CN104008305B - For ten million kilowatt of wind power base can power generating wind resource distribution method of estimation - Google Patents

Abstract

The invention discloses for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, mainly include：Assume that wind power base to be measured has Building N anemometer tower, obtain each anemometer tower T_iThe historical wind speed data of hourage are preset before (i=1,2 ..., N)；Beeline cluster is carried out to anemometer tower according to the difference of the anemometer tower wind speed average of front default hourage in each whole little time point, unknown point is found with all anemometer tower T_iThe geographic distance R of (i=1,2 ..., N)_iMinimum anemometer tower T_i(i=1,2 ..., N)；Determine anemometer tower T_iOther anemometer towers in the group of place, the wind speed and direction estimate that unknown point is obtained to the anemometer tower inverse distance-weighting method interpolation that chooses.Of the present invention for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, can overcome that prior art wind resource utilization is low, power supply reliability is poor and the defect such as Operation of Electric Systems stability difference, to realize that wind-resources utilization rate is high, power supply reliability is good and the advantage of Operation of Electric Systems good stability.

Description

For ten million kilowatt of wind power base can power generating wind resource distribution method of estimation

Technical field

The present invention relates to large-scale wind electricity base wind-resources analysis technical field, in particular it relates to it is used for ten million kilowatt of wind-powered electricity generation Base can power generating wind resource distribution method of estimation.

Background technology

Natural Resources in China distribution natural endowment determines that large-scale wind electricity base majority is located remotely from load center and locality and does not have The region that large-scale power supply is supported, large-scale wind power concentration is grid-connected to cause transmission line of electricity power fluctuation steady to power system security Fixed operation brings huge challenge.As Construction of Wind Power speed is generally faster than power grid construction speed, opened in wind electricity large-scale at a high speed Under the background that sends out, there is abandoning wind and rations the power supply problem in China's majority large-scale wind electricity base.

To large-scale wind electricity base can power generating wind resource distribution situation be analyzed assessment, contribute to rationally carrying out wind energy turbine set Planning, meanwhile, 1-2 can be shifted to an earlier date and planned and built supporting project of transmitting and converting electricity, be effectively improved wind-resources utilization rate, reduce and abandon wind limit Electricity, so that promote the healthy and sustainable development of new forms of energy.For ten million multikilowatt large-scale wind electricity base can power generating wind resource analysis ask Topic, has no related application.

Realize the present invention during, inventor find prior art at least exist wind-resources utilization rate low, power supply The defect such as poor reliability and Operation of Electric Systems stability difference.

Content of the invention

It is an object of the invention to, for the problems referred to above, propose the generated electricity wind-resources for ten million kilowatt of wind power base Distribution estimation method, to realize that wind-resources utilization rate is high, power supply reliability is good and the advantage of Operation of Electric Systems good stability.

For achieving the above object, the technical solution used in the present invention is：For ten million kilowatt of wind power base can power generating wind Resource distribution method of estimation, mainly includes：

A, the total Building N anemometer tower of hypothesis wind power base to be measured, obtain each anemometer tower T_iPreset before (i=1,2 ..., N) The historical wind speed data of hourage；

B, anemometer tower is carried out according to the difference of the anemometer tower wind speed average of front default hourage in each whole little time point most short Distance cluster, finds unknown point with all anemometer tower T_iThe geographic distance R of (i=1,2 ..., N)_iMinimum anemometer tower T_i(i= 1,2,...,N)；

C, determine anemometer tower T_iOther anemometer towers in the group of place, are inserted with inverse distance-weighting method to the anemometer tower that chooses It is worth to the wind speed and direction estimate of unknown point.

Further, in stepb, also include：

According to the latitude and longitude coordinates of the latitude and longitude coordinates and unknown point of each anemometer tower, unknown point and all anemometer tower T are calculated_i The geographic distance R of (i=1,2 ..., N)_i.

Further, the calculating unknown point and all anemometer tower T_iThe geographic distance R of (i=1,2 ..., N)_iOperation, Specifically include：

If point coordinates to be measured is (x_k,y_k), it is known that measuring point coordinate is (x_i,y_i), wherein i=1,2 ..., N, N be anemometer tower Number, then tested point to the geographic distance of i-th anemometer tower be

Further, the unknown point that finds is with all anemometer tower T_iThe geographic distance R of (i=1,2 ..., N)_iMinimum Anemometer tower T_iThe operation of (i=1,2 ..., N), specifically includes：

Seek minR_i, corresponding anemometer tower is tested point to the geographic distance R of i-th anemometer tower_iMinimum anemometer tower.

Further, in stepb, described in each whole little time point according to the anemometer tower wind speed average of front default hourage Difference carry out the operation of beeline cluster to anemometer tower, specifically include：

First, the distance of definition class and class is the minimum of a value of distance between all individualities in two classes；

Secondly, the distance matrix (m × m) of m element to be clustered is constructed, then in the off-diagonal element of this matrix, The object G of minimum of adjusting the distance_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_q}；

Again, by formula d_rk=min { d_pk,d_qk(k ≠ p, q) calculate new class G_rWith originally the distance between all kinds of, so New (m-1) rank distance matrix is obtained；

Finally, second step is returned to, is required until categorical measure meets, clustering algorithm terminates.

Further, the object G of the minimum of adjusting the distance_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_qOperation, Specifically include：

Distance according to the definition of beeline clustering procedure, class and class is the minimum of distance between all individualities in two classes Value, then class G_pWith class G_qThe distance between D_pqThen it is expressed as：

On this basis, the distance matrix (m × m) of m element to be clustered is constructed, then in the nondiagonal element of this matrix In element, the object G of minimum of adjusting the distance_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_q}.

Further, in stepb, described in each whole little time point according to the anemometer tower wind speed average of front default hourage Difference carry out the operation of beeline cluster to anemometer tower, specifically also include：

Air speed data correlation calculations method is：The wind series of two anemometer towers regard two stochastic variables X, Y as, then phase Relation number is shown below：

Further, in step c, the determination anemometer tower T_iThe operation of other anemometer towers in the group of place, specifically Including：

According to each cluster result, the group belonging to each anemometer tower is obtained, determine other anemometer towers in the group.

Further, in step c, the inverse distance-weighting method is specifically with the function of unknown point to known point distance As weighting function, the data of known point are weighted with the estimate for obtaining unknown point；

Weighting function W (r) has various ways, including：

W (r)=1/r²；

Wherein, r is the distance between unknown point and known point, and R is the radius of influence, represents that distance exceedes the known of the radius To the data of unknown point without reference to value, m is greater than 1 integer to point；Obviously, in this interpolation method, it is known that put from not Know a little nearer, then weight is bigger.

Further, in step c, the anemometer tower inverse distance-weighting method interpolation of described pair of selection obtains unknown point The operation of wind speed and direction estimate, specifically includes：

The rule of " distance is nearer, and resources characteristic similitude is higher " is taken full advantage of, using W (r)=1/r²Form enter Row modeling；Then, U, V both direction is that the interpolation wind vector in x, y direction is expressed as：

Various embodiments of the present invention for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, due to master Including：Assume that wind power base to be measured has Building N anemometer tower, obtain each anemometer tower T_iHour is preset before (i=1,2 ..., N) Several historical wind speed data；Anemometer tower is entered according to the difference of the anemometer tower wind speed average of front default hourage in each whole little time point Row beeline is clustered, and finds unknown point with all anemometer tower T_iThe geographic distance R of (i=1,2 ..., N)_iMinimum anemometer tower T_i(i=1,2 ..., N)；Determine anemometer tower T_iOther anemometer towers in the group of place, to the anemometer tower inverse distance-weighting that chooses Method interpolation obtains the wind speed and direction estimate of unknown point；Such that it is able to overcome prior art wind resource utilization low, power supply Poor reliability and the defect of Operation of Electric Systems stability difference, to realize that wind-resources utilization rate is high, power supply reliability is good and electric power The good advantage of system run all right.

Other features and advantages of the present invention will be illustrated in the following description, also, partly be become from specification Obtain it is clear that or being understood by implementing the present invention.

Below by drawings and Examples, technical scheme is described in further detail.

Description of the drawings

Accompanying drawing is used for providing a further understanding of the present invention, and constitutes a part for specification, the reality with the present invention Applying example is used for explaining the present invention together, is not construed as limiting the invention.In the accompanying drawings：

Fig. 1 be the present invention for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation flow process illustrate Figure.

Specific embodiment

The preferred embodiments of the present invention are illustrated below in conjunction with accompanying drawing, it will be appreciated that preferred reality described herein Apply example to be merely to illustrate and explain the present invention, be not intended to limit the present invention.

According to embodiments of the present invention, as shown in Figure 1, there is provided for the generated electricity wind-resources point of ten million kilowatt of wind power base Cloth method of estimation.

The present embodiment for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, mainly include following Step：

Step S10, it is assumed that the wind power base has Building N anemometer tower, obtains each anemometer tower T_i6 before (i=1,2 ..., N) The historical wind speed data of individual hour；

Step S20, is carried out to anemometer tower according to the difference of the anemometer tower wind speed average of first 6 hours in each whole little time point most short Distance cluster；

In step S20, the concrete grammar of the beeline cluster is as follows：

Beeline clustering method belongs to one kind of systemic clustering.The basic thought of systemic clustering is according to certain Similarity degree (or distance) between kind criterion calculation is of all categories, then to sample to be sorted constantly by two the most close classes A class is merged into, while recalculating the similarity degree (or distance) of new class and other classes, so circulation is carried out, until sample Till classification number reaches setting value.Due to the simple relation for only between consideration wind series originally studying a question, preliminary in order to realize Primary Stage Data packet target, without the need for using complicated clustering method, and consider that beeline clustering method physical significance is bright Really, the simple advantage of model, the method used below carry out Clustering.Equation Section3

Distance according to the definition of beeline clustering procedure, class and class is the minimum of distance between all individualities in two classes Value, then class G_pWith class G_qThe distance between D_pqThen it is expressed as：

On this basis, the distance matrix (m × m) of m element to be clustered is constructed, then in the nondiagonal element of this matrix In element, the object G of minimum of adjusting the distance_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_q, then press formula

d_rk=min { d_pk,d_qk}(k≠p,q) (2)；

Calculate new class G_rWith originally the distance between all kinds of, so available new (m-1) rank distance matrix；Similar Ground, then find minimum d in new distance matrix_ij, then corresponding class G_iAnd G_jMerger, and calculate new class and other classes away from From so going on, till total categorical measure reaches predetermined value.

That is, clustering method is：First, the distance for defining class and class is that between all individualities, distance is most in two classes Little value；Secondly, construct the distance matrix (m × m) of m element to be clustered, then in the off-diagonal element of this matrix, to away from From minimum object G_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_q}；Again, by formula d_rk=min { d_pk,d_qk}(k≠p, Q) new class G is calculated_rWith originally the distance between all kinds of, so available new (m-1) rank distance matrix；Finally, return to Second step, requires until categorical measure meets, clustering algorithm terminates.

Corresponding with knearest neighbour method, distance here is the difference (taking absolute value) of anemometer tower wind speed average.Additionally, due to wind Fast numerical values recited there may be larger fluctuation, here not prior designated packet quantity, but constantly carry out in assorting process Judge, until per group in wind speed average difference maximum be less than a certain threshold value till.

During concrete calculating, being primarily based on a pair of initiation sequences of index carries out first round packet, afterwards to the sequence in each group Second wheel packet is carried out as standard with index two, satisfactory group is finally given and is divided.

Due to the wind series of point to be estimated unknown, so being difficult to strictly to calculate the group of unknown point ownership.Here be given Hypothesis below：The anemometer tower wind-resources situation for thinking nearest apart from unknown point is closest with the wind-resources situation of unknown point, this One assume to typical flat country be applicable, but then need to use with caution under MODEL OVER COMPLEX TOPOGRAPHY.

Here, air speed data correlation calculations method is：The wind series of two anemometer towers regard as two stochastic variables X, Y, then coefficient correlation be shown below：

Step S30, it is assumed that the wind power base has Building N anemometer tower, calculates unknown point and all known anemometer tower T_i(i= 1,2 ..., N) geographic distance R_i；

In step s 30, unknown point can be calculated according to the latitude and longitude coordinates of each anemometer tower and the latitude and longitude coordinates of unknown point With known all anemometer tower geographic distance R_i.

According to the latitude and longitude coordinates of the latitude and longitude coordinates and unknown point of each anemometer tower, unknown point and known all survey wind are calculated Tower geographic distance R_i.

If point coordinates to be measured is (x_k,y_k), it is known that measuring point coordinate is (x_i,y_i), wherein i=1,2 ..., N, N be anemometer tower Number, then tested point to the geographic distance of i-th anemometer tower be

Step S40, finds R_iMinimum anemometer tower T_i(i=1,2 ..., N)：Seek minR_i, corresponding anemometer tower is Geographic distance R of the tested point to i-th anemometer tower_iMinimum anemometer tower；

Step S50, according to step S20 and the result of calculation of step S40, determines anemometer tower T_iOther surveys in the group of place Wind tower；

In step s 50, according to each cluster result, it is possible to obtain the group belonging to each anemometer tower, such that it is able to Determine other anemometer towers in the group；

Step S60, the wind speed and direction estimate that unknown point is obtained to the anemometer tower inverse distance-weighting method interpolation that chooses；

In step S60, the inverse distance weighted interpolation method is as follows：

Inverse distance weighted interpolation method is to the function of known point distance as weighting function, the number to known point using unknown point According to being weighted the estimate that obtains unknown point.Weighting function W (r) has various ways, for example, have：

W (r)=1/r²(6)；

Wherein, r is the distance between unknown point and known point, and R is the radius of influence, represents that distance exceedes the known of the radius To the data of unknown point without reference to value, m is greater than 1 integer to point.Obviously, in this interpolation method, it is known that put from not Know a little nearer, then weight is bigger.

Typically be modeled in the form of formula (6), take full advantage of here " distance nearer, resources characteristic similitude is got over The rule of height ".Then the interpolation wind vector of U, V both direction (i.e. x, y direction) is expressed as：

In sum, the various embodiments described above of the present invention for ten million kilowatt of wind power base can power generating wind resource distribution estimate Meter method, for ten million multikilowatt large-scale wind electricity base can power generating wind resource distribution estimated, supervised by known anemometer tower The wind-resources data of survey, carry out beeline using the Dynamic Packet method of the difference based on anemometer tower wind speed average and gather to anemometer tower Class, is then obtained the wind speed and direction estimate of unknown point, efficiently solves large-scale wind electricity base using inverse distance-weighting method interpolation The wind-resources monitoring blind area problem on ground, provides a kind of effective solution to the wind-resources distributional analysis of large-scale wind electricity base.

Finally it should be noted that：The preferred embodiments of the present invention are the foregoing is only, the present invention is not limited to, Although being described in detail to the present invention with reference to the foregoing embodiments, for a person skilled in the art, which still may be used To modify to the technical scheme described in foregoing embodiments, or equivalent is carried out to which part technical characteristic. All any modification, equivalent substitution and improvements that within the spirit and principles in the present invention, is made etc., should be included in the present invention's Within protection domain.

Claims (9)

1. be used for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, it is characterised in that mainly include：

A, the total Building N anemometer tower of hypothesis wind power base to be measured, obtain each anemometer tower T_iThe historical wind speed number of front default hourage According to；

B, beeline is carried out to anemometer tower according to the difference of the anemometer tower wind speed average of front default hourage in each whole little time point Cluster, finds unknown point with all anemometer tower T_iGeographic distance R_iMinimum anemometer tower T_i；

C, determine each anemometer tower T_iOther anemometer towers in the group of place, are inserted with inverse distance-weighting method to the anemometer tower that chooses It is worth to the wind speed and direction estimate of unknown point；

In stepb, described in each whole little time point according to the difference of the anemometer tower wind speed average of front default hourage to anemometer tower The operation of beeline cluster is carried out, is specifically included：

First, the distance of definition class and class is the minimum of a value of distance between all individualities in two classes；

Secondly, the distance matrix m × m of m element to be clustered is constructed, then in the off-diagonal element of this matrix, is adjusted the distance Minimum object G_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_q}；

Again, by formula d_rk=min { d_pk,d_qkCalculate new class G_rWith originally the distance between all kinds of, so available one is new M-1 rank distance matrixs；

Finally, second step is returned to, is required until categorical measure meets, clustering algorithm terminates.

2. according to claim 1 for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, which is special Levy and be, in stepb, also include：

According to the latitude and longitude coordinates of the latitude and longitude coordinates and unknown point of each anemometer tower, unknown point and all anemometer tower T are calculated_iGround Reason is apart from R_i.

3. according to claim 2 for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, which is special Levy and be, the calculating unknown point and all anemometer tower T_iGeographic distance R_iOperation, specifically include：

If point coordinates to be measured is (x_k,y_k), it is known that measuring point coordinate is (x_i,y_i), wherein i=1,2 ..., N, N be anemometer tower number, Then tested point to the geographic distance of i-th anemometer tower is：

$R_i=\sqrt{{(x_k-x_i)}^2+{(y_k-y_i)}^2}.$

4. according to claim 3 for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, which is special Levy and be, the unknown point that finds is with all anemometer tower T_iGeographic distance R_iMinimum anemometer tower T_iOperation, wherein i=1, 2 ..., N, specifically include：

Seek min R_i, corresponding anemometer tower is tested point to the geographic distance R of i-th anemometer tower_iMinimum anemometer tower.

5. according to claim 1 for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, which is special Levy and be, the object G of the minimum of adjusting the distance_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_qOperation, specifically include：

Distance according to the definition of beeline clustering procedure, class and class is the minimum of a value of distance between all individualities in two classes, that Class G_pWith class G_qThe distance between D_pqThen it is expressed as：

$D_{pq}=\underset{x_i\∈G_p,x_j\∈G_q}{min}{d}_{ij};$

On this basis, the distance matrix m × m of m element to be clustered is constructed, then in the off-diagonal element of this matrix, The object G of minimum of adjusting the distance_pAnd G_qMerger is carried out, new class G is obtained_r={ G_p,G_q}.

6. according to claim 1 for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, which is special Levy and be, in stepb, described in each whole little time point according to the difference of the anemometer tower wind speed average of front default hourage to surveying wind Tower carries out the operation of beeline cluster, specifically also includes：

Air speed data correlation calculations method is：The wind series of two anemometer towers regard two stochastic variables X, Y, then phase relation as Number is shown below：

$r_{x,y}=\frac{\frac{1}{n}\underset{i=1}{\overset{n}{\Σ}}(x_i-{\mathrm{\μ}}_x)(y_i-{\mathrm{\μ}}_y)}{{\mathrm{\σ}}_x{\mathrm{\σ}}_y}.$

7. according to any one of claim 1-4 for ten million kilowatt of wind power base can power generating wind resource distribution estimate Method, it is characterised in that in step c, the determination anemometer tower T_iThe operation of other anemometer towers in the group of place, concrete bag Include：

According to each cluster result, the group belonging to each anemometer tower is obtained, determine other anemometer towers in the group.

8. according to any one of claim 1-4 for ten million kilowatt of wind power base can power generating wind resource distribution estimate Method, it is characterised in that in step c, the inverse distance-weighting method is specifically with the function of unknown point to known point distance As weighting function, the data of known point are weighted with the estimate for obtaining unknown point；

Weighting function W (r) has various ways, including：

$W\left(r\right)=\left\{\begin{array}{cc}{\[\frac{R^2-r^2}{R^2+r^2}\]}^m& r\≤R\\ 0& r>R\end{array}\right.;$

W (r)=1/r²；

Wherein, r is the distance between unknown point and known point, and R is the radius of influence, represents that distance exceedes the known point pair of the radius Without reference to value, m is greater than 1 integer to the data of unknown point；Obviously, in this interpolation method, it is known that put from unknown point Nearer, then weight is bigger.

9. according to claim 8 for ten million kilowatt of wind power base can power generating wind resource distribution method of estimation, which is special Levy and be, in step c, the anemometer tower inverse distance-weighting method interpolation of described pair of selection obtains the wind speed and direction of unknown point and estimates The operation of evaluation, specifically includes：

The rule of " distance is nearer, and resources characteristic similitude is higher " is taken full advantage of, using W (r)=1/r²Form built Mould；Then, U, V both direction is that the interpolation wind vector in x, y direction is expressed as：

$U=\frac{\underset{i=1}{\overset{n}{\Σ}}\left(W\right(r_i\left)U_i\right)}{\underset{i=1}{\overset{n}{\Σ}}W\left(r_i\right)};$

$V=\frac{\underset{i=1}{\overset{n}{\Σ}}\left(W\right(r_i\left)V_i\right)}{\underset{i=1}{\overset{n}{\Σ}}W\left(r_i\right)}.$