CN107425743A - A kind of combining inverter MPC methods based on prediction deviation feedback correction - Google Patents
A kind of combining inverter MPC methods based on prediction deviation feedback correction Download PDFInfo
- Publication number
- CN107425743A CN107425743A CN201710315813.3A CN201710315813A CN107425743A CN 107425743 A CN107425743 A CN 107425743A CN 201710315813 A CN201710315813 A CN 201710315813A CN 107425743 A CN107425743 A CN 107425743A
- Authority
- CN
- China
- Prior art keywords
- grid
- moment
- prediction deviation
- connected current
- voltage
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000012937 correction Methods 0.000 title claims abstract description 28
- 239000013598 vector Substances 0.000 claims abstract description 66
- 239000003990 capacitor Substances 0.000 claims abstract description 7
- 230000008569 process Effects 0.000 claims description 14
- 238000006243 chemical reaction Methods 0.000 claims description 13
- 230000001360 synchronised effect Effects 0.000 claims description 10
- 238000005070 sampling Methods 0.000 claims description 8
- 230000005611 electricity Effects 0.000 claims description 4
- 230000003068 static effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 10
- 230000008859 change Effects 0.000 abstract description 6
- 230000009897 systematic effect Effects 0.000 abstract description 4
- 230000003044 adaptive effect Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 4
- 238000011217 control strategy Methods 0.000 description 3
- 238000002474 experimental method Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 238000007630 basic procedure Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 239000000470 constituent Substances 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005530 etching Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 230000005764 inhibitory process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
- H02M7/00—Conversion of ac power input into dc power output; Conversion of dc power input into ac power output
- H02M7/42—Conversion of dc power input into ac power output without possibility of reversal
- H02M7/44—Conversion of dc power input into ac power output without possibility of reversal by static converters
- H02M7/48—Conversion of dc power input into ac power output without possibility of reversal by static converters using discharge tubes with control electrode or semiconductor devices with control electrode
-
- H02J3/382—
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
- H02M1/00—Details of apparatus for conversion
- H02M1/0003—Details of control, feedback or regulation circuits
- H02M1/0012—Control circuits using digital or numerical techniques
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Inverter Devices (AREA)
Abstract
The invention discloses a kind of combining inverter MPC methods based on prediction deviation feedback correction, controlled using outer voltage, current inner loop MPC controls, phase locked-loop unit obtains line voltage azimuth θ to realize coordinate transform；Outer voltage exports grid-connected current set-point according to PWM converter DC capacitor voltageCurrent inner loop MPC control units are according to k+1 moment grid-connected current set-pointsWith k moment grid-connected current actual valuesAnd the feedback quantity of prediction deviation feedback unit, obtain voltage vector u of the inverter at k to the k+1 momentc(α,β)(k)；The voltage vector u that will be obtainedc(α,β)(k) it is converted into the signal s of driving combining inverter power deviceabc.The prediction deviation feedback unit of the present invention can realize adaptive corrective when systematic parameter mismatches, when the other reasonses such as load changing, temperature rise in combining inverter cause the grid-connected inductance and resistance to change, good current tracking effect is kept, improves robustness during combining inverter parameter variations.
Description
Technical field
It is particularly a kind of based on the grid-connected of prediction deviation feedback correction the present invention relates to a kind of control method of grid-connected inverter
Inverter MPC methods.
Background technology
With the development of the new energy such as photovoltaic, wind energy, distributed generation system is widely used, and combining inverter
The conversion equipment being connected to the grid as new energy, extensive concern and research are also obtained.Model Predictive Control (MPC) is used as one
Kind nonlinear control techniques, value letter is defined with its quick response speed, simple implementation and for restrictive condition
Several direct processing modes occupy an important position in the control mode of combining inverter.But conventional MPC control modes
It in system practice, usually can be deteriorated serious by the unmatched influence of parameter, control performance.Therefore during parameter variations,
The high-performance stability problem urgent need to resolve of the combining inverter of MPC controls.
At present, have for the way to solve the problem：Luenberger observers Feedforward Compensation, extended state observer
Feedforward Compensation, using simplified inductance observer revised law.Eliminated using Luenberger observers by feedforward compensation
System disturbance, but the control strategy is based on the unmatched problem of parameter in fixed-frequency-type Model Predictive Control, is not suitable for herein
Finite State Model PREDICTIVE CONTROL；Actual disturbance is estimated using extended state observer, and disturbance shadow is eliminated by feedforward compensation
Ring, but the algorithm structure is complicated, time-consuming for calculating, it is high to the performance requirement of controller.Using simplified inductance observer, profit
With the observation Real Time Correction System inductance parameters of inductance, robustness of the system when circuit inductance L changes is improved.But should
Strategy have ignored the influence of circuit impedance R variations.
The content of the invention
Goal of the invention：In view of the shortcomings of the prior art, a kind of combining inverter based on prediction deviation feedback correction is proposed
MPC methods, solve the problems, such as that grid-connected current distortion aggravates when systematic parameter mismatches, and improves the robustness of system.
Technical scheme：A kind of combining inverter MPC methods based on prediction deviation feedback correction, comprise the following steps：
Step 1: sampling three-phase line voltage usa、usb、uscWith grid-connected current ia、ib、ic, by Clark conversion respectively
Line voltage u under to two-phase rest framesα、usβWith grid-connected current iα、iβ；
Step 2: line voltage azimuth θ is obtained by phaselocked loop, using voltage vector angle θ to two-phase rest frame
Lower line voltage usα、usβPark conversion is carried out, obtains d, q component u of line voltage under synchronous rotating framed、uq；
Step 3: sampling DC capacitor voltage udc, grid-connected current d axles are obtained by determining DC voltage outer loop-control unit
The set-point of componentAnd the set-point of grid-connected current q axis components0 is set to, two-phase static coordinate is changed into their contravariant
The grid-connected current set-point at k+1 moment is obtained under system
Step 4: the k+1 moment grid-connected current set-points that step 3 is obtainedThe k moment obtained by step 1
Grid-connected current value iα,β(k) current inner loop MPC control units (9) are given, are sweared so as to obtain output voltage of the inverter at the k moment
Measure uc(α,β)(k) and the k+1 moment grid-connected current predicted value
Step 5: the grid-connected current predicted value that step 4 is obtainedWith output voltage vector uc(α,β)(k) and
The grid-connected current value i that step 1 obtainsα,β(k) prediction deviation feedback unit, is sent to, obtains the prediction deviation △ at k+1 momentα,β(k
+ 1), and by the prediction deviation at k+1 moment it is updated in the current inner loop MPC control units of step 4, to the electric current in step 4
Inner ring MPC control units are modified；
Step 6: the output voltage vector u that step 4 is obtainedc(α,β)(k) it is converted into the work(of driving combining inverter bridge arm
Rate device control signal sabc。
Further, the acquisition of the line voltage azimuth and to line voltage carry out Park conversion processes：
The line voltage azimuth θ that step 2.1 is fed back to obtain using phaselocked loop, to line voltage under two-phase rest frame
usα、usβPark conversion is carried out, obtains d, q component u of line voltage under synchronous rotating framed、uq；
Step 2.2 is by specified rateWith detecting obtained line voltage q axis components uqSubtract each other, by pi regulator,
Along with angular speed 314rad/s, by, with 2 π modulus, obtaining line voltage azimuth θ after an integrator.
Further, the DC voltage outer shroud control process of the step 3 is：With the set-point of DC voltageSubtract
DC capacitor voltage udc, the set-point of grid-connected current d axis components under synchronous rotating frame is obtained by pi regulator
Further, the current inner loop MPC control process of the step 4 is：
The k moment grid-connected currents i that step 4.1 obtains all output vectors of combining inverter and step 1α,β(k) pass through
MPC control units computing one by one, so as to obtain inverter in k moment all output voltage vectors and with right after this vector
The grid-connected current predicted value for the subsequent time answered；
Step 4.2 substitutes into the output iteration in last moment prediction deviation feedback unit, corrects and is calculated in step 4.1
The grid-connected current predicted value arrived, retrieve one group of corresponding number of combining inverter output voltage vector and grid-connected current predicted value
According to；
The k+1 moment grid-connected current set-points that step 4.3 obtains step 3With the institute obtained in step 4.2
There is grid-connected current predicted value to compare, take closest to inverter output vector conduct corresponding to the grid-connected current predicted value of set-point
The output voltage vector u at k momentc(α,β)(k) the grid-connected current predicted value, taken is designated as
Further, the acquisition of prediction deviation and the process of feedback control are in the step 5：
Grid-connected current predicted value vector under the k moment two-phase rest frames that step 5.1 obtains step 4.3
Subtract the grid-connected current value vector i under the two-phase rest frame obtained by step 1α,β(k) prediction deviation at k moment, is obtained
△α,β(k)；
Grid-connected current predicted value vector under the k-1 moment two-phase rest frames that step 5.2 obtains step 4.3Subtract the grid-connected current value vector i under the two-phase rest frame obtained by step 1α,β(k-1) the k moment, is obtained
Prediction deviation △α,β(k-1)；
Step 5.3 judges the difference u of k and k-1 inverter output voltagec(α,β)(k-1)-uc(α,β)(k-2) whether it is more than and sets
The threshold value put, if being more than threshold value, the k moment that step 5.1 and step 5.2 are obtained and the prediction deviation at k-1 moment really
Make it is poor, then divided by the two moment inverter output voltage difference, obtain amendment of the inverter output voltage to prediction deviation
COEFFICIENT Kαβ；If the difference of the two moment inverter output voltages is not more than threshold value, the amendment system that last computation obtains is continued to use
Number Kαβ；
It is poor that step 5.4 makees k moment inverter output voltage and k+1 moment inverter output voltage, is multiplied by correct
Number Kαβ, obtained result adds the prediction deviation △ at k momentα,β(k) prediction deviation at k+1 moment, is obtained, and using it as pre-
Survey the output of deviation feedback unit.
Further, the inverter output vector u that will be obtained in step 4c(α,β)(k) it is minimum according to bridge arm switch motion
Principle, distribute the power device control signal s of each combining inverter bridge armabc。
Beneficial effect：The prediction deviation feedback correction unit of the present invention, it can be realized when systematic parameter mismatches adaptive
Correction, when the other reasonses such as the load changing temperature rise of combining inverter cause grid-connected inductance and resistance to change, keep good
Good current tracking effect, the THD of grid-connected current is reduced, improve robust during the grid-connected reactor parameter variations of combining inverter
Property.
Brief description of the drawings
Fig. 1 is the control structure figure of the combining inverter MPC methods based on prediction deviation feedback correction of the present invention.
Fig. 2 is the prediction deviation feedback correction MPC algorithm structure chart of the present invention.
Fig. 3 is the prediction deviation feedback unit structure chart of the present invention.
Fig. 4 is the topology diagram of two level grid-connected inverters of the present invention.
Fig. 5 is grid-connected current and prediction deviation figure before and after the inductance parameters of the present invention mismatch.
The inductance parameters that Fig. 6 is the present invention mismatch front and rear feedback factor recognition effect figure.
Fig. 7 is that the inductance parameters of the present invention add prediction deviation figure before and after inventive algorithm when mismatching.
Fig. 8 is a phase grid-connected current tracking effect figures added when the inductance parameters of the present invention mismatch before and after algorithm.
Fig. 9 is that the resistance parameter of the present invention adds prediction deviation figure before and after algorithm when mismatching.
Figure 10 is that the resistance parameter of the present invention adds a phase grid-connected current tracking effect figures before and after algorithm when mismatching.
When Figure 11 is that inductive resistance parameter matches, the prediction deviation figure before and after addition algorithm.
In Fig. 1,1, electric network source；2nd, grid-connected resistance and inductance；3rd, three-phase PWM current transformer；4th, DC load；5th, power network electricity
The detection unit of pressure；6th, the detection unit of current on line side；7th, soft phase-locked loop unit；8th, DC voltage outer loop-control unit；9、MPC
Control unit；10th, prediction deviation feedback unit；11st, switching signal converting unit.
Embodiment
Further explanation is done to the present invention below in conjunction with the accompanying drawings.
A kind of combining inverter MPC methods based on prediction deviation feedback correction, comprise the following steps：
Step 1: sampling three-phase line voltage usa、usb、uscWith grid-connected current ia、ib、ic, by Clark conversion respectively
Line voltage u under to two-phase rest framesα、usβWith grid-connected current iα、iβ；
Step 2: line voltage azimuth θ is obtained by phaselocked loop 7, using voltage vector angle θ to two-phase rest frame
Lower line voltage usα、usβPark conversion is carried out, obtains d, q component u of line voltage under synchronous rotating framed、uq；
Step 3: sampling DC capacitor voltage udc, grid-connected current d is obtained by determining DC voltage outer loop-control unit 8
The set-point of axis componentAnd the set-point of grid-connected current q axis components0 is set to, the static seat of two-phase is changed into their contravariant
The grid-connected current set-point at k+1 moment is obtained under mark system
Step 4: the k+1 moment grid-connected current set-points that step 3 is obtainedThe k moment obtained by step 1
Grid-connected current value iα,β(k) current inner loop MPC control units 9 are given, so as to obtain output voltage vector of the inverter at the k moment
uc(α,β)(k) and the k+1 moment grid-connected current predicted value
Step 5: the grid-connected current predicted value that step 4 is obtainedWith output voltage vector uc(α,β)(k) and
The grid-connected current value i that step 1 obtainsα,β(k) prediction deviation feedback unit 10, is sent to, obtains the prediction deviation △ at k+1 momentα,β
(k+1), and by the prediction deviation at k+1 moment it is updated in the current inner loop MPC control units 9 of step 4, in step 4
Current inner loop MPC control units 9 are modified；
Step 6: the output voltage vector u that step 4 is obtainedc(α,β)(k) according to the minimum principle of bridge arm switch motion,
It is converted into the power device control signal s of driving combining inverter bridge armabc。
Wherein, the acquisition of line voltage azimuth and to line voltage carry out Park conversion processes：
Step 2.1 feeds back obtained line voltage azimuth θ using phaselocked loop 7, to power network electricity under two-phase rest frame
Press usα、usβPark conversion is carried out, obtains d, q component u of line voltage under synchronous rotating framed、uq；
Step 2.2 is by specified rateWith detecting obtained line voltage q axis components uqSubtract each other, by pi regulator,
Along with angular speed 314rad/s, by, with 2 π modulus, obtaining line voltage azimuth θ after an integrator.
The DC voltage outer shroud control process of step 3 is：With the set-point of DC voltageSubtract DC capacitor voltage
udc, the set-point of grid-connected current d axis components under synchronous rotating frame is obtained by pi regulator
The current inner loop MPC control process of step 4 is：
The k moment grid-connected currents i that step 4.1 obtains all output vectors of combining inverter and step 1α,β(k) pass through
The computing one by one of MPC control units 9, so as to obtain inverter after k moment all output voltage vectors and with this vector
The grid-connected current predicted value of corresponding subsequent time；
Step 4.2 substitutes into the output iteration in last moment prediction deviation feedback unit 10, corrects and is calculated in step 4.1
Obtained grid-connected current predicted value, retrieve combining inverter output voltage vector and grid-connected current predicted value one group are corresponding
Data；
The k+1 moment grid-connected current set-points that step 4.3 obtains step 3With the institute obtained in step 4.2
There is grid-connected current predicted value to compare, take closest to inverter output vector conduct corresponding to the grid-connected current predicted value of set-point
The output voltage vector u at k momentc(α,β)(k) the grid-connected current predicted value, taken is designated as
The acquisition of prediction deviation and the process of feedback control are in step 5：
Grid-connected current predicted value vector under the k moment two-phase rest frames that step 5.1 obtains step 4.3
Subtract the grid-connected current value vector i under the two-phase rest frame obtained by step 1α,β(k) prediction deviation at k moment, is obtained
△α,β(k)；
Grid-connected current predicted value vector under the k-1 moment two-phase rest frames that step 5.2 obtains step 4.3Subtract the grid-connected current value vector i under the two-phase rest frame obtained by step 1α,β(k-1) the k moment, is obtained
Prediction deviation △α,β(k-1)；
Step 5.3 judges the difference u of k and k-1 inverter output voltagec(α,β)(k-1)-uc(α,β)(k-2) whether it is more than and sets
The threshold value put, if being more than threshold value, the k moment that step 5.1 and step 5.2 are obtained and the prediction deviation at k-1 moment really
Make it is poor, then divided by the two moment inverter output voltage difference, obtain amendment of the inverter output voltage to prediction deviation
COEFFICIENT Kαβ；If the difference of the two moment inverter output voltages is not more than threshold value, the amendment system that last computation obtains is continued to use
Number Kαβ；
It is poor that step 5.4 makees k moment inverter output voltage and k+1 moment inverter output voltage, is multiplied by correct
Number Kαβ, obtained result adds the prediction deviation △ at k momentα,β(k) prediction deviation at k+1 moment, is obtained, and using it as pre-
Survey the output of deviation feedback unit.
Model Predictive Control strategy basic procedure：
The topology of three-phase voltage type synchronization inverter is as shown in Figure 4.Wherein usa、usb、uscFor three-phase power grid voltage, ia、ib、
icFor three-phase grid electric current, uca、ucb、uccThe phase voltage exported for inverter, R and L are respectively the resistance and electricity of grid-connected reactor
Sense.
Three-phase grid-connected inverter mathematical modeling is established under two-phase rest frame：
(1) is write into as vector form：
Wherein ucRepresent inverter output voltage vector, usLine voltage vector is represented, i represents grid-connected current vector, R and L
The resistance and inductance of respectively grid-connected reactor.
The prediction of grid-connected current is the discrete model based on system, therefore needs to use forward-difference method to system equation (2)
Discretization.Assuming that the sampling period of system is Ts, then it is approximately in k moment di/dt：
Obtain the discretization model of system：
In three-phase grid-connected inverter, the major control target of Model Predictive Control is grid-connected current.Predictive control algorithm
Control block diagram it is as shown in Figure 3.
By the grid-connected current value i for detecting the k momentα,βAnd line voltage value u (k)s(α,β)(k) it is, possible with reference to each
Inverter output voltage vector uc(α,β)(k) corresponding subsequent time grid-connected current value, is predicted by system discretization equation.
By the way that grid-connected current predicted value and command current value are obtained into cost function J as difference：
Select output quantity of the switching value as inverter corresponding to a minimum vector of cost function sum.
The mismatch problem of forecast model and real system：
Forecast model as the key link in Model Predictive Control strategy, close very much by the uniformity of itself and real system
Key.When forecast model, which establishes the factors such as inaccuracy, Parameters variation, to be influenceed, the current value of predictionAnd command current valueRelatively large deviation will be produced, this deviation can cause falsely dropping for inverter switching device amount to be selected, i.e., selected output voltage vector S
The minimum switching values of evaluation function J are not so that, and then cause the current track error of system to increase.
Prediction deviation mainly includes two aspects：1 is prediction model parameterses and actual thing caused by circuit R, L Parameters variation
Manage the mismatch of parameter；2 be the deviation of forecast model and real system continuous model based on system discretization equation.
Prediction deviation analysis when system line parameter mismatches：
The prediction deviation of system, first ignores system delay and discretization causes when being mismatched for independent analysis circuit parameter
Prediction deviation.Define R0、L0The line resistance and inductance parameters respectively set in forecast model, R and L are that circuit is actual
Resistance and inductance parameters, wherein WithRepresent actual resistance and inductance relative to model respectively
The departure of parameter.Then predicted current value of the system at the k+1 moment is：
And the actual electric current at system k+1 moment when line parameter circuit value mismatches should be as follows：
The current forecasting deviation at the k+1 moment of definition system is △k+1,
The prediction deviation in (6) and (7) substitution (8), obtaining system when line parameter circuit value mismatches is as follows,
Formula (9) illustrates that, when line parameter circuit value mismatches, current forecasting deviation not only has with actual R, L parameter of circuit
Close, and it is relevant with grid-connected current vector, the voltage output vector of line voltage vector and Systematic selection.It is inclined in model parameter
Residual quantityWith, can be according to the changing rule of prediction deviation heterogeneity by current forecasting deviation △ in the case of certainα,β(k+
1) it is divided intoWithTwo parts：
WhereinThat an amplitude is fixed, the vector of phase angle mechanical periodicity, it only and grid-connected current vector and
Line voltage vector correlation.Line voltage vector is usually that amplitude is fixed in the case of bulk power grid, and phase is with power frequency 50Hz speed
Degree turns clockwise；And grid-connected current vector amplitude approximately constant in chain type STATCOM steady-state operations, angular velocity of rotation also and
Line voltage is identical.So be transformed under alpha-beta coordinate system,Then become amplitude necessarily and will change similar to 50Hz
Simple alternating current amount.
AndThen be prediction deviation in aperiodic changing unit, it only and chain type STATCOM output voltage
Vector correlation, chain type STATCOM output voltage vectors in steady state operation are saltus steps, so being transformed into alpha-beta coordinate system
UnderAnd with output voltage vector saltus step.
The prediction deviation analysis that MPC models discretization process introduces：
MPC controls the discrete equation model formation (4) based on system, in chain type STATCOM, Model Predictive Control Algorithm
The grid-connected current of subsequent time is predicted by the discrete equation of system.But actual three-phase grid-connected inverter be not one from
Scattered system, but a continuous system, in a discrete periodic TsIn, grid-connected current and line voltage are consecutive variations
's.The grid-connected predicted value of this subsequent time for allowing for obtaining based on system discrete model and actual value generate deviation.In order to
This deviation of labor, it is assumed here that circuit system parameter matches.
In a PREDICTIVE CONTROL cycle, the predicted current at k+1 moment can be written as：
Wherein TsFor PREDICTIVE CONTROL cycle, in a PREDICTIVE CONTROL cycle, the u in system models(α,β)And i (k)α,β
(k) it is constant.Grid-connected current iα,βWith the same phase of line voltage.
And shown in the actual change equation such as formula (2) of system power, in a PREDICTIVE CONTROL cycle, real system
uS (α, β)And iα, βIt is consecutive variations, for the ease of comparative analysis, by predetermined period TsIt is divided into M parts：
Formula (13) is added item by item, obtained
When M is intended to infinity, formula (14) represents that system in the actual value of k+1 moment electric currents, is write as：
The prediction deviation that discretization model introduces
It can be found that the difference according to the actual current value of etching system when the k+1 moment predicted current values and k+1 of model calculating
For △ (k+1), it is also that an amplitude is fixed, and phase is transformed into power frequency 50Hz and grid-connected current synchronous rotary vector
Under alpha-beta coordinate system, the predicted current deviation is also the simple alternating current amount that an amplitude is certain and changes into 50HZ.In order to it is previous
The prediction deviation of partial analysis makes a distinction, the deviation △ that will be introduced here by system model discretizationα,β(k+1) it is expressed as
Prediction deviation constituent analysis：
The current forecasting total deviation that system can be obtained in summary at the k+1 moment is：
WhereinAll it is the 50Hz Stable State of Sine part in k+1 moment current forecasting deviations.They
Vector be transformed under alpha-beta coordinate system the Stable State of Sine current forecasting deviation of the form that can be written as, referred to as k moment.
AndIt is then the ladder hopping part in k+1 moment current forecasting deviations, being transformed into can under alpha-beta coordinate system
In the form of being write as (17), the referred to as ladder trip current prediction deviation at k moment.
Wherein A is in certain discrete periodic Ts, sampling period TmAnd keep constant under certain line parameter circuit value, and discrete week
Phase TsAlso very little, therefore in the adjacent cycle of PREDICTIVE CONTROL twice k+1 and k+2, the fractional prediction deviation approximation is constant.
In the adjacent cycle of PREDICTIVE CONTROL twice k+1 and k+2, due to combining inverter output voltage vector difference compared with
Greatly, soChanged greatly in the adjacent cycle of PREDICTIVE CONTROL twice.
Model Predictive Control based on prediction deviation feedback correction：
In order to eliminate three-phase grid-connected inverter current forecasting deviation, propose here a kind of based on prediction deviation feedback correction
MPC algorithm.Cardinal principle is the prediction deviation value △ by last momentα,β(k) the prediction deviation value of prediction subsequent time is gone
△α,β(k+1), and by this value it is updated in system model, completes the correction to subsequent time prediction deviation.Control block diagram is such as
Fig. 2：
The prediction deviation at any k moment and k+1 moment is according to knowable to the analysis of upper section
Due to prediction deviation Stable State of Sine part in predetermined period twice in succession it is approximate constant, you can to think：
So formula (20) is expressed as form：
And then obtain
△α,β(k+1)=△α,β(k)+Kα,β(uc(α,β)(k+1)-uc(α,β)(k)) (23)
That is prediction deviation of the system at the k+1 moment, it is that its prediction deviation at the k moment is added by k and k+1 moment inverters
One correction of current deviation caused by output voltage difference, wherein Kα,βIt is inverter output voltage under alpha-beta coordinate system to electric current
The correction factor of prediction deviation：
It can be seen that the K in the case where system line parameter and PREDICTIVE CONTROL cycle determineα,βFor a constant.Due toIt is system
Line inductance disturbance quantity, be uncertain, to calculate it and introduce state observer again so that system is more complicated.Consider
In adjacent several PREDICTIVE CONTROL cycles, Kα,βMay be considered it is metastable, so here pass through the first two PREDICTIVE CONTROL week
Interim data calculate Kα,βValue：
But so direct real-time update calculates and the phenomenon that denominator is zero occurs, causes the unstable of system, considers Kα,β
Change mainly caused by line inductance Parameters variation, and the change frequency of line inductance is far below pre- measured frequency, so here
Only in denominator uc(α,β)(k)-uc(α,β)(k-1) to K when being more than the threshold value of settingα,βCalculating renewal is carried out, otherwise continues to use the last time
The K being calculatedα,β.Control block diagram such as 3：
Substituted into after obtaining the prediction deviation of subsequent time in forecast model, correct original model, to cause reality
Prediction deviation reduce, reach good PREDICTIVE CONTROL effect.
In order to verify above-mentioned decoupling control method, build two level PWM rectifiers and carry out experimental verification.Wherein, core
Controller uses Infineon K15T1202 for high-performance the NI cRIO-9024, IGBT of NI companies, and driving chip is
IR2233。
Specifically experiment parameter is：
The experiment parameter of table 1
When inductance actual value is 6mH, grid-connected inductance nominal value from 6mH sport 2mH when, grid-connected current prediction deviation is obvious
Become big, grid-connected current wave distortion is seriously such as Fig. 5.Correction factor of the inverter output voltage identified to current forecasting deviation
Such as Fig. 6, it is seen that carry the inverter output voltage in algorithm to current forecasting drift correction COEFFICIENT KαRecognition effect it is good.Add
After entering carried prediction deviation feedback correction algorithm, prediction deviation substantially reduces such as Fig. 7.After adding algorithm, due to prediction deviation
Reduce, therefore grid-connected current irregularity of wave form is significantly reduced, current tracking effect improves such as Fig. 8.
In order to verify that carried algorithm matches to the unmatched inhibition of resistance parameter, the grid-connected inductance parameters of design system,
Grid-connected resistance actual value is 1 ohm, when nominal value is 10 ohm, adds current forecasting deviation such as Fig. 9 before and after algorithm, it is seen that prediction
Deviation effectively reduces, and grid-connected current tracking effect Figure 10, current tracking effect improves.
When line parameter circuit value all matches, the prediction deviation before and after addition algorithm is as follows, it is seen that system is still deposited when parameter matches
The prediction deviation introduced in discretization process, this fractional prediction deviation and PREDICTIVE CONTROL cycle, i.e. system discretization cycle have
Close, pre- measured frequency is 10 KHzs in this experiment, so fractional prediction deviation is smaller, has still been obtained effectively after adding algorithm
Suppress, such as Figure 11.
Described above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (6)
- A kind of 1. combining inverter MPC methods based on prediction deviation feedback correction, it is characterised in that comprise the following steps：Step 1: sampling three-phase line voltage usa、usb、uscWith grid-connected current ia、ib、ic, two are respectively obtained by Clark conversion Line voltage u under phase rest framesα、usβWith grid-connected current iα、iβ；Step 2: line voltage azimuth θ is obtained by phaselocked loop (7), using voltage vector angle θ under two-phase rest frame Line voltage usα、usβPark conversion is carried out, obtains d, q component u of line voltage under synchronous rotating framed、uq；Step 3: sampling DC capacitor voltage udc, grid-connected current d axles are obtained by determining DC voltage outer loop-control unit (8) The set-point of componentAnd the set-point of grid-connected current q axis components0 is set to, two-phase static coordinate is changed into their contravariant The grid-connected current set-point at k+1 moment is obtained under systemStep 4: the k+1 moment grid-connected current set-points that step 3 is obtainedThe k moment obtained by step 1 is grid-connected Current value iα,β(k) current inner loop MPC control units (9) are given, so as to obtain output voltage vector of the inverter at the k moment uc(α,β)(k) and the k+1 moment grid-connected current predicted valueStep 5: the grid-connected current predicted value that step 4 is obtainedWith output voltage vector uc(α,β)And step 1 (k) Obtained grid-connected current value iα,β(k) prediction deviation feedback unit (10), is sent to, obtains the prediction deviation △ at k+1 momentα,β(k+ 1), and by the prediction deviation at k+1 moment it is updated in the current inner loop MPC control units (9) of step 4, to the electricity in step 4 Stream inner ring MPC control units (9) are modified；Step 6: the output voltage vector u that step 4 is obtainedc(α,β)(k) it is converted into the power device of driving combining inverter bridge arm Part control signal sabc。
- 2. a kind of combining inverter MPC methods based on prediction deviation feedback correction according to claim 1, its feature exist In the acquisition of the line voltage azimuth simultaneously carries out Park conversion processes to line voltage：The line voltage azimuth θ that step 2.1 is obtained using phaselocked loop (7) feedback, to line voltage under two-phase rest frame usα、usβPark conversion is carried out, obtains d, q component u of line voltage under synchronous rotating framed、uq；Step 2.2 is by specified rateWith detecting obtained line voltage q axis components uqSubtract each other, by pi regulator, add Angular speed 314rad/s, by, with 2 π modulus, obtaining line voltage azimuth θ after an integrator.
- 3. a kind of combining inverter MPC methods based on prediction deviation feedback correction according to claim 1, its feature exist In the DC voltage outer shroud control process of the step 3 is：With the set-point of DC voltageSubtract DC capacitor voltage udc, the set-point of grid-connected current d axis components under synchronous rotating frame is obtained by pi regulator
- 4. a kind of combining inverter MPC methods based on prediction deviation feedback correction according to claim 1, its feature exist In the current inner loop MPC control process of the step 4 is：The k moment grid-connected currents i that step 4.1 obtains all output vectors of combining inverter and step 1α,β(k) controlled by MPC Unit (9) computing one by one processed, corresponded to so as to obtain inverter after k moment all output voltage vectors and with this vector Subsequent time grid-connected current predicted value；Step 4.2 substitutes into the output iteration in last moment prediction deviation feedback unit (10), corrects and is calculated in step 4.1 The grid-connected current predicted value arrived, retrieve one group of corresponding number of combining inverter output voltage vector and grid-connected current predicted value According to；The k+1 moment grid-connected current set-points that step 4.3 obtains step 3With obtained in step 4.2 it is all simultaneously Net current forecasting value compares, and takes closest to when inverter output vector is as k corresponding to the grid-connected current predicted value of set-point The output voltage vector u at quarterc(α,β)(k) the grid-connected current predicted value, taken is designated as
- 5. a kind of combining inverter MPC methods based on prediction deviation feedback correction according to claim 1, its feature exist In the acquisition of prediction deviation and the process of feedback control are in the step 5：Grid-connected current predicted value vector under the k moment two-phase rest frames that step 5.1 obtains step 4.3Subtract by Grid-connected current value vector i under the two-phase rest frame that step 1 obtainsα,β(k) the prediction deviation △ at k moment, is obtainedα,β (k)；Grid-connected current predicted value vector under the k-1 moment two-phase rest frames that step 5.2 obtains step 4.3Subtract The grid-connected current value vector i gone under the two-phase rest frame that is obtained by step 1α,β(k-1) prediction deviation at k moment, is obtained △α,β(k-1)；Step 5.3 judges the difference u of k and k-1 inverter output voltagec(α,β)(k-1)-uc(α,β)(k-2) whether it is more than what is set Threshold value, if being more than threshold value really, the prediction deviation work at the k moment that step 5.1 and step 5.2 are obtained and k-1 moment is poor, Again divided by the inverter output voltage at the two moment difference, obtain correction factor of the inverter output voltage to prediction deviation Kαβ；If the difference of the two moment inverter output voltages is not more than threshold value, the correction factor that last computation obtains is continued to use Kαβ；It is poor that step 5.4 makees k moment inverter output voltage and k+1 moment inverter output voltage, multiplied by with adjusted coefficient Kαβ, Obtained result adds the prediction deviation △ at k momentα,β(k) prediction deviation at k+1 moment, is obtained, and using it as prediction deviation The output of feedback unit.
- 6. a kind of combining inverter MPC methods based on prediction deviation feedback correction according to claim 1, its feature exist In the inverter output vector u that will be obtained in step 4c(α,β)(k) it is each according to the minimum principle of bridge arm switch motion, distribution The power device control signal s of combining inverter bridge armabc。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710315813.3A CN107425743B (en) | 2017-05-08 | 2017-05-08 | A kind of gird-connected inverter MPC method based on prediction deviation feedback correction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710315813.3A CN107425743B (en) | 2017-05-08 | 2017-05-08 | A kind of gird-connected inverter MPC method based on prediction deviation feedback correction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107425743A true CN107425743A (en) | 2017-12-01 |
| CN107425743B CN107425743B (en) | 2019-09-27 |
Family
ID=60424991
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710315813.3A Active CN107425743B (en) | 2017-05-08 | 2017-05-08 | A kind of gird-connected inverter MPC method based on prediction deviation feedback correction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107425743B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109861565A (en) * | 2019-01-25 | 2019-06-07 | 湖南大学 | A kind of Model Reference Adaptive Control Method of two-stage type AC/DC converter |
| CN110417329A (en) * | 2018-04-28 | 2019-11-05 | 株洲中车时代电气股份有限公司 | A kind of trailer system |
| CN111864793A (en) * | 2020-05-25 | 2020-10-30 | 国网江西省电力有限公司经济技术研究院 | Power grid control method based on model prediction algorithm and oriented to power energy Internet |
| CN112886609A (en) * | 2021-01-15 | 2021-06-01 | 北方工业大学 | Alternating current constant power load optimization control model and optimization control method |
| CN112994109A (en) * | 2021-04-25 | 2021-06-18 | 郑州轻工业大学 | LC filtering type grid-connected inverter weighting sliding mode model prediction capacitor voltage control method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100259204A1 (en) * | 2009-04-10 | 2010-10-14 | Denso Corporation | Control device for electric rotating machine |
| CN102916600A (en) * | 2012-10-26 | 2013-02-06 | 河南师范大学 | Self-correcting prediction control method of model of three-phase voltage type PWM (Pulse-Width Modulation) rectifier |
| CN104779830A (en) * | 2015-04-29 | 2015-07-15 | 厦门大学 | Variable-dead-time inversion control method |
-
2017
- 2017-05-08 CN CN201710315813.3A patent/CN107425743B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100259204A1 (en) * | 2009-04-10 | 2010-10-14 | Denso Corporation | Control device for electric rotating machine |
| CN102916600A (en) * | 2012-10-26 | 2013-02-06 | 河南师范大学 | Self-correcting prediction control method of model of three-phase voltage type PWM (Pulse-Width Modulation) rectifier |
| CN104779830A (en) * | 2015-04-29 | 2015-07-15 | 厦门大学 | Variable-dead-time inversion control method |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110417329A (en) * | 2018-04-28 | 2019-11-05 | 株洲中车时代电气股份有限公司 | A kind of trailer system |
| CN110417329B (en) * | 2018-04-28 | 2021-05-28 | 株洲中车时代电气股份有限公司 | Traction system |
| CN109861565A (en) * | 2019-01-25 | 2019-06-07 | 湖南大学 | A kind of Model Reference Adaptive Control Method of two-stage type AC/DC converter |
| CN109861565B (en) * | 2019-01-25 | 2020-11-06 | 湖南大学 | Model reference self-adaptive control method of two-stage AC/DC converter |
| CN111864793A (en) * | 2020-05-25 | 2020-10-30 | 国网江西省电力有限公司经济技术研究院 | Power grid control method based on model prediction algorithm and oriented to power energy Internet |
| CN111864793B (en) * | 2020-05-25 | 2023-12-05 | 国网江西省电力有限公司经济技术研究院 | Power grid control method based on model prediction algorithm and oriented to electric power energy Internet |
| CN112886609A (en) * | 2021-01-15 | 2021-06-01 | 北方工业大学 | Alternating current constant power load optimization control model and optimization control method |
| CN112994109A (en) * | 2021-04-25 | 2021-06-18 | 郑州轻工业大学 | LC filtering type grid-connected inverter weighting sliding mode model prediction capacitor voltage control method |
| CN112994109B (en) * | 2021-04-25 | 2023-03-14 | 郑州轻工业大学 | LC filtering type grid-connected inverter weighting sliding mode model prediction capacitor voltage control method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107425743B (en) | 2019-09-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107425743B (en) | A kind of gird-connected inverter MPC method based on prediction deviation feedback correction | |
| CN111130366B (en) | Prediction control method of non-grid voltage sensor of three-phase VIENNA rectifier | |
| CN105552959B (en) | Three-phase grid rectifier prediction direct Power Control method based on extended state observer | |
| Sharma et al. | Real-time implementation of shunt active power filter with reduced sensors | |
| CN103944428B (en) | A kind of control method of the Three-Phase PWM Rectifier being applicable to waveform distortion of the power supply network | |
| CN102916599A (en) | Model predication control method of three-phase PWM (pulse width modulation) rectifier under unbalanced voltage | |
| CN107611971B (en) | Network side inverter resonance full-order sliding mode control method for network voltage harmonic distortion | |
| CN104836232A (en) | Frequency wide-range self-adaptive repetition control method of active power filter | |
| CN103595069A (en) | Method for carrying out model prediction control on grid-side converter of photovoltaic power generation system under unbalanced voltage | |
| JP6406268B2 (en) | Periodic disturbance suppression control device | |
| CN114079399B (en) | Grid-connected inverter current loop control system and method based on linear active disturbance rejection control | |
| Liao et al. | Robust dichotomy solution-based model predictive control for the grid-connected inverters with disturbance observer | |
| CN103595280A (en) | Deadbeat control method of solar power generating system inverter under condition of voltage unbalance | |
| CN111817713A (en) | High-voltage direct-current phase-locked loop capable of rapidly synchronizing voltage phases under symmetric faults and algorithm | |
| Mohanraj et al. | A unified power quality conditioner for power quality improvement in distributed generation network using adaptive distributed power balanced control (ADPBC) | |
| Gong et al. | A low-cost phase-angle compensation method for the indirect matrix converters operating at the unity grid power factor | |
| CN107123991B (en) | Dynamic voltage restorer control method and device based on fuzzy particle swarm algorithm | |
| WO2024082730A1 (en) | Finite control set model predictive control method for llcl-type battery energy storage converter | |
| CN103001509B (en) | Self-compensating model-predictive control method for parameters of active front-end rectifier | |
| CN102969913A (en) | Method for compensating mismatching of model predictive control parameters for initiative front-end rectifier | |
| CN102946110B (en) | Fixed frequency model prediction control method for voltage type PWM (Pulse Width Modulation) rectifier in process of voltage unbalance drop | |
| Shi et al. | Modified model predictive control of voltage source inverter | |
| CN115276439A (en) | LCL type grid-connected inverter resonance suppression method adapting to impedance change of weak grid | |
| CN113452069A (en) | Multi-step model prediction control method for inverter station supplying power to passive network | |
| CN110048466B (en) | Prediction control method for grid-side converter model of double-fed wind power generation system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |