CN110190599A - A kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory - Google Patents
A kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory Download PDFInfo
- Publication number
- CN110190599A CN110190599A CN201910495672.7A CN201910495672A CN110190599A CN 110190599 A CN110190599 A CN 110190599A CN 201910495672 A CN201910495672 A CN 201910495672A CN 110190599 A CN110190599 A CN 110190599A
- Authority
- CN
- China
- Prior art keywords
- output
- voltage
- capacitance sensor
- frequency
- finite time
- 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
- 238000011217 control strategy Methods 0.000 title claims abstract description 23
- 238000004891 communication Methods 0.000 claims abstract description 30
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 19
- 238000007665 sagging Methods 0.000 claims abstract description 14
- 238000013461 design Methods 0.000 claims abstract description 9
- 238000000034 method Methods 0.000 claims abstract description 8
- 230000006641 stabilisation Effects 0.000 claims abstract description 6
- 238000011105 stabilization Methods 0.000 claims abstract description 6
- 230000001568 sexual effect Effects 0.000 claims abstract description 5
- 238000012545 processing Methods 0.000 claims abstract description 4
- 238000012549 training Methods 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 11
- 230000005284 excitation Effects 0.000 claims description 6
- 230000008878 coupling Effects 0.000 claims description 5
- 238000010168 coupling process Methods 0.000 claims description 5
- 238000005859 coupling reaction Methods 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 230000003068 static effect Effects 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 4
- 238000013528 artificial neural network Methods 0.000 claims description 3
- 230000005611 electricity Effects 0.000 claims description 3
- 238000013178 mathematical model Methods 0.000 claims description 3
- 230000001105 regulatory effect Effects 0.000 claims description 3
- 239000004744 fabric Substances 0.000 claims 1
- 230000000694 effects Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000002790 cross-validation Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000033228 biological regulation Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
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
-
- 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
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/381—Dispersed generators
-
- H02J3/382—
-
- 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
- H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
- H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
-
- 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
- H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
- H02J3/388—Islanding, i.e. disconnection of local power supply from the network
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Supply And Distribution Of Alternating Current (AREA)
- Feedback Control In General (AREA)
Abstract
The design method of the invention discloses a kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory, belongs to smart grid control field.The present invention is communicated in the isolated island micro-capacitance sensor based on distributed Two-stage control strategy, using reciprocity sparse network with contiguous agent, and by there is consistency algorithm processing to obtain information and local information；And consistent sexual deviation is fed back to the frequency and voltage controller designed by finite time consistency policy；Then, the output of controller is input into sagging control, is then transferred into voltage and current control loop, finally realizes frequency and voltage stabilization and traces into nominal value；Wherein, when breaking in the communications, the historical data of contiguous agent is predicted based on extreme learning machine is improved, and prediction result is input to the frequency designed by finite time consistency policy and voltage controller.To realize the secondary frequency and voltage adjusting under communication failure.
Description
Technical field
The invention belongs to smart grid control fields, and in particular to a kind of isolated island based on finite time congruity theory is micro-
The design method of power grid control strategy.
Background technique
With getting worse for energy crisis and problem of environmental pollution, the new energy including distributed generation technology
Using the extensive concern of society is received, micro-capacitance sensor is also rapidly developed.
The operation of AC micro-capacitance sensor can be divided into both of which: grid-connected and island mode.When island mode operation, using more
A distributed generation resource is run parallel, to improve output power quality.Generally use hierarchical control to power quality when island mode
It optimizes, and defines and execute different control targets using different layered structures.Sagging control is typically used as holding
The primary control that line frequency and voltage are adjusted, and power distribution may be implemented.But traditional sagging control and regulation effect
Bad, when distribution line impedance is uneven, power distribution effect is poor.Some existing improve include virtual impedance to balance
Route resistance and some adaptive sagging control strategies.Although there are many improved strategies for sagging control, due to inherently missing
Difference, simple sagging control always can not accurate track reference values.Therefore, in order to obtain more precise control effect, it is proposed that adopt
With Two-stage control strategy.Currently, existing Two-stage control strategy is generally divided into centralized and distributed, it is usually used in compensating sagging control
System.Centralized Two-stage control is formulated by micro-capacitance sensor master controller (MGCC), and MGCC is collected and calculated the entire net in micro-capacitance sensor
Network information, then issues commands to physical layer.Although centralized control precision is high, the communication pressure of system is very big, reliability
And poor expandability.Distributed secondary control does not need central node and is communicated, and the reciprocity sparse network of most of uses,
It is only communicated with contiguous agent, therefore has better scalability and feasibility.Distributed secondary control strategy can be with
Realize that frequency and voltage are restored, accurate power distribution.
But when communication failure interrupts, the unstable of whole system still will cause.
Summary of the invention
The invention proposes a kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory, it is intended to be based on one
Cause property is theoretical, and introduces finite time theory, and the two combines, and proposes a kind of new secondary control algorithm, and by communication disruption
Situation is taken into account, and is predicted based on extreme learning machine (ELM) is improved the historical data of contiguous agent, and by prediction result
Instead of the communication link information interrupted, predictive compensation is realized, improve system stability.
In order to solve the above technical problem, the present invention provides technical solution are as follows: one kind based on finite time consistency manage
The isolated island micro-capacitance sensor control strategy of opinion, which is characterized in that the steps include:
In the isolated island micro-capacitance sensor based on distributed Two-stage control strategy, carried out using reciprocity sparse network and contiguous agent
Communication, and by there is consistency algorithm processing to obtain information and local information；And by consistent sexual deviation feedback to by finite time
The frequency and voltage controller of consistency policy design；Then, the output of controller is input into sagging control, is then transferred into
Voltage and current control loop finally realizes frequency and voltage stabilization and traces into nominal value；
Wherein, when breaking in the communications, the historical data of contiguous agent is predicted based on extreme learning machine is improved, and will
Prediction result is input to the frequency designed by finite time consistency policy and voltage controller.
A further technical solution lies in the isolated island micro-capacitance sensor is divided into two parts of physical layer and network layer；The object
Reason layer mainly includes controller and distributed generation resource, and controller stablizes each distributed generation resource output, then by the defeated of distributed generation resource
Point of common coupling is connected to by static transfer switch out；Or controller stablizes each distributed generation resource output, then uses distributed electrical
Source is powered load；The network layer carries out the data exchange between different electronic power inverters, completes entire micro- electricity
The output of distributed generation resource is synchronous in netting.
A further technical solution lies in acting on behalf of DGiOutput error can use consistency algorithm are as follows:
Wherein, eiAct on behalf of the global output error of i, xiAct on behalf of the real output value of i, xjAct on behalf of the phase of i
Neighbour acts on behalf of the value that j is come by communications, xrefThe nominal value for acting on behalf of the output of i, the Leader as consistency algorithm
Node, aijIt indicates weighted adjacent coefficient, that is, acts on behalf of the network communication links of i, biThe weight for leading to Leader is represented, and if only if
When acting on behalf of the accessible Leader of i, bi> 0；
In order to realize the regulating effect of finite time, created symbol function:
The finite-time control device of design are as follows:
ui=β sig (ei)α
Wherein, α, β are finite-time control parameters, and α is index coefficient, and 0 < α < 1 improves system convergence performance, β is
Proportionality coefficient, β > 0 determine the step-length of control variable.
A further technical solution lies in carry out prediction step to the historical data of contiguous agent based on extreme learning machine is improved
Suddenly specifically:
(1) foundation of prediction model
For any N number of different sample (Xi,Yi), wherein Xi=[xi1 xi2 … xin]T∈RnIndicate that each DG history is defeated
Frequency or voltage out, Yi=[yi1 yi2 … yim]T∈RmThe frequency or voltage for indicating each DG desired output have L for one
The neural networks with single hidden layer of a hidden node can indicate are as follows:
Wherein g (x) is excitation function, generally Gaussian function, Wi=[wi1 wi2 … win]TFor input weight, γiFor
Export weight, ciFor the biasing of i-th of Hidden unit, ojIndicate prediction output；
When considering empiric risk and structure risk, the mathematical model of ES-ELM can be indicated are as follows:
Wherein, | | ε | |2Indicate empiric risk, | | γ | |2Indicate structure risk, η ∈ R indicates that two kinds of risk ratio parameters are logical
The best compromise point for crossing the mode of cross validation to determine；
LS-SVM algorithm is introduced, is Lagrange's equation by ES-ELM model conversation are as follows:
Wherein, λ=[λ1 λ2 … λN] indicating Lagrange multiplier, H indicates hidden layer output, γ=[γ1 γ2 …
γL]TIndicate output weight
According to KTT optimal conditions, seeking the gradient of Lagrange's equation and enabling it is 0, is obtained
To obtain
Wherein X+The generalized inverse matrix of representing matrix X；
(2) training of prediction model
First clear the following contents before training: the N number of different sample (X of training seti,Yi), hidden layer number L, excitation function g
(x),
Step 1: when data variation is larger, better training result, pre-processes data in order to obtain, specifically
Formula are as follows:
Wherein, x (i) and x'(i) respectively indicate initial data and treated data, ExAnd σiIndicate the equal of initial data
Value and standard deviation；
Step2: input weight wikWith biasing ciIt is arbitrarily set in (0,1) range, and calculates hidden layer output H；
Step 3: according toCalculate output weight γ and Lagrange multiplier λ.
(3) prediction of result
After training by ES-ELM, by historical data to the DG for having lost the communication informationiLocal prediction is carried out, point
The predicted value of frequency and voltage is not obtained.
The present invention having the beneficial effect that by adopting the above technical scheme
(1) present invention proposes a kind of based on distributed finite time to realize that more accurate frequency and voltage are restored
The voltage and frequency control strategy of consistency algorithm.Finite time strategy and consistency algorithm are combined, to reduce by unevenness
The coupling of frequency caused by even line impedance and voltage, and reach reference value in the finite time of setting.
(2) prove that proposed finite time consistency algorithm can accurately realize frequency and voltage with Lyapunov
Secondary control, and accurate track reference value.In addition, passing through the analysis of Lyapunov function, convergence rate and parameter can be obtained
Relationship between selection.
(3) when communication data is lost and communication disruption, by considering empiric risk and structure risk, ELM is changed
Into predicting for the historical data of frequency of training and voltage, and to the communication data of loss, then send prediction result
To pilot controller, to inhibit the influence of communication failure.
Detailed description of the invention
Fig. 1 is that the present invention is based on the isolated island micro-capacitance sensor simplification figures of distributed Two-stage control strategy；
Fig. 2 is the inverter block diagram in i-th of DG of the present invention；
Fig. 3 is the control block diagram of whole system；
Specific embodiment
Present invention is further described in detail with specific embodiment with reference to the accompanying drawing:
The embodiment of the present invention elaborates a kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory, special
Sign is, the steps include:
In the isolated island micro-capacitance sensor based on distributed Two-stage control strategy, carried out using reciprocity sparse network and contiguous agent
Communication, and by there is consistency algorithm processing to obtain information and local information；And by consistent sexual deviation feedback to by finite time
The frequency and voltage controller of consistency policy design；Then, the output of controller is input into sagging control, is then transferred into
Voltage and current control loop finally realizes frequency and voltage stabilization and traces into nominal value；
Wherein, when breaking in the communications, the historical data of contiguous agent is predicted based on extreme learning machine is improved, and will
Prediction result is input to the frequency designed by finite time consistency policy and voltage controller.
Particular content of the present invention is as follows:
(1) micro-capacitance sensor hierarchical control framework is designed
Micro-capacitance sensor control is broadly divided into two parts: physical layer and network layer, as shown in Figure 1.Physical layer mainly includes control
Device and distributed generation resource processed, controller stablizes each DG output, then the output of DG is passed through static transfer switch (static
Transfer switch, STS) it is connected to point of common coupling (point ofcommon coupling, PCC), it thus can be with
The common load externally hung on PCC is powered, naturally it is also possible to is powered with some DG to specific load.Network layer
It mainly realizes the data exchange between different electronic power inverters, realizes that all DG outputs synchronize in entire micro-capacitance sensor.
Fig. 2 shows the inverter block diagram inside DG, it is assumed that N number of DG unit is shared in micro-capacitance sensor, each DG is according to certain
Sequence be respectively defined as DGi(i=1,2 ..., N), DGiPCC is connected to by feeder line.The controller of inverter specifically includes that
1) power calculator；2) power controller；3) voltage control loop adjusts the exchange of DG unit and surveys voltage voi；4) current control is returned
Road adjusts the exchange of DG unit and surveys electric current ili。DGiThe three-phase electricity of output is converted with dp-frame, obtains vodi,voqi,
iodi,ioqi, then v is obtained by calculationoi,ili, to calculate DGiThe active and reactive power component of output.Power controller
It is the sagging control of power-handling capability progress in conjunction with them:
Wherein, miAnd niIt is cells D GiActive and idle sagging control coefrficient, ωiIt is DGiOutput frequency, ωrefIt is
System nominal frequency (2 π × 50red/s), vrefIt is nominal voltage of a system,It is that inverter ac surveys voltage voiReference value simultaneously
As the reference value of voltage and current ring, PiAnd QiIt is cells D G respectivelyiThe active and reactive power component of output.
(2) finite time consistency control strategy is designed, realizes the adjusting to frequency and voltage.
2.1 algebraic graph theory
In order to preferably describe micro-capacitance sensor communication network structure, it is theoretical with algebra digraph here.Such as Fig. 3 institute
Showing, communication network includes N number of more agencies (DGs) in micro-capacitance sensor, and it marks from 1 to N, is then mapped to digraph G (V, ε, A),
Interior joint collection V={ v1,v2,...,vNRepresent all DGs, edge aggregationIt represents and is able to carry out information exchange
Communication link.A=[aij]N×NIt is weighted adjacent matrix coefficient, aii=0 and aij>=0, and if only if (vi,vj) ∈ ε when, aij
> 0.The adjacent node of i-th of agency is defined as Ni={ vj∈V:(vi,vj) ∈ ε, the subordinated-degree matrix of digraph G
(degreematrix) it is set as D=diag { d1,d2,...,dN, andLaplacian Matrix Ω=D-A is
One symmetrical positive semidefinite matrix.
2.2 finite time consistency algorithms
The drop of bring the whole network frequency and voltage that simple sagging control can not be avoided that, therefore, this paper presents
One kind being based on finite time consistency algorithm, with Graph Analysis before, compensates to distributed generation resource output, finally makes
The output frequency and voltage for obtaining each agency are consistent and can continue respectively to trace into the frequency of setting and the mark of voltage
Title value ωrefAnd vref, improve the dynamic property of system.
For this acts on behalf of DGiOutput error can be calculated with consistency policy are as follows:
Wherein, eiAct on behalf of the global output error of i, xiAct on behalf of the real output value of i, xjAct on behalf of the phase of i
Neighbour acts on behalf of the value that j is come by communications, xrefThe nominal value for acting on behalf of the output of i, the Leader as consistency algorithm
Node, aijIt indicates weighted adjacent coefficient, that is, acts on behalf of the network communication links of i, biThe weight for leading to Leader is represented, and if only if
When acting on behalf of the accessible Leader of i, bi> 0.
In order to realize the regulating effect of finite time, created symbol function.
It ensure that the output global error for acting on behalf of i can be eliminated, and finally realize micro-capacitance sensor under conditions of some is suitble to
The output of interior each agency is consistent and levels off to nominal value.
The finite-time control device of design are as follows:
Wherein, α, β are finite-time control parameters, and α is index coefficient, and 0 < α < 1 improves system convergence performance, β is
Proportionality coefficient, β > 0 determine the step-length of control variable.
In order to verify the feasibility and stability that are designed with limit time controller, following Lyapunov letter is constructed
Number:
Therefore,
Wherein, e=[e1 e2 … eN], B=diag { b1 b2 … bN, Γ=Ω+B, Γ are a positive semidefinite matrixs,
It will be apparent thatDefine a bounded matrix Φ=δ ∈ R | δTδ=1, for ξ ⊥ 1, δ=β | ξα}.For even
Continuous δ ∈ Φ, function δTΓ δ is also continuous, andIn the presence of and be greater than 0, be defined as ρ.
Therefore,
It might as well setThen
It again assumes thatFinally, available:
According to proof aboveIt will be in finite timeLevel off to 0, therefore in time t*Interior generation
The output error e of reason can also level off to 0, it can realize
Namely
Detailed distribution Two-stage control strategy is as shown in Figure 3.As shown, obtaining adjacent DG by reciprocity sparse network
Information, and information and local information obtained are handled by consistency algorithm.Consistent sexual deviation feedback to by it is limited when
Between consistency policy design frequency and voltage controller.Then, the output of controller is input into sagging control, then shifts
To voltage and current control loop, finally realizes frequency and voltage stabilization and trace into nominal value.
(3) the predictive compensation strategy of communication disruption is designed
Traditional ELM is to be based on empirical risk minimization principle, and training error is minimized, but in the training process,
It will appear overfitting problem, reduce the generalization ability of model.Therefore, we introduce structural risk minimization theory, rationally weigh
Empiric risk and structure risk propose a kind of improved ELM model (ES-ELM).When communication data occurs to lose and interferes
When, it using improved extreme learning machine, is trained by the historical data to voltage and frequency, to predict script, this has
Data, finally the obtained data of prediction are input in consistency control, compared with conventional limit learning machine, ES-ELM has
Better generalization ability and robustness.
For any N number of different sample (Xi,Yi), wherein Xi=[xi1 xi2 … xin]T∈RnIndicate that each DG history is defeated
Frequency or voltage out, Yi=[yi1 yi2 … yim]T∈RmThe frequency or voltage for indicating each DG desired output have L for one
The neural networks with single hidden layer of a hidden node can indicate are as follows:
Wherein g (x) is excitation function, generally Gaussian function, Wi=[wi1 wi2 … win]TFor input weight, γiFor
Export weight, ciFor the biasing of i-th of Hidden unit, ojIndicate prediction output.
When considering empiric risk and structure risk, the mathematical model of ES-ELM can be indicated are as follows:
Wherein, | | ε | |2Indicate empiric risk, | | γ | |2Indicate structure risk, η ∈ R indicates that two kinds of risk ratio parameters are logical
The best compromise point for crossing the mode of cross validation to determine.
LS-SVM algorithm is introduced, is Lagrange's equation by ES-ELM model conversation are as follows:
Wherein, λ=[λ1 λ2 … λN] indicating Lagrange multiplier, H indicates hidden layer output, γ=[γ1 γ2 …
γL]TIndicate output weight
According to KTT optimal conditions, seeking the gradient of Lagrange's equation and enabling it is 0, is obtained
To obtain
The wherein generalized inverse matrix of X+ representing matrix X.
In conclusion ES-ELM considers empiric risk and structure risk simultaneously, so that the over-fitting of model reduces, extensive energy
Power also can be outstanding.
First clear the following contents before training: the N number of different sample (X of training seti,Yi), hidden layer number L, excitation function g
(x), mainly include several steps based on ES-ELM prediction process:
Step 1: when data variation is larger, better training result, pre-processes data in order to obtain, specifically
Formula are as follows:
Wherein, x (i) and x'(i) respectively indicate initial data and treated data, ExAnd σiIndicate the equal of initial data
Value and standard deviation.
Step 2: input weight wikWith biasing ciIt is arbitrarily set in (0,1) range, and calculates hidden layer output H.
Step 3: output weight γ and Lagrange multiplier λ is calculated according to formula (14).
After training by ES-ELM, by historical data to having lost communication information DGiCarry out local prediction, point
The predicted value of frequency and voltage is not obtained, and shorter predicted time can reduce system wild effect caused by loss of data.
, with predictive compensation mechanism, the stabilization of whole system is improved in the case that this breaks in the communications based on the above analysis
Property.
As it will be easily appreciated by one skilled in the art that the foregoing is merely preferred embodiments of the present invention, not to
The limitation present invention, all any modifications, equivalent replacements, and improvements etc. done within the spirit and principles in the present invention should all wrap
Containing within protection scope of the present invention.
Claims (4)
1. a kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory, which is characterized in that the steps include:
In the isolated island micro-capacitance sensor based on distributed Two-stage control strategy, led to using reciprocity sparse network with contiguous agent
Letter, and by there is consistency algorithm processing to obtain information and local information；And by consistent sexual deviation feedback to by finite time one
The frequency and voltage controller of cause property strategy design；Then, the output of controller is input into sagging control, is then transferred into electricity
Pressure and current controlled circuit finally realize frequency and voltage stabilization and trace into nominal value；
Wherein, when breaking in the communications, the historical data of contiguous agent is predicted based on extreme learning machine is improved, and will prediction
As a result the frequency designed by finite time consistency policy and voltage controller are input to.
2. a kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory according to claim 1, special
Sign is that the isolated island micro-capacitance sensor is divided into two parts of physical layer and network layer；The physical layer mainly includes controller and divides
Cloth power supply, controller stablizes each distributed generation resource output, then the output of distributed generation resource is connected by static transfer switch
To point of common coupling；Or controller stablizes each distributed generation resource output, then is powered with distributed generation resource to load；The net
Network layers carry out the data exchange between different electronic power inverters, and the output for completing distributed generation resource in entire micro-capacitance sensor is same
Step.
3. a kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory according to claim 1, special
Sign is, to acting on behalf of DGiOutput error can use consistency algorithm are as follows:
Wherein, eiAct on behalf of the global output error of i, xiAct on behalf of the real output value of i, xjAct on behalf of the adjacent generation of i
The value that reason j is come by communications, xrefThe nominal value for acting on behalf of the output of i, the Leader as consistency algorithm are saved
Point, aijIt indicates weighted adjacent coefficient, that is, acts on behalf of the network communication links of i, biThe weight for leading to Leader is represented, and if only if generation
When managing the accessible Leader of i, bi> 0；
In order to realize the regulating effect of finite time, created symbol function:
The finite-time control device of design are as follows:
ui=β sig (ei)α
Wherein, α, β are finite-time control parameters, and α is index coefficient, and 0 < α < 1 improves system convergence performance, and β is ratio
Coefficient, β > 0 determine the step-length of control variable.
4. a kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory according to claim 1, special
Sign is, carries out prediction steps to the historical data of contiguous agent based on extreme learning machine is improved specifically:
(1) foundation of prediction model
For any N number of different sample (Xi,Yi), wherein Xi=[xi1 xi2 … xin]T∈RnIndicate each DG history output frequency
Rate or voltage, Yi=[yi1 yi2 … yim]T∈RmThe frequency or voltage for indicating each DG desired output have L a hidden for one
The neural networks with single hidden layer of node layer can indicate are as follows:
Wherein g (x) is excitation function, generally Gaussian function, Wi=[wi1 wi2 … win]TFor input weight, γiFor output
Weight, ciFor the biasing of i-th of Hidden unit, ojIndicate prediction output；
When considering empiric risk and structure risk, the mathematical model of ES-ELM can be indicated are as follows:
Wherein, | | ε | |2Indicate empiric risk, | | γ | |2Indicate structure risk, η ∈ R indicates that two kinds of risk ratio parameters pass through friendship
The best compromise point for pitching the mode of verifying to determine；
LS-SVM algorithm is introduced, is Lagrange's equation by ES-ELM model conversation are as follows:
Wherein, λ=[λ1 λ2 … λN] indicating Lagrange multiplier, H indicates hidden layer output, γ=[γ1 γ2 … γL]TTable
Show output weight
According to KTT optimal conditions, seeking the gradient of Lagrange's equation and enabling it is 0, is obtained
To obtain
Wherein X+The generalized inverse matrix of representing matrix X；
(2) training of prediction model
First clear the following contents before training: the N number of different sample (X of training seti,Yi), hidden layer number L, excitation function g (x),
Step1: when data variation is larger, better training result, pre-processes data in order to obtain, specific formula
Are as follows:
Wherein, x (i) and x'(i) respectively indicate initial data and treated data, ExAnd σiIndicate initial data mean value and
Standard deviation；
Step2: input weight wikWith biasing ciIt is arbitrarily set in (0,1) range, and calculates hidden layer output H；
Step3: according toCalculate output weight γ and Lagrange multiplier λ；
(3) prediction of result
After training by ES-ELM, by historical data to the DG for having lost the communication informationiLocal prediction is carried out, respectively
To the predicted value of frequency and voltage.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910495672.7A CN110190599B (en) | 2019-06-10 | 2019-06-10 | Island microgrid control method based on finite time consistency theory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910495672.7A CN110190599B (en) | 2019-06-10 | 2019-06-10 | Island microgrid control method based on finite time consistency theory |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110190599A true CN110190599A (en) | 2019-08-30 |
| CN110190599B CN110190599B (en) | 2020-12-22 |
Family
ID=67720957
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910495672.7A Active CN110190599B (en) | 2019-06-10 | 2019-06-10 | Island microgrid control method based on finite time consistency theory |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110190599B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110611333A (en) * | 2019-11-11 | 2019-12-24 | 燕山大学 | Island microgrid distributed coordination control method considering communication problem |
| CN110994581A (en) * | 2019-11-14 | 2020-04-10 | 国网北京市电力公司 | Control processing method and device for direct-current micro-grid |
| CN112086996A (en) * | 2020-08-08 | 2020-12-15 | 东北电力大学 | Agent-based improved droop control method for parallel inverter |
| CN112260251A (en) * | 2020-10-12 | 2021-01-22 | 国网河北省电力有限公司经济技术研究院 | Microgrid control period stability analysis method and system |
| CN112713581A (en) * | 2020-12-17 | 2021-04-27 | 华中科技大学 | Distributed fixed-time voltage regulation and current equalization method and system for direct-current micro-grid |
| CN112788048A (en) * | 2021-01-22 | 2021-05-11 | 新华三信息安全技术有限公司 | Authentication information synchronization method and device |
| CN112909954A (en) * | 2020-12-30 | 2021-06-04 | 燕山大学 | Microgrid hierarchical control strategy based on generation of antagonistic neural network |
| CN113485126A (en) * | 2021-08-23 | 2021-10-08 | 安徽工业大学 | Improved dynamic matrix control cubic control method for direct-current micro-grid cluster |
| CN114243767A (en) * | 2021-12-06 | 2022-03-25 | 电子科技大学长三角研究院(湖州) | Island microgrid secondary controller design method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106129999A (en) * | 2016-07-01 | 2016-11-16 | 东南大学 | Based on finite time conforming direct-current grid distributed collaboration control method |
| US20180138705A1 (en) * | 2016-03-10 | 2018-05-17 | Southeast University | Distributed predictive control based voltage restoration scheme for microgrids |
| CN108075487A (en) * | 2017-12-05 | 2018-05-25 | 燕山大学 | The hierarchical control method for the isolated island micro-capacitance sensor that adaptive sagging and uniformity is combined |
| CN108667067A (en) * | 2018-04-04 | 2018-10-16 | 燕山大学 | A kind of isolated island micro-capacitance sensor hierarchical control method based on dual SMC- congruity theories |
| CN109687526A (en) * | 2019-03-06 | 2019-04-26 | 华北电力大学 | A kind of isolated island micro-capacitance sensor layered distribution type control strategy based on congruity theory |
-
2019
- 2019-06-10 CN CN201910495672.7A patent/CN110190599B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180138705A1 (en) * | 2016-03-10 | 2018-05-17 | Southeast University | Distributed predictive control based voltage restoration scheme for microgrids |
| CN106129999A (en) * | 2016-07-01 | 2016-11-16 | 东南大学 | Based on finite time conforming direct-current grid distributed collaboration control method |
| CN108075487A (en) * | 2017-12-05 | 2018-05-25 | 燕山大学 | The hierarchical control method for the isolated island micro-capacitance sensor that adaptive sagging and uniformity is combined |
| CN108667067A (en) * | 2018-04-04 | 2018-10-16 | 燕山大学 | A kind of isolated island micro-capacitance sensor hierarchical control method based on dual SMC- congruity theories |
| CN109687526A (en) * | 2019-03-06 | 2019-04-26 | 华北电力大学 | A kind of isolated island micro-capacitance sensor layered distribution type control strategy based on congruity theory |
Non-Patent Citations (2)
| Title |
|---|
| 孙伟 等: "基于改进极限学习机的电力需求预测研究", 《计算机与数字工程》 * |
| 顾伟 等: "基于有限时间一致性的直流微电网分布式协同控制", 《电力系统自动化》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110611333B (en) * | 2019-11-11 | 2022-12-13 | 燕山大学 | Island microgrid distributed coordination control method considering communication problem |
| CN110611333A (en) * | 2019-11-11 | 2019-12-24 | 燕山大学 | Island microgrid distributed coordination control method considering communication problem |
| CN110994581A (en) * | 2019-11-14 | 2020-04-10 | 国网北京市电力公司 | Control processing method and device for direct-current micro-grid |
| CN112086996A (en) * | 2020-08-08 | 2020-12-15 | 东北电力大学 | Agent-based improved droop control method for parallel inverter |
| CN112260251A (en) * | 2020-10-12 | 2021-01-22 | 国网河北省电力有限公司经济技术研究院 | Microgrid control period stability analysis method and system |
| CN112713581B (en) * | 2020-12-17 | 2022-08-05 | 华中科技大学 | Distributed fixed-time voltage regulation and current equalization method and system for direct-current micro-grid |
| CN112713581A (en) * | 2020-12-17 | 2021-04-27 | 华中科技大学 | Distributed fixed-time voltage regulation and current equalization method and system for direct-current micro-grid |
| CN112909954A (en) * | 2020-12-30 | 2021-06-04 | 燕山大学 | Microgrid hierarchical control strategy based on generation of antagonistic neural network |
| CN112909954B (en) * | 2020-12-30 | 2023-12-15 | 燕山大学 | Micro-grid layered control strategy based on generation of antagonistic neural network |
| CN112788048B (en) * | 2021-01-22 | 2022-04-01 | 新华三信息安全技术有限公司 | Authentication information synchronization method and device |
| CN112788048A (en) * | 2021-01-22 | 2021-05-11 | 新华三信息安全技术有限公司 | Authentication information synchronization method and device |
| CN113485126A (en) * | 2021-08-23 | 2021-10-08 | 安徽工业大学 | Improved dynamic matrix control cubic control method for direct-current micro-grid cluster |
| CN114243767A (en) * | 2021-12-06 | 2022-03-25 | 电子科技大学长三角研究院(湖州) | Island microgrid secondary controller design method |
| CN114243767B (en) * | 2021-12-06 | 2023-11-21 | 电子科技大学长三角研究院(湖州) | Island micro-grid secondary controller design method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110190599B (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110190599A (en) | A kind of isolated island micro-capacitance sensor control strategy based on finite time congruity theory | |
| CN109687526B (en) | Island micro-grid layered distributed control strategy based on consistency theory | |
| Lai et al. | Distributed multi-DER cooperative control for master-slave-organized microgrid networks with limited communication bandwidth | |
| Peng et al. | Voltage-based distributed optimal control for generation cost minimization and bounded bus voltage regulation in DC microgrids | |
| CN106505632B (en) | The planing method of the distributed electrical source position and capacity of middle pressure independent micro-grid system | |
| WO2023179003A1 (en) | Control method for multi-time-scale active distribution network voltage | |
| Zou | Design of reactive power optimization control for electromechanical system based on fuzzy particle swarm optimization algorithm | |
| CN105391071A (en) | Multifunctional grid-connected parallel inverter group intelligent control method used in microgrid | |
| CN114362267B (en) | Distributed coordination optimization method for AC/DC hybrid power distribution network considering multi-objective optimization | |
| CN113516278B (en) | Active power distribution network multi-time scale active and reactive power coordination optimization scheduling method and system | |
| CN113541146B (en) | Power flow calculation optimization method of power system considering distributed power supply | |
| Wu et al. | Distributed optimal operation of hierarchically controlled microgrids | |
| CN111725798A (en) | Distributed economic dispatching prediction control method for direct-current micro-grid cluster | |
| CN103178534A (en) | Calculating method for prevention and control strategy of small interference stabilization | |
| CN110445186B (en) | Self-synchronizing microgrid control system and secondary frequency modulation control method | |
| CN108536917A (en) | A kind of distributed computing method of transmission and distribution network overall situation Voltage Stability Control | |
| CN114421526A (en) | Distributed photovoltaic multi-cluster voltage control method and system and storage medium | |
| CN110429578A (en) | A kind of distributed DC grid control method | |
| CN104300546B (en) | A kind of Voltage Stability Constraints idle work optimization method based on wide area measurement information | |
| CN110350540A (en) | A kind of fine cutting load method based on On-line Estimation frequency character of load | |
| Bhukya et al. | Coordinated control and parameters optimization for PSS, POD and SVC to enhance the transient stability with the integration of DFIG based wind power systems | |
| CN116937606A (en) | Reactive voltage autonomous coordination control method for power distribution network | |
| CN105470976A (en) | Coordinated configuration method for SVC and TCSC under steady state condition | |
| Li et al. | Two‐stage voltage control strategy in distribution networks with coordinated multimode operation of PV inverters | |
| Gong et al. | The real-time optimization of active distribution system based on deep deterministic policy gradient |
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 | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240311 Address after: 071000 200m north of the intersection of Dingzhou commercial street and Xingding Road, Baoding City, Hebei Province (No. 1910, 19th floor, building 3, Jueshan community) Patentee after: Hebei Kaitong Information Technology Service Co.,Ltd. Country or region after: China Address before: 066004 No. 438 west section of Hebei Avenue, seaport District, Hebei, Qinhuangdao Patentee before: Yanshan University Country or region before: China |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240327 Address after: 200000 No. 1954, Huashan Road, Shanghai, Xuhui District Patentee after: Ai Qian Country or region after: China Address before: 071000 200m north of the intersection of Dingzhou commercial street and Xingding Road, Baoding City, Hebei Province (No. 1910, 19th floor, building 3, Jueshan community) Patentee before: Hebei Kaitong Information Technology Service Co.,Ltd. Country or region before: China |
|
| TR01 | Transfer of patent right |