in combining inverter, it is characterised in that the voltage vector formula of described compensation is as follows:

$\left\{\begin{array}{c}{U}_{\mathrm{\α}comp}=\frac{4*Ud*Td}{3T}\×\cos (\mathrm{\π}/3)\\ U_{\mathrm{\β}comp}=\frac{4*Ud*Td}{3T}\×\sin (\mathrm{\π}/3)\end{array}\right.$

Wherein Ud is the voltage swing of dc bus, and Td is Dead Time, and T is switch periods, and the sector then passing through current polarity judges that the size of the voltage vector of compensation is applied directly to voltage to be given.