CN106849082B - The Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic - Google Patents
The Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic Download PDFInfo
- Publication number
- CN106849082B CN106849082B CN201710140169.0A CN201710140169A CN106849082B CN 106849082 B CN106849082 B CN 106849082B CN 201710140169 A CN201710140169 A CN 201710140169A CN 106849082 B CN106849082 B CN 106849082B
- Authority
- CN
- China
- Prior art keywords
- formula
- upqc
- output
- control
- distribution network
- 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.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/01—Arrangements for reducing harmonics or ripples
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E40/00—Technologies for an efficient electrical power generation, transmission or distribution
- Y02E40/40—Arrangements for reducing harmonics
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The present invention provides a kind of Research on Unified Power Quality Conditioner Harmonic Control Method of power distribution network containing photovoltaic, implemented using UPQC, the UPQC includes signal acquisition module, signal conditioning module, microcontroller and drive circuit module, the following steps are included: the first step, voltage signal acquisition: acquisition power distribution network three-phase bus voltage ua,ub,ucWith load current ia,ib,ic；Second step acquires signal condition；Third step calculates harmonic compensation instruction；4th step, using the PI control parameter (k of adaptive line spectrum enhancer adjusting UPQCp, ki)；5th step, with selected optimal PI control parameter (kp, ki) implement PI control, output current harmonics compensate component.The present invention can control effectively to the harmonic wave of the power distribution network containing photovoltaic by using the PI control method based on adaptive line spectrum enhancer adjusting PI control parameter；A kind of new method is provided to administer distribution network electric energy quality using UPQC.
Description
Technical field
The present invention relates to Distribution Network Harmonics control methods, and in particular to a kind of to utilize Research on Unified Power Quality Conditioner to containing light
Lie prostrate the Harmonic Control Method of power distribution network.
Background technique
The research of power quality control technology has huge economic and social benefit, is one in electric power research field
Hot spot.Voltage swell, voltage dip, three-phase imbalance voltage, harmonic voltage, reactive current, harmonic current, out-of-balance current etc.
The harm of power quality problem is on the rise.But current existing device mostly access system in parallel or serial fashion, and
It can only solve part power quality problem.With increasingly sophisticated, the various power qualities of distribution net work structure and electric load ingredient
The case where problem occurs simultaneously in same distribution system or in same power load also can be more and more.If matched same
Existing voltage-sensitive load has nonlinear-load and impact load again on goddess of lightning's line, it is necessary at the same install voltage compensating device and
Current compensator.If all individually taking a type of controlling device for each power quality problem, it will increase greatly
Add treatment cost, also will increase the complexity of device operation and maintenance, and there is also cooperation problems between each device.
Unified Power Quality Controller (Unified Power Quality Conditioner, abbreviation UPQC) is used as function
The powerful electric energy quality synthesis compensation apparatus of energy, independent operating realizes respective function after series, parallel unit can decouple, can also
Combined operating realizes unified comprehensive function, becomes the research hotspot in power quality controlling field in recent years.Such as Publication No.
The Chinese patent literature of CN103326397A is related to a kind of Unified Power Quality Controller of mixed frequency control；Publication No.
The Chinese patent literature of CN204633344U is related to a kind of Research on Unified Power Quality Conditioner with uninterruptible power supply function.So
And traditional three-phase power distribution system is all based on greatly to the research of Unified Power Quality Controller itself and its application at present, and it is right
It is less in the power distribution network research of the new-energy grid-connecteds such as photovoltaic.Since the new energy such as photovoltaic have the characteristics that randomness, intermittence, and
Carry out a large amount of harmonic pollutions to conventional electrical distribution mesh belt after net, harmonic pollution is carried out using traditional control method using UPQC at this time
Control effect is bad, thus, research and utilization UPQC implements effectively control to the harmonic wave in the power distribution network containing photovoltaic using new method
System, it appears very necessary.
Linear neural network is a kind of simple neuroid, it can be made of one or more linear neurons.
Adaline network (the Adaptive proposed by Stanford Univ USA professor Berhard Widrow in 1962
Linear Element, Adaline) it is the earliest Typical Representative of linear neural network, it is one by input layer and output layer
The single layer feed-forward type network of composition, it and perceptron are the difference is that the transfer function of each of which neuron is linear letter
Number, therefore the output of adaptive line spectrum enhancer can take arbitrary value, and the output of perceptron can only be 1 or 0.Linear neural
Network uses a kind of new learning rules proposed jointly by Berhard Widrow and Marcian Hoff.Adaptive line mind
Learning algorithm through network is all greatly improved than the convergence rate and precision of the learning algorithm of perceptron.Adaptive line mind
It is mainly used for function approximation, signal detection, System Discrimination, the fields such as pattern-recognition and control through network.
Adaptive line spectrum enhancer structure perceptron, the difference is that the transfer function of each of which neuron is linear
Function.For with M input, the L linear neural network exported.The input summation net of i-th of neuron of output layeri
For
In formula, wijFor training weighting coefficient, xjFor input vector, M is the number of nodes of input layer, θiFor output layer neuron
The threshold value of i；The number inputted；
The output of i-th of neuron of output layer is respectively as follows: yi=f (neti)；yi=f (neti) it is activation primitive, it is
The transfer function of linear function,
The study of adaptive line spectrum enhancer uses Wideow-HOFF learning rules, during the training period, constantly with training
The each mode concentrated is to training network.When giving a certain training mode, output unit can generate a reality output vector,
With desired output and the difference of actual output come corrective networks connection weight.How by adaptive line spectrum enhancer and UPQC phase
In conjunction with, it is effective to the harmonic wave implementation in the power distribution network containing photovoltaic to control, it is the interested problem of those skilled in the art.
Summary of the invention
The object of the present invention is to provide a kind of Research on Unified Power Quality Conditioner Harmonic Control Method of power distribution network containing photovoltaic,
This method is implemented using Research on Unified Power Quality Conditioner, adjusts PI control parameter by using based on adaptive line spectrum enhancer
PI control method, control effectively to the harmonic wave of the power distribution network containing photovoltaic.
The technical scheme is that the Research on Unified Power Quality Conditioner harmonic controling side of the power distribution network of the invention containing photovoltaic
Method is implemented using UPQC, and the UPQC includes signal acquisition module, signal conditioning module, microcontroller and driving circuit mould
Block, comprising the following steps:
The first step, voltage signal acquisition:
The signal acquisition module of UPQC acquires power distribution network three-phase bus voltage ua,ub,ucWith load current ia,ib,icConcurrently
Give the signal conditioning module of UPQC；
Second step acquires signal condition:
The power distribution network three-phase bus voltage u that the signal conditioning module of UPQC sends signal acquisition modulea,ub,ucAnd load
Electric current ia,ib,icAfter improving into the acceptable signal of microprocessor, it is sent to microprocessor；
Third step calculates harmonic compensation instruction:
1. microcontroller is by threephase load electric current ia、ib、icI is obtained through Park Transformationα、iβ, transformation for mula such as formula (1):
2. microcontroller is by iα、iβI is calculated according to instantaneous power theoryp、iq, transformation for mula such as formula (2):
In formula, ω is network voltage frequency；
3. obtaining i through the LPF low-pass filtering that low-pass cut-off frequencies are 50HZp、iqFundamental positive sequence ipf、iqf；
4. microcontroller is according to fundamental positive sequence ipf、iqf, by CωtInverse transformation and Park Transformation C32Inverse transformation fortune
Calculation obtains threephase load current first harmonics component iaf、ibf、icf, then with threephase load electric current ia、ib、icSubtract each other and obtains harmonic compensation
Instruct iah、ibh、ich, harmonic compensation instruction iah、ibh、ichIt is abbreviated as；
4th step adjusts the PI control parameter (k of UPQCp, ki):
If the PI controller of UPQC is increment type PI controller, control error is
The variable quantity △ u (k) and e (k) of the output u (k) of PI controller has shown relationship as the formula:
△ u (k)=kp(e(k)-e(k-1))+kie(k) (5)
In formula, scale parameter kp, integral parameter kiFor the parameter of adjusting needed for PI controller, microprocessor is using adaptive
Linear neural network adjusts PI control parameter (k by step in detail belowp, ki):
1. initialization: linear neural network input is kp、kiTwo parameters export as the k after optimizationp、kiTwo parameters,
Number of training N=100, weighting coefficient wijValue range be [0.1,20], initialize all weighting coefficients be it is the smallest with
Machine number；
2. providing training set: for parameter kp, i.e. x(1), value range is [1,100]；For parameter ki, i.e. x(2), value
Range is [0.001,1], to two parameters in value range, by assignment in order at random, provides 100 training samples,
Provide 100 input vector x(1), x(2), 100 desired output vector t(1), t(2)；
3. calculating the output of each neuron of output layer:
In the study stage of training network, input under selecting a sample p in 100 training samples to act on/defeated
Mode is to { x outpAnd { tp, carry out network training, input of i-th of the neuron of output layer under the action of sample p are as follows:
In formula, wijFor training weighting coefficient, value is the random number of initialization；θiFor the threshold value of output layer neuron i, just
Beginning value is 0.5,For the input under sample p effect；
The output of i-th of neuron of output layer are as follows:
In formula, f () is the linear activation primitive that the input of network is directly switched to output, expression formula are as follows:
4. calculating the desired value of all training samples and the error of actual value:
For the quadratic form error function of the input pattern pair of each sample p are as follows:
In formula,Indicate the desired output of i-th of neuron under sample p effect；It indicates under sample p effect
The reality output of i-th of neuron, eiIndicate the error between sample p desired output and reality output；
The performance index function for taking PI to control is J, expression formula are as follows:
kp, kiAdjustment algorithm use gradient descent method:
In formula, △ kpFor kpChange of gradient amount, △ kiFor kiChange of gradient amount；η is learning rate,α value is
1；
5. adjusting the weighting coefficient w of output layerijAnd threshold θi:
According to gradient method, the weighting coefficient correction formula of any neuron i of output layer can be obtained are as follows:
According to formula (10) and formula (11),
Therefore △ kp=e (k) △ wij·[e(k)-e(k-1)] (13)
△ki=e (k) △ wij·e(k) (14)
The weighting coefficient correction formula of any neuron i of output layer are as follows:
Threshold θiCorrection formula are as follows:
△θi=η (ti-yi)=η ei (16)
η is with input sample xpIt is adaptively adjusted；
Step is calculated 3. 6. returning, kp、kiTuning process is carried out by the direction that the performance index function J of PI control reduces, when
When J obtains minimum value, corresponding PI control parameter (kp、ki) it is the optimal value adjusted；
5th step, output current harmonics compensate component:
Microcontroller is with selected optimal PI control parameter (kp, ki) implement PI control, it exports u (k), then generates phase
The pwm signal answered is sent to drive circuit module；Drive circuit module generates and exports corresponding current harmonics compensation component ic,
Be connected to the grid electric current is, realize the harmonic controling to load current.
The present invention has the effect of positive: the Research on Unified Power Quality Conditioner harmonic controling of the power distribution network of the invention containing photovoltaic
Method is implemented using Research on Unified Power Quality Conditioner, adjusts PI control parameter by using based on adaptive line spectrum enhancer,
It is control effectively with the optimal PI control parameter of adjusting to the harmonic wave of the power distribution network containing photovoltaic.The present invention is neural by adaptive line
The PI of network and Research on Unified Power Quality Conditioner control organically combines, and UPQC carries out the Distribution Network Harmonics containing photovoltaic
Effectively control provides a kind of new method to administer distribution network electric energy quality using UPQC.
Detailed description of the invention
Fig. 1 is the structural schematic diagram of UPQC employed in the embodiment of the present invention, also schematically shows UPQC in figure
With the electrical connection of the power distribution network containing photovoltaic；
Fig. 2 is the hardware configuration schematic block for participating in implementing harmonic compensation control in UPQC employed in the embodiment of the present invention
Scheme, also schematically shows its electrical connection with power distribution network in figure；
Fig. 3 is the schematic illustration that UPQC employed in the embodiment of the present invention calculates harmonic compensation instruction；
Fig. 4 is that UPQC employed in the embodiment of the present invention uses the PI control adjusted based on adaptive line spectrum enhancer
Method, the schematic illustration that the harmonic wave of the power distribution network containing photovoltaic is controlled.
Appended drawing reference in above-mentioned attached drawing is as follows:
Series filtering unit 1, parallel filtering unit 2, power supply unit 3.
Specific embodiment
The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.
(embodiment 1)
The Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic of the present embodiment, passes through existing unification
Electric energy regulator (hereinafter referred to as UPQC) is implemented, and uses adaptive line spectrum enhancer using Research on Unified Power Quality Conditioner
The PI control method for adjusting PI control parameter, control effectively to the harmonic wave of the power distribution network containing photovoltaic.
See Fig. 1, the distribution network system containing photovoltaic is made of alternating current distribution, grid-connected photovoltaic and nonlinear load, distribution netting gear
There is distribution bus；The UPQC of aforementioned use mainly includes series filtering unit 1, parallel filtering unit 2 and power supply unit 3, UPQC
The mode for accessing grid-connected power distribution network is as shown in Figure 1.Power supply unit 3 is mainly by photovoltaic array and battery connected to it
Device composition, for providing UPQC itself working power.
Referring to fig. 2, it includes for acquiring network voltage and bearing which, which participates in the specific functional modules of harmonic compensation control,
Carry current signal signal acquisition module, for improving the signal conditioning module of collection voltages current signal, for receive letter
Signal that number conditioning module is sent and the microcontroller for being handled and being controlled, for executing the sent out control instruction phase of microcontroller
The drive circuit module for implementing harmonic compensation control to power distribution network bus should be generated.In the present embodiment, microcontroller is preferably used
DSP。
For UPQC when implementing harmonic compensation control to power distribution network, microcontroller, which uses, is based on adaptive line spectrum enhancer
The PI control method of adjusting is controlled, and realizes that the compensation to Distribution Network Harmonics controls by drive circuit module.
Referring to Fig. 3 and Fig. 4, the Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic of the present embodiment,
Specific step is as follows:
The first step, voltage signal acquisition:
The signal acquisition module of UPQC acquires power distribution network three-phase bus voltage ua,ub,ucIt (is only symbolically labelled in Fig. 3
ua) and load current ia,ib,ic(label is in Fig. 4l) and be sent to the signal conditioning module of UPQC.
Second step acquires signal condition:
The power distribution network three-phase bus voltage u that signal conditioning module sends signal acquisition modulea,ub,ucAnd load current
ia,ib,icAfter improving into the acceptable signal of microprocessor, it is sent to microprocessor.
Third step calculates harmonic compensation instruction:
1. microcontroller is by threephase load electric current ia、ib、icI is obtained through Park Transformationα、iβ, transformation for mula C32It is as follows:
2. microcontroller is by iα、iβI is calculated according to instantaneous power theoryp、iq, transformation for mula CωtIt is as follows:
Wherein, ω is network voltage frequency, as shown in Figures 2 and 3, by acquiring distribution network voltage, and utilizes phaselocked loop
It can get stable network voltage frequency；
3. obtaining i through the LPF low-pass filtering that low-pass cut-off frequencies are 50HZp、iqFundamental positive sequence ipf、iqf；
4. microcontroller is according to fundamental positive sequence ipf、iqf, by CωtInverse transformation is (i.e. in Fig. 3) and Park Transformation C32's
Inverse transformation (i.e. C in Fig. 323) operation obtains threephase load current first harmonics component iaf、ibf、icf, then with threephase load electric current
ia、ib、icSubtract each other and show that harmonic compensation instructs iah、ibh、ich(in Fig. 4 label for)。
4th step adjusts the PI control parameter (k of UPQCp, ki):
Harmonic detecting is instructed i* by microcontrollercAs input signal, adjust based on adaptive line spectrum enhancer
PI control, if the PI controller of UPQC be increment type PI controller, control error be
K is sampling step number, and the output of PI controller is u (k):
U (k)=u (k-1)+△ u (k) (4)
△ u (k) is the variable quantity that PI controller exports u (k), the output of PI controller when u (k-1) is kth -1 time sampling,
△ u (k) and e (k) has shown relationship as the formula:
△ u (k)=kp(e(k)-e(k-1))+kie(k) (5)
In formula, scale parameter kp, integral parameter kiFor the parameter of adjusting needed for PI controller, microprocessor is using adaptive
Linear neural network adjusts PI control parameter (kp, ki) specific step is as follows:
1. initialization: linear neural network input is kp、kiTwo parameters export as the k after optimizationp、kiTwo parameters,
Number of training N=100, weighting coefficient wijValue range be [0.1,20], initialize all weighting coefficients be it is the smallest with
Machine number；
2. providing training set: for parameter kp, i.e. x(1), value range is [1,100]；For parameter ki, i.e. x(2), value
Range is [0.001,1], to two parameters in value range, by assignment in order at random, provides 100 training samples,
Provide 100 input vector x(1), x(2), 100 desired output vector t(1), t(2)；
3. calculating the output of each neuron of output layer:
In the study stage of training network, for 100 training samples, first select under wherein some sample p effect
Input/output mode is to { xpAnd { tp, network training is carried out, i-th of neuron of output layer is defeated under the action of sample p
Enter are as follows:
In formula, wijFor training weighting coefficient, value is the random number of initialization；θiFor the threshold value of output layer neuron i, just
Beginning value is 0.5,For the input under sample p effect；
The output of i-th of neuron of output layer are as follows:
In formula, f () is linear activation primitive, and the input of network is directly switched to export by it, therefore expression formula are as follows:
4. calculating the desired value of all training samples and the error of actual value:
For the quadratic form error function of the input pattern pair of each sample p are as follows:
In formula,Indicate the desired output of i-th of neuron under sample P effect；It indicates under sample p effect
The reality output of i-th of neuron, eiIndicate the error between sample p desired output and reality output；
The performance index function for taking PI to control is J, expression formula are as follows:
kp, kiAdjustment algorithm use gradient descent method:
In formula, △ kpFor kpChange of gradient amount, △ kiFor kiChange of gradient amount；η is learning rate,α is normal
Value can make algorithmic statement as 0 < α < 2, and α value is 1；
5. adjusting the weighting coefficient w of output layerijAnd threshold θi:
According to gradient method, the weighting coefficient correction formula of any neuron i of output layer can be obtained are as follows:
According to formula (10) and formula (11),
Therefore △ kp=e (k) △ wij·[e(k)-e(k-1)] (13)
△ki=e (k) △ wij·e(k) (14)
The weighting coefficient correction formula of any neuron i of output layer are as follows:
Similarly, threshold θiCorrection formula are as follows:
△θi=η (ti-yi)=η ei (16)
η is with input sample xpIt is adaptively adjusted；
Step is calculated 3. 6. returning, kp、kiTuning process is carried out by the direction that J reduces, corresponding when J obtains minimum value
PI control parameter (kp、ki) it is optimal value.
5th step, output current harmonics compensate component:
Microcontroller is with selected optimal PI control parameter (kp, ki) implement PI control, it exports u (k), then generates phase
The pwm signal answered is sent to drive circuit module, and drive circuit module generates and exports corresponding current harmonics compensation component ic,
Be connected to the grid electric current is, to realize to load current il(i.e. three-phase current ia,ib,ic) harmonic controling.
Above embodiments are the explanations to a specific embodiment of the invention, rather than limitation of the present invention, related technology
The technical staff in field without departing from the spirit and scope of the present invention, can also make various transformation and variation and obtain
To corresponding equivalent technical solution, therefore all equivalent technical solutions should be included into patent protection model of the invention
It encloses.
Claims (1)
1. a kind of Research on Unified Power Quality Conditioner Harmonic Control Method of power distribution network containing photovoltaic, is implemented using UPQC, described
UPQC includes signal acquisition module, signal conditioning module, microcontroller and drive circuit module, which is characterized in that including following
Step:
The first step, voltage signal acquisition:
The signal acquisition module of UPQC acquires power distribution network three-phase bus voltage ua,ub,ucWith load current ia,ib,icAnd it is sent to
The signal conditioning module of UPQC；
Second step acquires signal condition:
The power distribution network three-phase bus voltage u that the signal conditioning module of UPQC sends signal acquisition modulea,ub,ucAnd load current
ia,ib,icAfter improving into the acceptable signal of microprocessor, it is sent to microprocessor；
Third step calculates harmonic compensation instruction:
1. microcontroller is by threephase load electric current ia、ib、icI is obtained through Park Transformationα、iβ, transformation for mula such as formula (1):
2. microcontroller is by iα、iβI is calculated according to instantaneous power theoryp、iq, transformation for mula such as formula (2):
In formula, ω is network voltage frequency；
3. obtaining i through the LPF low-pass filtering that low-pass cut-off frequencies are 50HZp、iqFundamental positive sequence ipf、iqf；
4. microcontroller is according to fundamental positive sequence ipf、iqf, by CωtInverse transformation and Park Transformation C32Transform operation obtain
Threephase load current first harmonics component iaf、ibf、icf, then with threephase load electric current ia、ib、icSubtract each other and show that harmonic compensation instructs
iah、ibh、ich, harmonic compensation instruction iah、ibh、ichIt is abbreviated as
4th step adjusts the PI control parameter (k of UPQCp, ki):
If the PI controller of UPQC is increment type PI controller, control error is
The variable quantity △ u (k) and e (k) of the output u (k) of PI controller has shown relationship as the formula:
△ u (k)=kp(e(k)-e(k-1))+kie(k) (5)
In formula, scale parameter kp, integral parameter kiFor the parameter of adjusting needed for PI controller, microprocessor uses adaptive line
Neural network adjusts PI control parameter (k by step in detail belowp, ki):
1. initialization: linear neural network input is kp、kiTwo parameters export as the k after optimizationp、kiTwo parameters, training sample
This number N=100, weighting coefficient wijValue range is [0.1,20], and initializing all weighting coefficients is the smallest random number；
2. providing training set: for parameter kp, i.e. x(1), value range is [1,100]；For parameter ki, i.e. x(2), value range
For [0.001,1], to two parameters in value range, by assignment in order at random, 100 training samples is provided, that is, are given
100 input vector x out(1), x(2), 100 desired output vector t(1), t(2)；
3. calculating the output of each neuron of output layer:
Input/output mould in the study stage of training network, under selecting a sample p in 100 training samples to act on
Formula is to { xpAnd { tp, carry out network training, input of i-th of the neuron of output layer under the action of sample p are as follows:
In formula, wijFor training weighting coefficient, value is the random number of initialization；θiFor the threshold value of output layer neuron i, initially take
Value is 0.5,For the input under sample p effect；
The output of i-th of neuron of output layer are as follows:
In formula, f () is the linear activation primitive that the input of network is directly switched to output, expression formula are as follows:
4. calculating the desired value of all training samples and the error of actual value:
For the quadratic form error function of the input pattern pair of each sample p are as follows:
In formula,Indicate the desired output of i-th of neuron under sample p effect；Indicate i-th under sample p effect
The reality output of a neuron, eiIndicate the error between sample p desired output and reality output；
The performance index function for taking PI to control is J, expression formula are as follows:
kp, kiAdjustment algorithm use gradient descent method:
In formula, △ kpFor kpChange of gradient amount, △ kiFor kiChange of gradient amount；η is learning rate,α value is 1；
5. adjusting the weighting coefficient w of output layerijAnd threshold θi:
According to gradient method, the weighting coefficient correction formula of any neuron i of output layer can be obtained are as follows:
According to formula (10) and formula (11),
Therefore △ kp=e (k) △ wij·[e(k)-e(k-1)] (13)
△ki=e (k) △ wij·e(k) (14)
The weighting coefficient correction formula of any neuron i of output layer are as follows:
Threshold θiCorrection formula are as follows:
△θi=η (ti-yi)=η ei (16)
η is with input sample xpIt is adaptively adjusted；
Step is calculated 3. 6. returning, kp、kiTuning process is carried out by the direction that the performance index function J of PI control reduces, when J takes
When obtaining minimum value, corresponding PI control parameter (kp、ki) it is the optimal value adjusted；
5th step, output current harmonics compensate component:
Microcontroller is with selected optimal PI control parameter (kp, ki) implement PI control, it exports u (k), then generates corresponding
Pwm signal is sent to drive circuit module；Drive circuit module generates and exports corresponding current harmonics compensation component ic, it is incorporated to
Power network current is, realize the harmonic controling to load current.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710140169.0A CN106849082B (en) | 2017-03-10 | 2017-03-10 | The Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710140169.0A CN106849082B (en) | 2017-03-10 | 2017-03-10 | The Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106849082A CN106849082A (en) | 2017-06-13 |
| CN106849082B true CN106849082B (en) | 2019-06-11 |
Family
ID=59143728
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710140169.0A Active CN106849082B (en) | 2017-03-10 | 2017-03-10 | The Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106849082B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107645166B (en) * | 2017-10-30 | 2021-01-08 | 燕山大学 | Novel photovoltaic integrated unified power quality adjusting device and control method thereof |
| CN107908113A (en) * | 2017-12-29 | 2018-04-13 | 安徽理工大学 | Based on the automatic disturbance rejection controller parameter tuning method for improving TLBO algorithms |
| CN108958098A (en) * | 2018-05-11 | 2018-12-07 | 云南电网有限责任公司昆明供电局 | Take into account the controller of the UPQC device of telecommunication |
| CN110474337A (en) * | 2019-07-10 | 2019-11-19 | 国网辽宁省电力有限公司电力科学研究院 | It is a kind of that control strategy is promoted based on the micro-capacitance sensor power quality for improving UPQC detection method |
| CN111478327B (en) * | 2020-05-18 | 2021-05-18 | 山东山大世纪科技有限公司 | New energy power quality control system and method based on dynamic strategy analysis |
| CN112398131B (en) * | 2020-09-30 | 2023-08-04 | 沈阳工业大学 | UPQC control system based on time lag compensation and control method thereof |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102832621A (en) * | 2012-09-18 | 2012-12-19 | 河海大学常州校区 | Adaptive RBF (radial basis function) neural network control technique for three-phase parallel active filters |
| CN104037766A (en) * | 2014-05-28 | 2014-09-10 | 河海大学常州校区 | Method for self-adaptive neural inversion control of three-phase parallel connection type active filter |
| CN104833154A (en) * | 2015-05-28 | 2015-08-12 | 河海大学常州校区 | Chilled water loop control method based on fuzzy PID and neural internal model |
-
2017
- 2017-03-10 CN CN201710140169.0A patent/CN106849082B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102832621A (en) * | 2012-09-18 | 2012-12-19 | 河海大学常州校区 | Adaptive RBF (radial basis function) neural network control technique for three-phase parallel active filters |
| CN104037766A (en) * | 2014-05-28 | 2014-09-10 | 河海大学常州校区 | Method for self-adaptive neural inversion control of three-phase parallel connection type active filter |
| CN104833154A (en) * | 2015-05-28 | 2015-08-12 | 河海大学常州校区 | Chilled water loop control method based on fuzzy PID and neural internal model |
Non-Patent Citations (2)
| Title |
|---|
| 一种基于自适应线性神经网络的瞬时谐波检测方法;赵江稳;《山西冶金》;20101231(第2期);全文 |
| 基于Adaline神经网络参数辨识的含DG配电网的电压稳定性分析;杨志淳;《电力与能源》;20150430;第36卷(第2期);全文 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106849082A (en) | 2017-06-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106849082B (en) | The Research on Unified Power Quality Conditioner Harmonic Control Method of the power distribution network containing photovoltaic | |
| Agarwal et al. | LMF-based control algorithm for single stage three-phase grid integrated solar PV system | |
| Singh et al. | Double-stage three-phase grid-integrated solar PV system with fast zero attracting normalized least mean fourth based adaptive control | |
| Kececioglu et al. | Power quality improvement using hybrid passive filter configuration for wind energy systems | |
| CN107317352B (en) | Control method of miniature photovoltaic inverter with three-phase unbalance treatment function | |
| Gupta et al. | Fuzzy logic controlled shunt active power filter for reactive power compensation and harmonic elimination | |
| Zue et al. | Simulation and stability analysis of a 100 kW grid connected LCL photovoltaic inverter for industry | |
| CN105226664B (en) | A kind of active distribution network reactive voltage layer distributed control method for coordinating | |
| Singh et al. | Load leveling and voltage control of permanent magnet synchronous generator-based DG set for standalone supply system | |
| CN108767869B (en) | Static reactive power compensator voltage adjusting method based on artificial neural network | |
| CN106786590B (en) | A kind of grid-connected Distribution Network Harmonics detection control method | |
| CN108808704A (en) | A kind of control method and device of virtual synchronous generator | |
| Benachaiba et al. | Smart control of UPCQ within microgrid energy system | |
| CN106816877A (en) | A kind of distribution network voltage containing photovoltaic falls detection compensation method | |
| Yeganeh et al. | Intelligent secondary control of islanded AC microgrids: A brain emotional learning-based approach | |
| Sindi et al. | Robust control of adaptive power quality compensator in Multi-Microgrids for power quality enhancement using puzzle optimization algorithm | |
| Suresh et al. | Shunt active power filter with solar photovoltaic system for long-term harmonic mitigation | |
| Tan et al. | Intelligent controlled dynamic voltage restorer for improving transient voltage quality | |
| Malla et al. | Supplementary control for virtual synchronous machine based on adaptive dynamic programming | |
| Meena et al. | Control of DSTATCOM in distribution systems by artificial intelligence based controller | |
| CN108258692A (en) | A kind of harmonic compensation method and device of parameter adaptive adjustment | |
| Khan et al. | Intelligent control of fixed capacitor-thyristor controlled reactor for power quality improvement | |
| Patra et al. | Design and Development of ANFIS based Controller for Three Phase Grid Connected System. | |
| Kumar et al. | A frequency observer based control for solar energy conversion system | |
| Coronado-Mendoza et al. | Variable frequency control of a photovoltaic Boost converter system with power quality indexes based on dynamic phasors |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP03 | Change of name, title or address |
Address after: No.27, junqian street, Changzhou City, Jiangsu Province, 213000 Patentee after: STATE GRID JIANGSU ELECTRIC POWER Co.,Ltd. CHANGZHOU POWER SUPPLY BRANCH Patentee after: STATE GRID JIANGSU ELECTRIC POWER Co. Patentee after: STATE GRID CORPORATION OF CHINA Address before: No. 27, Tianning District Bureau, front street, Changzhou, Jiangsu Province Patentee before: STATE GRID CHANG ZHOU CURRENT SUPPLY COMPANY OF JIANGSU ELECTRIC POWER Co. Patentee before: STATE GRID JIANGSU ELECTRIC POWER Co. Patentee before: State Grid Corporation of China |
|
| CP03 | Change of name, title or address | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20221124 Address after: No.27, junqian street, Changzhou City, Jiangsu Province, 213000 Patentee after: STATE GRID JIANGSU ELECTRIC POWER Co.,Ltd. CHANGZHOU POWER SUPPLY BRANCH Patentee after: JIANGSU University OF TECHNOLOGY Patentee after: STATE GRID JIANGSU ELECTRIC POWER Co.,Ltd. Patentee after: STATE GRID CORPORATION OF CHINA Address before: No.27, junqian street, Changzhou City, Jiangsu Province, 213000 Patentee before: STATE GRID JIANGSU ELECTRIC POWER Co.,Ltd. CHANGZHOU POWER SUPPLY BRANCH Patentee before: STATE GRID JIANGSU ELECTRIC POWER Co. Patentee before: STATE GRID CORPORATION OF CHINA |
|
| TR01 | Transfer of patent right |