CN102263417B - Method for designing hybrid damping parameter of LCL filter in photovoltaic grid-connected inverter - Google Patents

Abstract

The invention discloses a method for designing a hybrid damping parameter of an LCL filter in a photovoltaic grid-connected inverter, and belongs to the technical fields of new energy sources and distributed grid-connected generation. The method is characterized in that: the resonance of the LCL filter is suppressed by adopting an active damping/passive damping cooperative working method, wherein a filter capacitive current feedback control method is adopted for an active damping part and the method of connecting filtering capacitance in series with damping resistance is adopted for a passive damping part. For the LCL filter, the damping resistance required by passive damping and a feedback coefficient of active damping are computed under the condition of ensuring a sufficient damping coefficient, and a designed cooperative damping control scheme is checked to make the loss of the damping resistance as low as possible, avoid the influence of system control delay on damping running; and the damping is adapted to strong power grid access and the weak power grid access of remote areas. The method is applied to photovoltaic grid-connected generation systems adopting the LCL filter, fuel cells and grid-connected wind power generation systems.

Description

Grid-connected with LCL filter mixing damping parameter method for designing in the inverter

Technical field

The present invention relates to a kind ofly reduce to control non-ideal factor influence and strengthen damping adaptive grid-connected with LCL filter damping method, belong to the new forms of energy technical field of generating electricity by way of merging two or more grid systems.

Background technology

Along with fossil energy is progressively exhausted, renewable energy power generation becomes the problem that the world payes attention to.Photovoltaic generation especially receives publicity as the important new forms of energy that rising.Solar energy is a kind of cleaning, inexhaustible regenerative resource, and China's solar radiation aboundresources is can the exploitation amount huge.The whole world over past ten years the installed capacity of photovoltaic generation present the trend of exponential increase, and 90% is grid-connected system.Therefore, development also improves the parallel network power generation technology, efficiently utilizes solar energy, injects the clean energy resource of the high quality of power supply more to electrical network, for the protection of the saving of the mankind's sustainable development, resource, environment with improve highly beneficial.

Will inject solar energy to electrical network, just must make electric energy satisfy the quality of power supply standard that meets the electrical network regulation, namely the grid-connected current harmonic content must be lower than a certain limit value.For photovoltaic DC-to-AC converter, the switch motion of its device will produce harmonic wave inevitably, and grid-connected current wave apparatus after filtration just can meet the standard of being incorporated into the power networks.For the low capacity photovoltaic DC-to-AC converter, single inductor filter can be realized filter function.And the excessive dynamic property variation that makes system of hicap inductance value, the loss on the filter is excessive.Nineteen ninety-five, people such as M.Lindgren have proposed to replace single L filter with the LCL filter, and it is big to solve big volume change device filter inductance, the problem of dynamic performance variation.LCL filtering has not only reduced total filter inductance value effectively, and further strengthens for the damping capacity of high-frequency harmonic.Yet the inherent characteristic of LCL filter has determined that its harmonic impedance for a certain characteristic frequency is zero, and this will cause resonance problems and this subharmonic is amplified, and cause system's instability.Solution to this problem mainly contains at present: the passive damping method, and namely series damping resistor plays attenuation to resonance on the electric capacity of LCL filter; The active damping method is namely eliminated resonance by the improvement of control algolithm, mainly comprises the many feedback quantity control methods of many rings and zero POLE PLACEMENT USING method.The former mainly be new send out feedback amount such as filter capacitor electric current by increasing, rate is carried out the inhibition of resonance point from inductive drop, filter capacitor voltage etc., the latter eliminates the limit of resonance by zero POLE PLACEMENT USING in current regulator.The active damping of passive damping and the many feedback quantity controls of many rings has had ripe product to come out at present.

No matter be active damping or passive damping, all can be subjected to the influence of some factors.Passive damping is comparatively simple, the influence of uncontrolled device non-ideal factor as postponing etc., but the electrical network inductance can weaken damping when the light current net inserts; Because the main characteristics of resonance are the rapid increases of capacitance current, so be that the virtual resistance method of additional feedback amount is the most common in active damping with the capacitance current.No matter be how the many feedback quantities of ring are controlled, still zero POLE PLACEMENT USING method, all strictness is controlled the influence that parameter particularly postpones.Though zero POLE PLACEMENT USING method has been omitted transducer, parameter adaptation is relatively poor.So active damping and passive damping have some complementary characteristics, two kinds of methods are combined the collaborative damped control system of formation, can satisfy grid adaptability simultaneously, the influence with control lag and parameter variation simultaneously reduces.

Summary of the invention

The present invention is directed to conventional grid-connected LCL filter damping and control a kind of photovoltaic parallel in system LCL filter damping method for designing that is applied in that is difficult to take into account the problem of electrical network parameter adaptability, damping resistance loss, the influence of control non-ideal factor simultaneously and proposes.Collaborative work by active damping and passive damping part, on the one hand thereby less damping resistance reduces damping loss, makes that on the other hand damping system still is that the parameter non-ideal factor of controller all has stronger adaptability for the electrical network variation inductance.Present existing home products adopts passive damping mostly, the bigger generating efficiency that reduced of the damping resistance excessive then loss of design; Damping resistance is less than normal then can to cause damping weakened greatly because of the electrical network inductance when the light current net inserts.And adopt the product of active damping comparatively rare, the part operation is on-the-spot the problem that high order harmonic component appears in particular resonance still to have occurred, and offshore company adopts active damping comparatively general.

The LCL filter damping of photovoltaic parallel in system control at present all is single form, exists some defectives of aspect such as robustness.The present invention is directed to the concrete condition of the photovoltaic grid-connected device of LCL filtering, the method that adopts active damping to combine with passive damping has reduced damping loss, has strengthened robustness and parameter adaptation.The invention is characterized in:

Step (1): the parameter of initialization LCL filter:

L ₁Inversion side inductance value for the LCL filter;

L ₂Some net side inductance value for the LCl filter;

C _fFilter capacitor value for the LCL filter;

The electrical network inductance L of access point _gThe variation span: less than 1.6mH;

Step (2): initialization total damping coefficient ζ, active damping coefficient ζ _ADAnd passive damping coefficient ζ _PD, wherein: ζ=ζ _AD+ ζ _PD:

ζ=0.1，ζ _AD=0.05，ζ _PD=0.05；

Calculate capacitor current feedback coefficient k and the passive damping resistance R of active damping according to following formula _d:

$k=2{\mathrm{\&zeta;}}_{\mathrm{AD}}{L}_1\sqrt{\frac{{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+L_2}{L_1{L}_2{C}_f}};$

$R_d=\frac{2{\mathrm{\&zeta;}}_{\mathrm{PD}}}{{\mathrm{<figure-callout\; id="1"\; label="C\; f\; L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">C</figure-callout>}}_f\sqrt{\frac{L_{}}{}}\sqrt{\frac{{}_1+L_2}{L_1{L}_2{C}_f}}};$

Step (3): the passive damping resistance value that step (2) is obtained rounds up, and obtains the new value R ' that meets the resistance specification _d

Step (4): (1) result of obtaining is to initialized active damping coefficient ζ in the step (2) set by step _AD, passive damping coefficient ζ _PDRevise, obtain revised active damping coefficient ξ ' _ADWith revised passive damping coefficient ζ ' _PD:

${\mathrm{\&zeta;}}_{\mathrm{PD}}^{\&prime;}=\frac{R_d{\mathrm{<figure-callout\; id="1"\; label="C\; f\; L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">C</figure-callout>}}_f\sqrt{\frac{L_{}}{}}\sqrt{\frac{{}_1+L_2}{L_1{L}_2{\mathrm{<figure-callout\; id="2"\; label="C\; f"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">C</figure-callout>}}_f}}}{2},$ Get L herein _g=0,

ζ ' _AD=0.1-ζ ' _PD, this moment ζ=0.1,

Step (5): the capacitor current feedback coefficient k of the active damping that obtains by following formula correction step (2), use k ' expression correction result:

$k^{\&prime;}=2{\mathrm{\&zeta;}}_{\mathrm{AD}}^{\&prime;}{L}_1\sqrt{\frac{{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+L_2}{L_1{L}_2{C}_f}};$

Step (6): press total damping coefficient ζ described in the following formula plot step (2) at the revised damping resistance value R ' of passive damping _d, active damping capacitor current feedback coefficient correction value k ' and electrical network inductance L _gWhen 0 to 1.6mH changes, the change curve of described total damping coefficient ζ:

$\mathrm{\&zeta;}=\frac{R_d^{\&prime;}{\mathrm{<figure-callout\; id="1"\; label="C\; f\; L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">C</figure-callout>}}_f\sqrt{\frac{L_{}}{}}\sqrt{\frac{{}_1+{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g}{L_1({\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g){\mathrm{<figure-callout\; id="2"\; label="C\; f"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">C</figure-callout>}}_f}}}{2}+\frac{k^{\&prime;}}{{2L}_1\sqrt{\frac{{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+{\mathrm{<figure-callout\; id="2"\; label="l"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">l</figure-callout>}}_2+L_g}{L_1({\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g)C_f}}},$

If: the undulating value that total damping coefficient ζ changes with the electrical network inductance surpasses 20% o'clock of total damping coefficient settings value that step (2) sets, then returns step (2), under the constant prerequisite of the value of described total damping coefficient ζ, to passive damping coefficient ζ _PDWith active damping coefficient ζ _ADRedistribute;

If: total damping coefficient ζ is no more than 20% of total damping coefficient value ζ that step (2) sets with electrical network variation inductance undulating value, then changes step (7) over to;

Step (7): the following damping parameter that mixes of output:

Total damping coefficient ζ=0.1, revised active damping coefficient ζ ' _AD, revised passive damping coefficient ζ ' _PD, revised passive damping resistance R ' _dAnd the capacitor current feedback coefficient k of revised active damping '.

The design can alleviate photovoltaic inversion control delay factor to the adverse effect of the stability of a system with respect to pure active damping system, can strengthen the grid adaptability of damping with respect to pure passive damping system, reduces the damping resistance loss, improves system changeover efficient.

Description of drawings

Fig. 1. the inventive method flow chart.

Fig. 2. adopt the photovoltaic parallel in system circuit diagram of LCL filtering.

Fig. 3. the one phase equivalent circuit of the photovoltaic parallel in system of LCL filtering when considering the electrical network non-ideal factor.

Fig. 4. damping coefficient is with electrical network variation inductance curve chart in the embodiment.

Fig. 5. adopt LCL filtering not adopt the photovoltaic parallel in system current loop control of damping control

Block diagram.

Fig. 6. initiatively with the LCL filtering photovoltaic parallel in system current loop control block diagram of passive damping Collaborative Control.

Fig. 7. active realizes schematic diagram with photovoltaic parallel in system circuit diagram and the control thereof of the LCL filtering of passive damping Collaborative Control.MPPT(MPPT maximum power point tracking wherein) link is carried out MPPT maximum power point tracking to photovoltaic array, and is provided the active current output order; By measuring grid-connected current and obtaining meritorious and idle component through the dq coordinate transform, the active current command value that provides with MPPT respectively and referenced reactive current value (generally getting 0) are carried out PI and are regulated control and form closed loop again; By measuring line voltage and coordinate transform obtains the dq conversion that an amount of angle of line voltage is used for grid-connected current through α β, make output current and the line voltage of inverter with the frequency homophase; The output of pi regulator add again to transform under the α β coordinate system behind the cross decoupling item provide the SVPWM(space vector pulse width modulation) instruction drives inverter work.The present invention relies on this system, and in the series damping resistor, the electric current of measuring simultaneously on the filter capacitor carries out subtracting each other the new closed loop feedback of introducing with the output of pi regulator after the dq conversion on the filter capacitor of LCL.

Embodiment

The present invention is starting point with adaptability and the robustness that improves damping, carries out new design, is used for photovoltaic parallel in system.This damping scheme relies on the control of existing photovoltaic combining inverter, owing on existing LCL filter and inverter control basis, only need to increase hardware and the control of damping portion, so except the relevant parameter of damping, other parameter is all constant.

This design is that method flow as shown in Figure 1 for the photovoltaic parallel in system that adopts LCL filtering provides the damping solution.Be that example is carried out damping of the present invention design with a photovoltaic inversion grid connection system, the parameter of LCL filter is as shown in table 1, and the circuit diagram of grid-connected system and single-phase reduced graph thereof are as shown in Figures 2 and 3.Embodiment is as follows:

The parameter value of table 1LCL filter

1. clear and definite requirement of system design and parameter.System adopts initiatively and the passive damping cooperative control method carries out damping control, and when changing between 0 to 1.6mH for the electrical network inductance, and damping coefficient changes and is no more than 20%.Filter parameter is as shown in table 1, and system parameters is as shown in table 2.

The parameter value of table 2LCL filter

2. set total damping coefficient ζ=0.1 of mixing damping, active damping and passive damping play impartial effect, i.e. active damping coefficient ζ _AD=0.05, the passive damping coefficient

ζ _PD=0.05。

3. get L ₁=1.6mH, L ₂=0.6mH, C _f=22 μ F, calculate the key parameter that mixes damping, wherein the capacitor current feedback coefficient of active damping:

$k=2{\mathrm{\&zeta;}}_{\mathrm{AD}}{L}_1\sqrt{\frac{{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+L_2}{L_1{L}_2{C}_f}}=1.6$

The damping resistance value of passive damping:

$R_d=\frac{2{\mathrm{\&zeta;}}_{\mathrm{PD}}}{{\mathrm{<figure-callout\; id="1"\; label="C\; f\; L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">C</figure-callout>}}_f\sqrt{\frac{L_{}}{}}\sqrt{\frac{{}_1+L_2}{L_1{L}_2{C}_f}}}0.45\mathrm{\&Omega;}$

4. the result in the step 3 is handled, damping resistance rounds up and gets R ' in the reality _d=0.5 Ω, then

${\mathrm{\&zeta;}}_{\mathrm{PD}}^{\&prime;}=\frac{R_d^{\&prime;}{\mathrm{<figure-callout\; id="1"\; label="C\; f\; L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">C</figure-callout>}}_f\sqrt{\frac{L_{}}{}}\sqrt{\frac{{}_1+L_2}{L_1({\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g){\mathrm{<figure-callout\; id="2"\; label="C\; f"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">C</figure-callout>}}_f}}}{2}=0.057$

Thereby, ζ ' _AD=0.1-ζ ' _PD=0.043, get L this moment _g=0, calculate k again and obtain, $k^{\&prime;}=2{\mathrm{\&zeta;}}_{\mathrm{AD}}^{\&prime;}{L}_1\sqrt{\frac{{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+L_2}{L_1{L}_2{C}_f}}=1.4.$

5. whether the damping system of check design adapts to electrical network variation inductance under the different access conditions, in the check electrical network variation inductance scope get (0,1.6mH], can satisfy examination requirements.In the check according to following formula make damping coefficient along with the electrical network variation inductance as shown in Figure 4:

$\mathrm{\&zeta;}=\frac{R_d^{\&prime;}{\mathrm{<figure-callout\; id="1"\; label="C\; f\; L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">C</figure-callout>}}_f\sqrt{\frac{L_{}}{}}\sqrt{\frac{{}_1+{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g}{L_1({\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g){\mathrm{<figure-callout\; id="2"\; label="C\; f"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">C</figure-callout>}}_f}}}{2}+\frac{k^{\&prime;}}{{2L}_1\sqrt{\frac{{\mathrm{<figure-callout\; id="1"\; label="L"\; filenames="HDA0000085880180000022.png,HDA0000085880180000031.png"\; state="\{\{state\}\}">L</figure-callout>}}_1+{\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g}{L_1({\mathrm{<figure-callout\; id="2"\; label="L"\; filenames="HDA0000085880180000022.png"\; state="\{\{state\}\}">L</figure-callout>}}_2+L_g)C_f}}},$

By accompanying drawing 4 as can be known, when the electrical network inductance changed between 0 to 1.6mH, total damping coefficient excursion was less than 20%, and when adopting passive damping merely, damping coefficient reduces along with the increase of electrical network inductance.

7. output design result, as shown in table 3;

Table 3 mixes damping design parameter value

8. the damping scheme of implementing is provided, as shown in Figure 7, in original photovoltaic grid-connected inverting system, for filter capacitor series connection resistance is the damping resistance of 0.5 Ω, and measures the filter capacitor current i _cBe amplified into electric current loop through 1.4 times and carry out active damping control simultaneously.

Claims (1)

1. LCL filter mixing damping parameter method for designing in the photovoltaic grid-connected inverting system is characterized in that, carries out design of Simulation in computer successively according to the following steps:

Step (1): the parameter of initialization LCL filter:

L ₁Inversion side inductance value for the LCL filter;

L ₂Some net side inductance value for the LCl filter;

C _fFilter capacitor value for the LCL filter;

The electrical network inductance L of access point _gThe variation span: less than 1.6mH;

Step (2): initialization total damping coefficient ζ, active damping coefficient ζ _ADAnd passive damping coefficient ζ _PD, wherein: ζ=ζ _AD+ ζ _PD:

ζ=0.1，ζ _AD=0.05，ζ _PD=0.05；

Calculate capacitor current feedback coefficient k and the passive damping resistance R of active damping according to following formula _d:

$k=2{\mathrm{\&zeta;}}_{\mathrm{AD}}{L}_1\sqrt{\frac{L_1+L_2}{L_1{L}_2{C}_f}};$

$R_d=\frac{2{\mathrm{\&zeta;}}_{\mathrm{PD}}}{C_f\sqrt{\frac{L_1+L_2}{L_1{L}_2{C}_f}}};$

Step (3): the passive damping resistance value that step (2) is obtained rounds up, and obtains the new value R ' that meets the resistance specification _d

Step (4): (1) result of obtaining is to initialized active damping coefficient ζ in the step (2) set by step _AD, passive damping coefficient ζ _PDRevise, obtain revised active damping coefficient ζ ' _ADWith revised passive damping coefficient ζ ' _PD:

${\mathrm{\&zeta;}}_{\mathrm{PD}}^{\&prime;}=\frac{R_d{C}_f\sqrt{\frac{L_1+L_2}{L_1{L}_2{C}_f}}}{2},$ Get L herein _g=0,

ζ ' _AD=0.1-ζ ' _PD, this moment ζ=0.1,

Step (5): the capacitor current feedback coefficient k of the active damping that obtains by following formula correction step (2), use k ' expression correction result:

$k^{\&prime;}=2{\mathrm{\&zeta;}}_{\mathrm{AD}}^{\&prime;}{L}_1\sqrt{\frac{L_1+L_2}{L_1{L}_2{C}_f}};$

Step (6): press total damping coefficient ζ described in the following formula plot step (2) at the revised damping resistance value R ' of passive damping _d, active damping capacitor current feedback coefficient correction value k ' and electrical network inductance L _gWhen 0 to 1.6mH changes, the change curve of described total damping coefficient ζ:

$\mathrm{\&zeta;}=\frac{R_d^{\&prime;}{C}_f\sqrt{\frac{L_1+L_2+L_g}{L_1(L_2+L_g)C_f}}}{2}+\frac{k^{\&prime;}}{{2L}_1\sqrt{\frac{L_1+l_2+L_g}{L_1(L_2+L_g)C_f}}},$

If: the undulating value that total damping coefficient ζ changes with the electrical network inductance surpasses 20% o'clock of total damping coefficient settings value that step (2) sets, then returns step (2), under the constant prerequisite of the value of described total damping coefficient ζ, to passive damping coefficient ζ _PDWith active damping coefficient ζ _ADRedistribute;

If: total damping coefficient ζ is no more than 20% of total damping coefficient value ζ that step (2) sets with electrical network variation inductance undulating value, then changes step (7) over to;

Step (7): the following damping parameter that mixes of output:

Total damping coefficient ζ=0.1, revised active damping coefficient ζ ' _AD, revised passive damping coefficient ζ ' _PD, revised passive damping resistance R ' _dAnd the capacitor current feedback coefficient k of revised active damping '.

Images