CN108054753A - A kind of directly driven wind-powered field group of planes division methods of meter and low-voltage crossing characteristic - Google Patents
A kind of directly driven wind-powered field group of planes division methods of meter and low-voltage crossing characteristic Download PDFInfo
- Publication number
- CN108054753A CN108054753A CN201711394587.9A CN201711394587A CN108054753A CN 108054753 A CN108054753 A CN 108054753A CN 201711394587 A CN201711394587 A CN 201711394587A CN 108054753 A CN108054753 A CN 108054753A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- wind
- wind turbines
- cluster centre
- 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 16
- 238000007599 discharging Methods 0.000 claims abstract description 31
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 29
- UUTFFHNHMRSARV-UHFFFAOYSA-N Setarin Chemical compound C1=CC=CC2=C1OC(=O)C=C2OC(=C)C UUTFFHNHMRSARV-UHFFFAOYSA-N 0.000 claims description 6
- 230000005611 electricity Effects 0.000 claims description 6
- 238000012217 deletion Methods 0.000 claims description 4
- 230000037430 deletion Effects 0.000 claims description 4
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical class C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 3
- 230000009471 action Effects 0.000 claims description 3
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 125000004122 cyclic group Chemical group 0.000 claims description 3
- 230000006855 networking Effects 0.000 claims description 3
- 238000005192 partition Methods 0.000 claims description 3
- 201000010099 disease Diseases 0.000 claims 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims 1
- 238000012512 characterization method Methods 0.000 abstract description 8
- 230000006872 improvement Effects 0.000 abstract description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000011217 control strategy Methods 0.000 description 4
- 230000001052 transient effect Effects 0.000 description 3
- HEFNNWSXXWATRW-UHFFFAOYSA-N Ibuprofen Chemical compound CC(C)CC1=CC=C(C(C)C(O)=O)C=C1 HEFNNWSXXWATRW-UHFFFAOYSA-N 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001427 coherent effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000003786 synthesis reaction Methods 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
- H02J3/24—Arrangements for preventing or reducing oscillations of power in networks
-
- H02J3/386—
-
- 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]
-
- 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
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/70—Wind energy
- Y02E10/76—Power conversion electric or electronic aspects
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Wind Motors (AREA)
- Control Of Eletrric Generators (AREA)
Abstract
The invention discloses it is a kind of count and low-voltage crossing characteristic directly driven wind-powered field group of planes division methods, including：1st, the detailed model of directly driven wind-powered field is established；2nd, the input wind energy of each Wind turbines is gathered；3rd, the set end voltage value of each Wind turbines of fault moment is gathered；Net side current transformer watt current reference value when the 4th, setting grid voltage sags；5th, each Wind turbines discharging circuit conducting situation is judged；6th, a directly driven wind-powered field group of planes is completed using the improvement K mean cluster algorithm that random k values and sensitive cluster centre is immunized to divide.The present invention can each Wind turbines discharging circuit conducting situation in the directly driven wind-powered field of accurate characterization, directly driven wind-powered field group of planes division quality is improved, so as to effectively improve fitting precision of the directly driven wind-powered field Equivalent Model to wind power plant output external characteristic.
Description
Technical field
The present invention relates to directly driven wind-powered field equivalent modeling technical field, counted and low-voltage crossing characteristic more particularly to a kind of
Directly driven wind-powered field group of planes division methods.
Background technology
Direct-drive permanent-magnetism Wind turbines eliminate the gear-box of high failure rate, have mechanical loss small, and operational efficiency is high, safeguards
It is at low cost, the advantages that speed-regulating range width, it is connected by full power convertor with power grid, generator is not direct to be coupled with power grid, with
Double-fed fan motor unit, which is compared, has stronger low-voltage crossing ability, has become the mainstream wind-powered electricity generation type of current wind power plant.With
The fast development of wind generating technology, the scale of integrated wind plant gradually expand, if every Wind turbines all use detailed mould
Type will appear from the problems such as the simulation calculation time is too long, committed memory is excessive, trend is difficult to restrain.Therefore, it is suitable to establish
Wind power plant Equivalent Model is very necessary.
At present, the equivalence method of directly driven wind-powered field mainly has unit characterization method and multimachine characterization method.Unit characterization method is applicable in
In scale is smaller, Wind turbines are distributed on geographical location compares concentration, Wind turbines electrical distance and input wind speed difference compared with
Small wind power plant.As wind power plant scale constantly expands, unit characterization method would generally generate large error.Multimachine characterization method increases
The division link of Coherent Generator Group, main thought are that have close operating point as group of planes division principle using Wind turbines, using gathering
Class algorithm carries out group of planes division.At present, for direct-drive permanent-magnetism Wind turbines, common point of group's index mainly has input wind speed, wind
Motor group surveys active data and divides group's index by the synthesis that wind speed, active power, voltage, electric current etc. form.Kmeans gathers
Class algorithm is a kind of commonly used clustering algorithm of group of planes division, has and simply easily realizes, operating rate is fast, can handle large-scale number
The advantages of according to collection.
However, existing multimachine characterization method when carrying out group of planes division, fails to take into full account Wind turbines generator terminal electricity mostly
Pressure falls difference, has ignored influence of the discharging circuit conducting situation to direct-drive permanent-magnetism Wind turbines transient characterisitics otherness, it is impossible to
The output external characteristic of enough accurately fitting wind power plants.In addition, there is tradition kmeans clustering algorithms cluster number k values need to give in advance
It is fixed；The selection of initial cluster center is depended critically upon, cluster result is made easily to be absorbed in locally optimal solution；Easily by isolated point data
The defects of influence.The disadvantages described above of traditional kmeans clustering algorithms significantly reduces clustering result quality, and a group of planes is caused to divide
As a result inaccuracy and unstability.
The content of the invention
In place of the present invention is avoids above-mentioned the deficiencies in the prior art, propose it is a kind of count and low-voltage crossing characteristic it is directly driven wind-powered
Field group of planes division methods to each Wind turbines discharging circuit conducting situation in energy accurate characterization wind power plant, improve straight wind dispelling
An electric field group of planes divides quality, so as to effectively improve fitting essence of the directly driven wind-powered field Equivalent Model to wind power plant output external characteristic
Degree.
The present invention adopts the following technical scheme that solve technical problem：
The characteristics of directly driven wind-powered field group of planes division methods of a kind of meter of the present invention and low-voltage crossing characteristic, is, including following
Step：
Step 1, the detailed model for establishing directly driven wind-powered field；
The directly driven wind-powered field is made of the identical direct-drive permanent-magnetism Wind turbines of n bench-types number, and the detailed model includes：
The unit model of each Wind turbines, current collection circuit model, generator terminal transformer model and main transformer model in wind power plant；Its
In, the unit model of the Wind turbines includes：Wind energy conversion system model, magneto alternator model, full power convertor and its
Control system model, variable-pitch control system model, discharging circuit and its control system model；
Step 2, the input wind energy for gathering each Wind turbines；
Gather the real-time wind speed of the n platform direct-drive permanent-magnetism Wind turbines of the directly driven wind-powered field, and by each Wind turbines
Input wind energy is denoted as { P1w,P2w,···,Pjw,···,Pnw, PjwRepresent the input wind energy of jth platform Wind turbines；
Step 3, the set end voltage value for gathering each Wind turbines of fault moment；
It is set in t0Moment sets three phase short circuit fault in the exit of the directly driven wind-powered field, and gathers t0Each of moment
The set end voltage value of Wind turbines, is denoted as { U1(t0),U2(t0),···,Uj(t0),···,Un(t0), Uj(t0) represent t0
The set end voltage value of moment jth platform Wind turbines, takes perunit value；In t1Moment cuts off the three phase short circuit fault；
Net side current transformer watt current reference value when step 4, setting grid voltage sags；
Step 4.1, the command value i using net side current transformer DC voltage pi regulator during formula (1) acquisition stable statedref1：
In formula (1), KdpAnd KdiThe proportionality coefficient and integration of DC bus-bar voltage control outer shroud respectively in net side current transformer
Coefficient；UdcFor stable state when Wind turbines DC bus-bar voltage, take perunit value；Join for the DC bus-bar voltage of Wind turbines
Value is examined, takes perunit value；
Step 4.2, during the directly driven wind-powered field three phase short circuit fault, when Wind turbines grid entry point voltage is less than nominal
Voltage 20% when, Wind turbines are cut out from power grid；
When Wind turbines grid entry point voltage is in 20%~90% section of nominal voltage, formula (2) is utilized to calculate net side
The reactive current reference value i of current transformerqref2(t)：
iqref2(t)=1.5 (0.9-Ug(t))IN(0.2pu≤Ug(t)≤0.9pu) (2)
In formula (2), Ug(t) it is the perunit value of Wind turbines grid entry point voltage；INFor the perunit of net side current transformer rated current
Value；
Step 4.3 obtains watt current reference value limits value i using formula (3)dref2(t)：
In formula (3), imaxThe maximum current value allowed for net side current transformer；
The command value i of DC voltage pi regulator when step 4.4, selection stable statedref1With watt current reference value limits value
idref2(t) net side current transformer watt current reference value i when smaller value is as grid voltage sags indref(t)；
Step 5 judges each Wind turbines discharging circuit conducting situation；
Step 5.1, in the t1Moment obtains jth typhoon motor networking side power converter dc bus using formula (4)
Voltage Ujdc(t1)：
In formula (4), CjFor the dc-link capacitance value of jth platform Wind turbines, SBFor the reference capacity of net side current transformer；
Step 5.2 compares t1The DC bus-bar voltage U of moment jth platform Wind turbinesjdc(t1) and discharging circuit action threshold value
Udc_inSize, if Ujdc(t1) it is more than Udc_in, then judge that jth platform Wind turbines discharging circuit turns on；Conversely, judge jth typhoon
Motor group discharging circuit does not turn on；
Step 6 improves the directly driven wind-powered field machine of K mean cluster algorithm partition using immune random k values and sensitive cluster centre
Group；
The Wind turbines of all discharging circuit conductings in n platform direct-drive permanent-magnetism Wind turbines are divided into a machine by step 6.1
Group；
Step 6.2, using the set end voltage value of each Wind turbines of fault moment as group's index is divided, using immune random k values
K mean cluster algorithm is improved with sensitive cluster centre, the m platform Wind turbines that remaining discharging circuit does not turn on are divided into k
A group of planes.
The characteristics of directly driven wind-powered field group of planes division methods of the present invention, lies also in, and in the step 6.2, is immunized random
It is to carry out as follows that k values and sensitive cluster centre, which improve K mean cluster algorithm,：
Step 6.2.1, the set end voltage value { U for the m platform Wind turbines for not turning on remaining discharging circuit1(t0),U2
(t0),···,Up(t0),···,Uq(t0),···,Um(t0) as sample data intersection, it is denoted as S={ x1,
x2,···,xp,···,xq,···,xm}；Wherein, Up(t0) and Uq(t0) pth platform and q typhoon motors are represented respectively
The set end voltage value of group, takes perunit value, and by Up(t0) and Uq(t0) data object x is denoted as respectivelypAnd xq, p, q=1,
2,···,m,p≠q；
Step 6.2.2, data object x is calculatedpAnd xqEuclidean distance dist (xp,xq)；
Step 6.2.3, being averaged between any two data object in the sample data intersection S is obtained using formula (5)
Distance MeanDist：
In formula (5),To appoint the number of combinations for taking 2 data objects from n data object；
Step 6.2.4, define：With data object xpCentered on, the region using average distance MeanDist as radius is known as
Data object xpNeighborhood, the data object xpNeighborhood in the number of data object be known as data object xpBased on distance
The density parameter of MeanDist；
Data object x is obtained using formula (6)pDensity parameter density (xp, MeanDist), so as to obtain m data
The density parameter of object：
In formula (6), u (MeanDist-dist (xp,xq)) represent a function, and have：
Step 6.2.5, by the density parameter of the m data object, the preceding M data object of density parameter maximum
Density parameter is added in alternative point set D, D={ density (xp, MeanDist), p=1,2, M }；
Step 6.2.6, cyclic variable r is defined, and initializes r=1；Definition cluster number is k, and initializes k=1；It is fixed
Maximum similarity is AMS between the adopted class that is initially averagedr-1, and initialize AMSr-1；It defines cluster centre collection and is combined into Ar-1, and initialize
For empty set；Using the alternative point set D as the r-1 times alternative point set Dr-1；
Step 6.2.7, from the r-1 times alternative point set Dr-1In select density parameter maximum two data objects average
As the r times cluster centre crIt is put into cluster centre set Ar-1In, obtain the r times cluster centre set Ar；Meanwhile by selected by
Two data objects corresponding to density parameter from the r-1 times alternative point set Dr-1Middle deletion, so as to obtain the r times alternative point
Collect Dr；
Step 6.2.8, from the r times alternative point set DrMiddle selection and the r times cluster centre set ArIn cluster centre distance
Farthest data object, as the r+1 times cluster centre cr+1It is put into the r times cluster centre set ArIn, it obtains the r+1 times and gathers
Class centralization Ar+1；Meanwhile by the r+1 times cluster centre cr+1Corresponding density parameter is from the r times alternative point set DrIn delete
It removes, so as to obtain the r+1 times alternative point set Dr+1；
Step 6.2.9, k+1 is assigned to k；
Step 6.2.10, remaining data object in the sample data intersection S is calculated respectively in being clustered with the r+1 times
Heart set Ar+1In each cluster centre Euclidean distance, and each data object is assigned in the nearest cluster of Euclidean distance
In class where the heart, so as to obtain k class；
Step 6.2.11, the data object x of arbitrary i-th class in k class is obtained using formula (8)pTo the cluster centre of the i-th class
ciThe distance between average si：
In formula (8), PiFor the total number of data object in the i-th class；I=1,2, k；
Step 6.2.12, the cluster centre c of the i-th class is obtained using formula (9)iWith the cluster centre c of jth classjThe distance between
di,j：
di,j=dist (ci,cj) (9)
In formula (9), i, j=1,2, k, i ≠ j；
Step 6.2.13, maximum similarity AMS between the r times average class of formula (10) acquisition is utilizedr：
Step 6.2.14, AMS is judgedr＜ AMSr-1It is whether true, if so, then go to step 6.2.15；Otherwise, go to
Step 6.2.19；
Step 6.2.15, newer cluster centre c is obtained using formula (11)i′：
In formula (11), PiFor the total number of data object in the i-th class, xpFor the data object of arbitrary i-th class, i=1,
2,···,k；
Step 6.2.16, described the r+1 times alternative point set D is calculated respectivelyr+1In any one density parameter corresponding to
The sum of Euclidean distance between data object and all newer cluster centres, so as to obtain the number corresponding to all density parameters
According to the set of object the sum of Euclidean distance between all newer cluster centres respectively, from the set of the sum of Euclidean distance
The data object corresponding to maximum is chosen as the r+2 times cluster centre cr+2And it is put into the r+1 times cluster centre set Ar+1
In, so as to obtain the r+2 times cluster centre set Ar+2；
Step 6.2.17, by the r+2 times cluster centre cr+2Corresponding density parameter is from the r+1 times alternative point set Dr+1In
It deletes, so as to obtain the r+2 times alternative point set Dr+2；
Step 6.2.18, after r+1 being assigned to r, 6.2.9 is gone to step；
Step 6.2.19, with AMSr-1In initial clustering of the k corresponding cluster centre as kmeans clustering algorithms
The heart carries out kmeans clustering algorithms to the sample data intersection S, obtains a k group of planes.
Compared with the prior art, the present invention has the beneficial effect that：
1st, directly driven wind-powered field group of planes division methods proposed by the present invention, with the off-load of each Wind turbines in directly driven wind-powered field
Circuit turn-on situation is as dividing group's foundation, using each Wind turbines set end voltage value in the directly driven wind-powered field of fault moment as dividing group
Index is completed a directly driven wind-powered field group of planes using the improvement K mean cluster algorithm that random k values and sensitive cluster centre is immunized and is divided,
The effective output external characteristic for being fitted directly driven wind-powered field accurately reflects the actual motion state of directly driven wind-powered field, improves
The accuracy and validity of directly driven wind-powered field group of planes division result.
2nd, the present invention adjusts net according to the drop depth of Wind turbines grid entry point voltage during grid side generation short trouble
The reactive current reference value of side converter so as to set the watt current reference value of net side current transformer, embodies direct-drive permanent-magnetism wind
The handoff procedure of motor group control strategy under low-voltage crossing pattern meets the Wind turbines of grid-connected directive/guide proposition in China in electricity
Off-grid does not run continuously and the requirement of reactive power support is provided to power grid when short trouble occurs for net side.
3rd, the immune random k values and the improvement K mean cluster algorithm of sensitive cluster centre that the present invention uses, by comparing flat
The value of maximum similarity index AMS automatically determines optimal cluster number k between equal class, solves traditional kmeans algorithms k values
The problem of need to giving in advance, improves the quality of cluster.
4th, the immune random k values and the improvement K mean cluster algorithm of sensitive cluster centre that the present invention uses, can dynamically adjust
It saves current cluster centre and dynamically adds next cluster centre, the k cluster centre finally obtained is by k-1 dynamic
The optimal k cluster centre for distributing and obtaining, generates in strict accordance with the characteristics of data object, effectively prevents cluster result
Fluctuation, solve the problems, such as traditional kmeans algorithms sensitivity initial cluster center, ability of searching optimum improved, larger
The Stability and veracity of cluster is improved in degree.
5th, the immune random k values and the improvement K mean cluster algorithm of sensitive cluster centre that the present invention uses, can will isolate
Point data is individually divided into a group of planes, reduces influence of the isolated point data to group of planes division result, avoids cluster centre
Tend to isolated point data away from data-intensive district, improve the quality of cluster.
Description of the drawings
Fig. 1 is the flow chart of directly driven wind-powered field group of planes division methods proposed by the present invention；
Fig. 2 is the topology diagram of directly driven wind-powered field in the present invention；
Fig. 3 is the topology diagram of direct-drive permanent-magnetism Wind turbines in the present invention；
Fig. 4 is net side current transformer control strategy block diagram in the present invention；
Fig. 5 improves K mean cluster algorithm flow chart for random k values and sensitive cluster centre are immunized in the present invention.
Specific embodiment
Below in conjunction with the accompanying drawings, embodiments of the present invention are illustrated.
In the present embodiment, a kind of directly driven wind-powered field group of planes division methods of meter and low-voltage crossing characteristic, it is contemplated that grid side
When breaking down, discharging circuit turns on influence of the situation to direct-drive permanent-magnetism Wind turbines transient response characteristic, according to dc bus
The variation of voltage judges the discharging circuit conducting situation of each Wind turbines, with each typhoon motor in the directly driven wind-powered field of fault moment
Group set end voltage value is as group of planes Classification Index, using the improvement K mean cluster algorithm that random k values and sensitive cluster centre is immunized
Directly driven wind-powered field group of planes division is completed, the Stability and veracity of group of planes division is improved, optimizes wind power plant multimachine equivalence and build
Mould method is laid a good foundation for directly driven wind-powered field electro-magnetic transient equivalent modeling.
The particular flow sheet of directly driven wind-powered field group of planes division methods proposed by the present invention is as shown in Figure 1, mainly include following
Step：
Step 1, the detailed model for establishing directly driven wind-powered field；
Directly driven wind-powered field is made of the identical direct-drive permanent-magnetism Wind turbines of n bench-types number, and detailed model includes：In wind power plant
Unit model, current collection circuit model, generator terminal transformer model and the main transformer model of each Wind turbines；Directly driven wind-powered field
Topology diagram as shown in Fig. 2, every direct-drive permanent-magnetism Wind turbines by its generator terminal transformer boosting after, pass through overhead transmission line
It is connected in wind power plant and presses busbar, then after wind power plant main transformer secondary booster, pass through from directly driven wind-powered field exit double
Return overhead transmission line access power grid；
The topology diagram of direct-drive permanent-magnetism Wind turbines is as shown in figure 3, the unit model of Wind turbines includes：Wind energy conversion system mould
Type, magneto alternator model, full power convertor and its control system model, variable-pitch control system model, unload it is charged
Road and its control system model；
Step 2, the input wind energy for gathering each Wind turbines；
Every the real-time wind speed of the n platform direct-drive permanent-magnetism Wind turbines of 5 minutes acquisition directly driven wind-powered fields, the company of randomly selecting
It is continuous 10 it is small when interior 50 groups of real-time wind speed of each Wind turbines and averaged, the input wind energy of each Wind turbines is denoted as
{P1w,P2w,···,Pjw,···,Pnw, PjwRepresent the input wind energy of jth platform Wind turbines.
Step 3, the set end voltage value for gathering each Wind turbines of fault moment；
It is set in t0At the moment, the exit of directly driven wind-powered field sets three phase short circuit fault, and gathers t0Moment each typhoon motor
The set end voltage value of group, is denoted as { U1(t0),U2(t0),···,Uj(t0),···,Un(t0)}；Uj(t0) represent t0Moment
The set end voltage value of j platform Wind turbines, takes perunit value；In t1Moment cuts off three phase short circuit fault；
The outlet that three phase short circuit fault point is arranged on directly driven wind-powered field everywhere, can fully demonstrate the circuit in wind power plant
Influence of the impedance to Wind turbines discharging circuit conducting situation.
Net side current transformer watt current reference value when step 4, setting grid voltage sags；
Net side current transformer control strategy block diagram is as shown in Figure 4.According to control strategy shown in Fig. 4, net side unsteady flow during stable state
Device watt current reference value and net side current transformer reactive current reference value use the command value of outer voltage pi regulator；Wind-powered electricity generation
During set grid-connection point Voltage Drop, joined by Wind turbines grid entry point Voltage Drop degree to set net side current transformer reactive current
Value is examined, so as to set net side current transformer watt current reference value.
Step 4.1, when network voltage is normal, net side current transformer performs active preferential maximum power tracing control mould
Formula, the command value i of net side current transformer DC voltage pi regulator when obtaining stable state using formula (1)dref1：
In formula (1), KdpAnd KdiThe proportionality coefficient and integration of DC bus-bar voltage control outer shroud respectively in net side current transformer
Coefficient；UdcFor stable state when Wind turbines DC bus-bar voltage, take perunit value；Join for the DC bus-bar voltage of Wind turbines
Value is examined, takes perunit value；
Step 4.2, during directly driven wind-powered field three phase short circuit fault, when Wind turbines grid entry point voltage be less than nominal voltage
20% when, Wind turbines are cut out from power grid；
When Wind turbines grid entry point voltage is in 20%~90% section of nominal voltage, since net side current transformer is active
Preferential control model is difficult to give full play to reactive power support ability of the direct-drive permanent-magnetism Wind turbines to power grid, it is therefore desirable to by net side
Current transformer is switched to idle preferential static reactive operational mode.At this point, net side current transformer is without PI control models,
But the reactive current reference value of net side current transformer is adjusted according to the drop depth of Wind turbines grid entry point voltage, utilize formula
(2) the reactive current reference value i of net side current transformer is calculatedqref2(t)：
iqref2(t)=1.5 (0.9-Ug(t))IN(0.2pu≤Ug(t)≤0.9pu) (2)
In formula (2), Ug(t) it is the perunit value of Wind turbines grid entry point voltage；INFor the perunit of net side current transformer rated current
Value；
Step 4.3, since static reactive operational mode is using reactive current as main control object, it is therefore desirable to having
Work(current reference value is limited, and watt current reference value limits value i is obtained using formula (3)dref2(t)：
In formula (3), imaxThe maximum current value allowed for net side current transformer；
The command value i of DC voltage pi regulator when step 4.4, selection stable statedref1With watt current reference value limits value
idref2(t) net side current transformer watt current reference value i when smaller value is as grid voltage sags indref(t)；
As the command value i of DC voltage pi regulatordref1Less than watt current reference value limits value idref2(t) when, explanation
Net side current transformer DC voltage outer shroud can still be adjusted DC voltage, net side current transformer active power reference value idref
(t) still it is set as the command value i of DC voltage pi regulatordref1；As the command value i of DC voltage pi regulatordref1It is more than
Watt current reference value limits value idref2(t) when, illustrate that DC voltage outer shroud can not effectively keep DC voltage
Stablize, need to put into DC side-discharging circuit at this time, consume the excess energy of DC side accumulation, be maintained at DC voltage
In safe range, net side current transformer active power reference value idref(t) it is set as watt current reference value limits value idref2(t)。
Step 5 judges each Wind turbines discharging circuit conducting situation；
Step 5.1, during directly driven wind-powered field three phase short circuit fault, the input wind energy of each Wind turbines remains unchanged,
In t1It moment, can be according to the input wind energy P of failure foreground Wind turbinesjwHaving for power grid is sent into Wind turbines net side current transformer
Work(power obtains jth typhoon motor networking side power converter DC bus-bar voltage U using formula (4)jdc(t1)：
In formula (4), CjFor the dc-link capacitance value of jth platform Wind turbines, SBFor the reference capacity of net side current transformer；
Step 5.2 compares t1The DC bus-bar voltage U of moment jth platform Wind turbinesjdc(t1) and discharging circuit action threshold value
Udc_inSize, if Ujdc(t1) it is more than Udc_in, then judge that jth platform Wind turbines discharging circuit turns on；Conversely, judge jth typhoon
Motor group discharging circuit does not turn on；
Step 6 improves the directly driven wind-powered field machine of K mean cluster algorithm partition using immune random k values and sensitive cluster centre
Group；
The Wind turbines of all discharging circuit conductings in n platform direct-drive permanent-magnetism Wind turbines are divided into a machine by step 6.1
Group；
Step 6.2, during directly driven wind-powered field three phase short circuit fault, due to full power convertor Fault Isolation act on,
Each Wind turbines set end voltage value is held essentially constant, and the set end voltage value using each Wind turbines of fault moment is divides group to refer to
Mark improves K mean cluster algorithm, the m that remaining discharging circuit is not turned on using immune random k values and sensitive cluster centre
Platform Wind turbines are divided into a k group of planes.
Fig. 5 is the flow chart for the improvements K mean cluster algorithm that random k values and sensitive cluster centre is immunized, mainly including with
Under several steps：
Step 6.2.1, the set end voltage value { U for the m platform Wind turbines for not turning on remaining discharging circuit1(t0),U2
(t0),···,Up(t0),···,Uq(t0),···,Um(t0) as sample data intersection, it is denoted as S={ x1,
x2,···,xp,···,xq,···,xm}；Wherein, Up(t0) and Uq(t0) pth platform and q typhoon motors are represented respectively
The set end voltage value of group, takes perunit value, and by Up(t0) and Uq(t0) data object x is denoted as respectivelypAnd xq, p, q=1,
2,···,m,p≠q；
Step 6.2.2, data object x is calculatedpAnd xqEuclidean distance dist (xp,xq)；
Similitude between data object is weighed with Euclidean distance, the smaller explanation of the Euclidean distance between data object
Data object is more similar；
Step 6.2.3, being averaged between any two data object in the sample data intersection S is obtained using formula (5)
Distance MeanDist：
In formula (5),To appoint the number of combinations for taking 2 data objects from n data object；
Step 6.2.4, define：With data object xpCentered on, the region using average distance MeanDist as radius is known as
Data object xpNeighborhood, the data object xpNeighborhood in the number of data object be known as data object xpBased on distance
The density parameter of MeanDist；
Data object x is obtained using formula (6)pDensity parameter density (xp, MeanDist), so as to obtain m data
The density parameter of object：
In formula (6), u (MeanDist-dist (xp,xq)) represent a function, and have：
Data object xpNeighborhood in data object number it is more, i.e. data object xpDensity parameter density (xp,
MeanDist it is) bigger, illustrate with data object xpClustering Effect as cluster centre is better.
Step 6.2.5, by the density parameter of m data object, the density of the preceding M data object of density parameter maximum
Parameter is added in alternative point set D, D={ density (xp, MeanDist), p=1,2, M }；In general, M values are
m/2；
Cluster centre is selected from M alternate data object of high density parameter, this can ensure in same class, gather
Class center is the data object of an opposite close quarters, ensure that the high similitude in class.
Step 6.2.6, cyclic variable r is defined, and initializes r=1；Definition cluster number is k, and initializes k=1；It is fixed
Maximum similarity is AMS between the adopted class that is initially averagedr-1, and initialize AMSr-1；It defines cluster centre collection and is combined into Ar-1, and initialize
For empty set；Using the alternative point set D as the r-1 times alternative point set Dr-1；
Step 6.2.7, from the r-1 times alternative point set Dr-1In select density parameter maximum two data objects average
As the r times cluster centre crIt is put into cluster centre set Ar-1In, obtain the r times cluster centre set Ar；Meanwhile by selected by
Two data objects corresponding to density parameter from the r-1 times alternative point set Dr-1Middle deletion, so as to obtain the r times alternative point
Collect Dr；
Step 6.2.8, from the r times alternative point set DrMiddle selection and the r times cluster centre set ArIn cluster centre distance
Farthest data object, as the r+1 times cluster centre cr+1It is put into the r times cluster centre set ArIn, it obtains the r+1 times and gathers
Class centralization Ar+1；Meanwhile by the r+1 times cluster centre cr+1Corresponding density parameter is from the r times alternative point set DrIn delete
It removes, so as to obtain the r+1 times alternative point set Dr+1；
Step 6.2.9, k+1 is assigned to k；
Step 6.2.10, remaining data object in sample data intersection S is calculated and the r+1 times cluster centre collection respectively
Close Ar+1In each cluster centre Euclidean distance, and each data object is assigned to the nearest cluster centre institute of Euclidean distance
Class in, so as to obtain k class；
Step 6.2.11, the data object x of arbitrary i-th class in k class is obtained using formula (8)pTo the cluster centre of the i-th class
ciThe distance between average si：
In formula (8), PiFor the total number of data object in the i-th class；I=1,2, k；
Step 6.2.12, the cluster centre c of the i-th class is obtained using formula (9)iWith the cluster centre c of jth classjThe distance between
di,j：
di,j=dist (ci,cj) (9)
In formula (9), i, j=1,2, L, k, i ≠ j；
Step 6.2.13, maximum similarity AMS between the r times average class of formula (10) acquisition is utilizedr：
Maximum similarity AMS represents the average of maximum similarity between each class between average class, when AMS obtains minimum value
When, illustrate that Clustering Effect at this time is optimal, k values at this time are exactly optimal cluster number.
Step 6.2.14, AMS is judgedr＜ AMSr-1It is whether true, if so, then go to step 6.2.15；Otherwise, go to
Step 6.2.19；
Step 6.2.15, newer cluster centre c ' is obtained using formula (11)i：
In formula (11), PiFor the total number of data object in the i-th class, xpFor the data object of arbitrary i-th class, i=1,
2,···,k；
Step 6.2.16, described the r+1 times alternative point set D is calculated respectivelyr+1In any one density parameter corresponding to
The sum of Euclidean distance between data object and all newer cluster centres, so as to obtain the number corresponding to all density parameters
According to the set of object the sum of Euclidean distance between all newer cluster centres respectively, from the set of the sum of Euclidean distance
The data object corresponding to maximum is chosen as the r+2 times cluster centre cr+2And it is put into the r+1 times cluster centre set Ar+1
In, so as to obtain the r+2 times cluster centre set Ar+2；
Selection and all newer cluster centre c ' from M alternate data object of high density parameteriBetween it is European
The maximum data object of the sum of distance is added to as new cluster centre in current cluster centre set, and so doing can make not
Similar cluster centre is as mutually exclusive as possible, so as to ensure that the low similitude between inhomogeneity.
Step 6.2.17, by the r+2 times cluster centre cr+2Corresponding density parameter is from the r+1 times alternative point set Dr+1In
It deletes, so as to obtain the r+2 times alternative point set Dr+2；
Step 6.2.18, after r+1 being assigned to r, 6.2.9 is gone to step；
Step 6.2.19, with AMSr-1In initial clustering of the k corresponding cluster centre as kmeans clustering algorithms
The heart carries out kmeans clustering algorithms to the sample data intersection S, obtains a k group of planes.
Claims (2)
1. the directly driven wind-powered field group of planes division methods of a kind of meter and low-voltage crossing characteristic, which is characterized in that comprise the following steps：
Step 1, the detailed model for establishing directly driven wind-powered field；
The directly driven wind-powered field is made of the identical direct-drive permanent-magnetism Wind turbines of n bench-types number, and the detailed model includes：Wind-powered electricity generation
Unit model, current collection circuit model, generator terminal transformer model and the main transformer model of each Wind turbines in；Wherein,
The unit model of the Wind turbines includes：Wind energy conversion system model, magneto alternator model, full power convertor and its control
System model, variable-pitch control system model, discharging circuit and its control system model；
Step 2, the input wind energy for gathering each Wind turbines；
Gather the real-time wind speed of the n platform direct-drive permanent-magnetism Wind turbines of the directly driven wind-powered field, and by the input of each Wind turbines
Wind energy is denoted as { P1w,P2w,…,Pjw,…,Pnw, PjwRepresent the input wind energy of jth platform Wind turbines；
Step 3, the set end voltage value for gathering each Wind turbines of fault moment；
It is set in t0Moment sets three phase short circuit fault in the exit of the directly driven wind-powered field, and gathers t0Moment each typhoon electricity
The set end voltage value of unit, is denoted as { U1(t0),U2(t0),…,Uj(t0),…,Un(t0), Uj(t0) represent t0Moment jth typhoon
The set end voltage value of motor group, takes perunit value；In t1Moment cuts off the three phase short circuit fault；
Net side current transformer watt current reference value when step 4, setting grid voltage sags；
Step 4.1, the command value i using net side current transformer DC voltage pi regulator during formula (1) acquisition stable statedref1：
<mrow>
<msub>
<mi>i</mi>
<mrow>
<mi>d</mi>
<mi>r</mi>
<mi>e</mi>
<mi>f</mi>
<mn>1</mn>
</mrow>
</msub>
<mo>=</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<msub>
<mi>K</mi>
<mrow>
<mi>d</mi>
<mi>p</mi>
</mrow>
</msub>
<mo>+</mo>
<mfrac>
<msub>
<mi>K</mi>
<mrow>
<mi>d</mi>
<mi>i</mi>
</mrow>
</msub>
<mi>s</mi>
</mfrac>
<mo>)</mo>
</mrow>
<mrow>
<mo>(</mo>
<msubsup>
<mi>U</mi>
<mrow>
<mi>d</mi>
<mi>c</mi>
</mrow>
<mo>*</mo>
</msubsup>
<mo>-</mo>
<msub>
<mi>U</mi>
<mrow>
<mi>d</mi>
<mi>c</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (1), KdpAnd KdiThe proportionality coefficient and integral coefficient of DC bus-bar voltage control outer shroud respectively in net side current transformer；
UdcFor stable state when Wind turbines DC bus-bar voltage, take perunit value；For the DC bus-bar voltage reference value of Wind turbines,
Take perunit value；
Step 4.2, during the directly driven wind-powered field three phase short circuit fault, when Wind turbines grid entry point voltage be less than nominal voltage
20% when, Wind turbines are cut out from power grid；
When Wind turbines grid entry point voltage is in 20%~90% section of nominal voltage, formula (2) is utilized to calculate net side unsteady flow
The reactive current reference value i of deviceqref2(t)：
iqref2(t)=1.5 (0.9-Ug(t))IN(0.2pu≤Ug(t)≤0.9pu) (2)
In formula (2), Ug(t) it is the perunit value of Wind turbines grid entry point voltage；INFor the perunit value of net side current transformer rated current；
Step 4.3 obtains watt current reference value limits value i using formula (3)dref2(t)：
<mrow>
<msub>
<mi>i</mi>
<mrow>
<mi>d</mi>
<mi>r</mi>
<mi>e</mi>
<mi>f</mi>
<mn>2</mn>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msqrt>
<mrow>
<msubsup>
<mi>i</mi>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
<mn>2</mn>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>i</mi>
<mrow>
<mi>q</mi>
<mi>r</mi>
<mi>e</mi>
<mi>f</mi>
<mn>2</mn>
</mrow>
<mn>2</mn>
</msubsup>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
</mrow>
</msqrt>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>3</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (3), imaxThe maximum current value allowed for net side current transformer；
The command value i of DC voltage pi regulator when step 4.4, selection stable statedref1With watt current reference value limits value idref2
(t) net side current transformer watt current reference value i when smaller value is as grid voltage sags indref(t)；
Step 5 judges each Wind turbines discharging circuit conducting situation；
Step 5.1, in the t1Moment obtains jth typhoon motor networking side power converter DC bus-bar voltage using formula (4)
Ujdc(t1)：
<mrow>
<munderover>
<mo>&Integral;</mo>
<msub>
<mi>t</mi>
<mn>0</mn>
</msub>
<msub>
<mi>t</mi>
<mn>1</mn>
</msub>
</munderover>
<msub>
<mi>C</mi>
<mi>j</mi>
</msub>
<msub>
<mi>U</mi>
<mrow>
<mi>j</mi>
<mi>d</mi>
<mi>c</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mfrac>
<mrow>
<msub>
<mi>dU</mi>
<mrow>
<mi>j</mi>
<mi>d</mi>
<mi>c</mi>
</mrow>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<mi>d</mi>
<mi>t</mi>
</mrow>
</mfrac>
<mo>=</mo>
<munderover>
<mo>&Integral;</mo>
<msub>
<mi>t</mi>
<mn>0</mn>
</msub>
<msub>
<mi>t</mi>
<mn>1</mn>
</msub>
</munderover>
<mrow>
<mo>(</mo>
<msub>
<mi>P</mi>
<mrow>
<mi>j</mi>
<mi>w</mi>
</mrow>
</msub>
<mo>-</mo>
<mfrac>
<mn>3</mn>
<mn>2</mn>
</mfrac>
<msub>
<mi>U</mi>
<mi>g</mi>
</msub>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
<msub>
<mi>i</mi>
<mrow>
<mi>d</mi>
<mi>r</mi>
<mi>e</mi>
<mi>f</mi>
</mrow>
</msub>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
<msub>
<mi>S</mi>
<mi>B</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>4</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (4), CjFor the dc-link capacitance value of jth platform Wind turbines, SBFor the reference capacity of net side current transformer；
Step 5.2 compares t1The DC bus-bar voltage U of moment jth platform Wind turbinesjdc(t1) and discharging circuit action threshold value Udc_in
Size, if Ujdc(t1) it is more than Udc_in, then judge that jth platform Wind turbines discharging circuit turns on；Conversely, judge jth typhoon motor
Group discharging circuit does not turn on；
Step 6 improves the directly driven wind-powered field group of planes of K mean cluster algorithm partition using immune random k values and sensitive cluster centre；
The Wind turbines of all discharging circuit conductings in n platform direct-drive permanent-magnetism Wind turbines are divided into a group of planes by step 6.1；
Step 6.2, using the set end voltage value of each Wind turbines of fault moment to divide group's index, using immune random k values and quick
Feel cluster centre and improve K mean cluster algorithm, the m platform Wind turbines that remaining discharging circuit does not turn on are divided into a k group of planes.
2. the directly driven wind-powered field group of planes division methods according to claims 1, which is characterized in that in the step 6.2, exempt from
It is to carry out as follows that the random k values of epidemic disease and sensitive cluster centre, which improve K mean cluster algorithm,：
Step 6.2.1, the set end voltage value { U for the m platform Wind turbines for not turning on remaining discharging circuit1(t0),U2(t0),…,
Up(t0),…,Uq(t0),…,Um(t0) as sample data intersection, it is denoted as S={ x1,x2,…,xp,…,xq,…,xm}；Its
In, Up(t0) and Uq(t0) the set end voltage values of pth platform and q platform Wind turbines is represented respectively, take perunit value, and by Up(t0) and
Uq(t0) data object x is denoted as respectivelypAnd xq, p, q=1,2 ..., m, p ≠ q；
Step 6.2.2, data object x is calculatedpAnd xqEuclidean distance dist (xp,xq)；
Step 6.2.3, the average distance in the sample data intersection S between any two data object is obtained using formula (5)
MeanDist：
<mrow>
<mi>M</mi>
<mi>e</mi>
<mi>a</mi>
<mi>n</mi>
<mi>D</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<msubsup>
<mi>C</mi>
<mi>n</mi>
<mn>2</mn>
</msubsup>
</mfrac>
<mo>&times;</mo>
<munderover>
<mo>&Sigma;</mo>
<mtable>
<mtr>
<mtd>
<mrow>
<mi>p</mi>
<mo>,</mo>
<mi>q</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mi>p</mi>
<mo>&NotEqual;</mo>
<mi>q</mi>
</mrow>
</mtd>
</mtr>
</mtable>
<mi>m</mi>
</munderover>
<mi>d</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>p</mi>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mi>q</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>5</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (5),To appoint the number of combinations for taking 2 data objects from n data object；
Step 6.2.4, define：With data object xpCentered on, the region using average distance MeanDist as radius is known as data pair
As xpNeighborhood, the data object xpNeighborhood in the number of data object be known as data object xpBased on distance MeanDist
Density parameter；
Data object x is obtained using formula (6)pDensity parameter density (xp, MeanDist), so as to obtain m data object
Density parameter：
<mrow>
<mi>d</mi>
<mi>e</mi>
<mi>n</mi>
<mi>s</mi>
<mi>i</mi>
<mi>t</mi>
<mi>y</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>p</mi>
</msub>
<mo>,</mo>
<mi>M</mi>
<mi>e</mi>
<mi>a</mi>
<mi>n</mi>
<mi>D</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<munderover>
<mo>&Sigma;</mo>
<mtable>
<mtr>
<mtd>
<mrow>
<mi>q</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mi>q</mi>
<mo>&NotEqual;</mo>
<mi>p</mi>
</mrow>
</mtd>
</mtr>
</mtable>
<mi>m</mi>
</munderover>
<mi>u</mi>
<mrow>
<mo>(</mo>
<mi>M</mi>
<mi>e</mi>
<mi>a</mi>
<mi>n</mi>
<mi>D</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>-</mo>
<mi>d</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>(</mo>
<mrow>
<msub>
<mi>x</mi>
<mi>p</mi>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mi>q</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>6</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (6), u (MeanDist-dist (xp,xq)) represent a function, and have：
<mrow>
<mi>u</mi>
<mrow>
<mo>(</mo>
<mi>M</mi>
<mi>e</mi>
<mi>a</mi>
<mi>n</mi>
<mi>D</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>-</mo>
<mi>d</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>(</mo>
<mrow>
<msub>
<mi>x</mi>
<mi>p</mi>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mi>q</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mrow>
<mn>1</mn>
<mo>,</mo>
<mi>M</mi>
<mi>e</mi>
<mi>a</mi>
<mi>n</mi>
<mi>D</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mo>-</mo>
<mi>d</mi>
<mi>i</mi>
<mi>s</mi>
<mi>t</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>x</mi>
<mi>p</mi>
</msub>
<mo>,</mo>
<msub>
<mi>x</mi>
<mi>q</mi>
</msub>
<mo>)</mo>
</mrow>
<mo>&GreaterEqual;</mo>
<mn>0</mn>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<mn>0</mn>
<mo>,</mo>
<mi>e</mi>
<mi>l</mi>
<mi>s</mi>
<mi>e</mi>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>7</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 6.2.5, by the density parameter of the m data object, the density of the preceding M data object of density parameter maximum
Parameter is added in alternative point set D, D={ density (xp, MeanDist), p=1,2 ..., M }；
Step 6.2.6, cyclic variable r is defined, and initializes r=1；Definition cluster number is k, and initializes k=1；Definition is just
Maximum similarity is AMS between the average class that beginsr-1, and initialize AMSr-1；It defines cluster centre collection and is combined into Ar-1, and it is initialized as sky
Collection；Using the alternative point set D as the r-1 times alternative point set Dr-1；
Step 6.2.7, from the r-1 times alternative point set Dr-1In select density parameter maximum two data objects average conduct
The r times cluster centre crIt is put into cluster centre set Ar-1In, obtain the r times cluster centre set Ar；Meanwhile by selected two
Density parameter corresponding to a data object is from the r-1 times alternative point set Dr-1Middle deletion, so as to obtain the r times alternative point set Dr；
Step 6.2.8, from the r times alternative point set DrMiddle selection and the r times cluster centre set ArIn cluster centre distance it is farthest
Data object, as the r+1 times cluster centre cr+1It is put into the r times cluster centre set ArIn, it obtains in the r+1 times cluster
Heart set Ar+1；Meanwhile by the r+1 times cluster centre cr+1Corresponding density parameter is from the r times alternative point set DrMiddle deletion, from
And obtain the r+1 times alternative point set Dr+1；
Step 6.2.9, k+1 is assigned to k；
Step 6.2.10, remaining data object in the sample data intersection S is calculated and the r+1 times cluster centre collection respectively
Close Ar+1In each cluster centre Euclidean distance, and each data object is assigned to the nearest cluster centre institute of Euclidean distance
Class in, so as to obtain k class；
Step 6.2.11, the data object x of arbitrary i-th class in k class is obtained using formula (8)pTo the cluster centre c of the i-th classiIt
Between apart from average si：
<mrow>
<msub>
<mi>s</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<msub>
<mi>P</mi>
<mi>i</mi>
</msub>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>p</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>P</mi>
<mi>i</mi>
</msub>
</munderover>
<mo>|</mo>
<mo>|</mo>
<msub>
<mi>x</mi>
<mi>p</mi>
</msub>
<mo>-</mo>
<msub>
<mi>c</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
<mo>|</mo>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>8</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (8), PiFor the total number of data object in the i-th class；I=1,2 ..., k；
Step 6.2.12, the cluster centre c of the i-th class is obtained using formula (9)iWith the cluster centre c of jth classjThe distance between di,j：
di,j=dist (ci,cj) (9)
In formula (9), i, j=1,2 ..., k, i ≠ j；
Step 6.2.13, maximum similarity AMS between the r times average class of formula (10) acquisition is utilizedr：
<mrow>
<msub>
<mi>AMS</mi>
<mi>r</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>k</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>k</mi>
</munderover>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
<mo>{</mo>
<mfrac>
<mrow>
<msub>
<mi>s</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<msub>
<mi>s</mi>
<mi>j</mi>
</msub>
</mrow>
<msub>
<mi>d</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
</mfrac>
<mo>}</mo>
<mo>,</mo>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
<mo>,</mo>
<mn>2</mn>
<mo>,</mo>
<mn>...</mn>
<mo>,</mo>
<mi>k</mi>
<mo>,</mo>
<mi>i</mi>
<mo>&NotEqual;</mo>
<mi>j</mi>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>10</mn>
<mo>)</mo>
</mrow>
</mrow>
Step 6.2.14, AMS is judgedr＜ AMSr-1It is whether true, if so, then go to step 6.2.15；Otherwise, step is gone to
6.2.19；
Step 6.2.15, newer cluster centre c ' is obtained using formula (11)i：
<mrow>
<msubsup>
<mi>c</mi>
<mi>i</mi>
<mo>&prime;</mo>
</msubsup>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<msub>
<mi>P</mi>
<mi>i</mi>
</msub>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>p</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<msub>
<mi>P</mi>
<mi>i</mi>
</msub>
</munderover>
<msub>
<mi>x</mi>
<mi>p</mi>
</msub>
<mo>-</mo>
<mo>-</mo>
<mo>-</mo>
<mrow>
<mo>(</mo>
<mn>11</mn>
<mo>)</mo>
</mrow>
</mrow>
In formula (11), PiFor the total number of data object in the i-th class, xpFor the data object of arbitrary i-th class, i=1,2 ..., k；
Step 6.2.16, described the r+1 times alternative point set D is calculated respectivelyr+1In any one density parameter corresponding to data
The sum of Euclidean distance between object and all newer cluster centres, so as to obtain the data pair corresponding to all density parameters
As the set of the sum of the Euclidean distance respectively between all newer cluster centres, chosen from the set of the sum of Euclidean distance
Data object corresponding to maximum is as the r+2 times cluster centre cr+2And it is put into the r+1 times cluster centre set Ar+1In, from
And obtain the r+2 times cluster centre set Ar+2；
Step 6.2.17, by the r+2 times cluster centre cr+2Corresponding density parameter is from the r+1 times alternative point set Dr+1In delete
It removes, so as to obtain the r+2 times alternative point set Dr+2；
Step 6.2.18, after r+1 being assigned to r, 6.2.9 is gone to step；
Step 6.2.19, with AMSr-1Initial cluster center of the k corresponding cluster centre as kmeans clustering algorithms, it is right
The sample data intersection S carries out kmeans clustering algorithms, obtains a k group of planes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711394587.9A CN108054753B (en) | 2017-12-21 | 2017-12-21 | Direct-drive wind power plant cluster division method considering low-voltage ride through characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711394587.9A CN108054753B (en) | 2017-12-21 | 2017-12-21 | Direct-drive wind power plant cluster division method considering low-voltage ride through characteristics |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108054753A true CN108054753A (en) | 2018-05-18 |
| CN108054753B CN108054753B (en) | 2020-05-22 |
Family
ID=62131163
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711394587.9A Expired - Fee Related CN108054753B (en) | 2017-12-21 | 2017-12-21 | Direct-drive wind power plant cluster division method considering low-voltage ride through characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108054753B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110704995A (en) * | 2019-11-28 | 2020-01-17 | 电子科技大学中山学院 | Cable layout method and computer storage medium for multiple types of fans of multi-substation |
| CN111367254A (en) * | 2020-02-26 | 2020-07-03 | 哈尔滨工业大学 | Photovoltaic power station analytic single machine equivalence method, system and equipment |
| CN113346483A (en) * | 2021-05-20 | 2021-09-03 | 华中科技大学 | Low-voltage ride-through operation control method and system of power electronic transformer |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105243230A (en) * | 2015-11-02 | 2016-01-13 | 合肥工业大学 | Equivalent modeling method for direct-drive wind farm of considering off-load circuit |
-
2017
- 2017-12-21 CN CN201711394587.9A patent/CN108054753B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105243230A (en) * | 2015-11-02 | 2016-01-13 | 合肥工业大学 | Equivalent modeling method for direct-drive wind farm of considering off-load circuit |
Non-Patent Citations (1)
| Title |
|---|
| 潘险险等: "基于免疫离群数据和敏感初始中心的K-means算法的风电场机群划分", 《中国电机工程学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110704995A (en) * | 2019-11-28 | 2020-01-17 | 电子科技大学中山学院 | Cable layout method and computer storage medium for multiple types of fans of multi-substation |
| CN110704995B (en) * | 2019-11-28 | 2020-05-01 | 电子科技大学中山学院 | Cable layout method and computer storage medium for multiple types of fans of multi-substation |
| CN111367254A (en) * | 2020-02-26 | 2020-07-03 | 哈尔滨工业大学 | Photovoltaic power station analytic single machine equivalence method, system and equipment |
| CN111367254B (en) * | 2020-02-26 | 2021-05-07 | 哈尔滨工业大学 | Photovoltaic power station analytic single machine equivalence method, system and equipment |
| CN113346483A (en) * | 2021-05-20 | 2021-09-03 | 华中科技大学 | Low-voltage ride-through operation control method and system of power electronic transformer |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108054753B (en) | 2020-05-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zou et al. | A fuzzy clustering algorithm-based dynamic equivalent modeling method for wind farm with DFIG | |
| WO2013174145A1 (en) | Large-scale wind power grid-integration reactive voltage optimization method based on improved artificial fish swarm hybrid optimization algorithm | |
| CN102856917B (en) | Reactive power optimization method of power distribution network | |
| AU2018101070A4 (en) | Automatic voltage control method, device and system for wind farm | |
| CN103094920B (en) | A kind of equivalence method of direct-drive-type wind turbine generator wind power plant | |
| CN102684201B (en) | Voltage threshold probability-based reactive power optimizing method for grid containing wind power plant | |
| CN106684905B (en) | A kind of wind power plant Dynamic Equivalence considering wind-powered electricity generation uncertainty in traffic | |
| CN102611132B (en) | Method for adjusting parameters of additional frequency controller of double-fed variable-speed wind turbine generator | |
| CN110571794B (en) | Transient model equivalent calculation method suitable for doubly-fed wind power plant | |
| CN109063276A (en) | Wind power plant dynamic equivalent modeling method suitable for long time domain fluctuations in wind speed | |
| CN108054753A (en) | A kind of directly driven wind-powered field group of planes division methods of meter and low-voltage crossing characteristic | |
| CN103368186A (en) | Reactive optimization method of wind power system | |
| US20240162715A1 (en) | Method for group coordinated voltage control of photovoltaic inverters in low-voltage distribution network | |
| CN111027179B (en) | Equivalent modeling method for double-fed wind power plant considering auxiliary frequency modulation service | |
| CN114513004A (en) | New energy station equivalence method based on improved k-means algorithm and application | |
| CN106780119A (en) | Power distribution network regenerative resource based on many active managements is dissolved method | |
| CN105656055A (en) | Wind power plant clustered reactive power and voltage optimized control system and control method thereof | |
| CN109816261A (en) | Marine half direct drive wind power field equivalence method, system and device | |
| CN106786766B (en) | A method of the raising wind-powered electricity generation maximum grid connection capacity based on P2G technology | |
| CN110323779B (en) | Method and system for dynamically aggregating power of distributed power generation and energy storage device | |
| CN103956767A (en) | Wake-effect-considering wind power plant grid-connection stability analysis method | |
| CN109741207A (en) | A kind of Distribution system method for improving new energy and receiving ability | |
| CN116054218A (en) | Network-structured energy storage capacity optimization method, device, terminal and storage medium | |
| CN112952869B (en) | Method and system for expanding and planning AC-DC hybrid system considering wind power access | |
| CN103259269A (en) | Wind power plant reactive power optimization configuration method based on double-fed induction fan reactive power generating capacity |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20200522 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |