CN105225252B - Particle clouds motion Forecasting Methodology - Google Patents
Particle clouds motion Forecasting Methodology Download PDFInfo
- Publication number
- CN105225252B CN105225252B CN201510603236.9A CN201510603236A CN105225252B CN 105225252 B CN105225252 B CN 105225252B CN 201510603236 A CN201510603236 A CN 201510603236A CN 105225252 B CN105225252 B CN 105225252B
- Authority
- CN
- China
- Prior art keywords
- mrow
- point
- point set
- pixel
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 35
- 239000002245 particle Substances 0.000 title claims abstract description 27
- 238000005457 optimization Methods 0.000 claims abstract description 22
- 239000011159 matrix material Substances 0.000 claims description 51
- 230000002068 genetic effect Effects 0.000 claims description 9
- 230000008569 process Effects 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 108090000623 proteins and genes Proteins 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 210000000349 chromosome Anatomy 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000006073 displacement reaction Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000009499 grossing Methods 0.000 description 2
- 238000005286 illumination Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013213 extrapolation Methods 0.000 description 1
- 230000008303 genetic mechanism Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30192—Weather; Meteorology
Landscapes
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Economics (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of particle clouds motion Forecasting Methodology, by the pixel point set for first obtaining cloud cluster edge in sky image at different moments, the corresponding relation of the pixel at cloud cluster edge in sky image at different moments is solved by optimization algorithm based on predetermined optimization aim, position based on the corresponding relation and corresponding pixel, it is predicted for the subsequent motion position of cloud cluster image border.Thus, it is possible to carry out more accurate prediction for particle clouds motion, providing data for the prediction of photovoltaic generation power minute level supports.
Description
Technical field
The present invention relates to image procossing and photovoltaic power electric powder prediction, and in particular to a kind of particle clouds motion prediction side
Method, more particularly, to the particle clouds motion Forecasting Methodology under a kind of Optimum Theory framework.
Background technology
Photovoltaic generation belongs to fluctuation and intermittent power supply as wind-power electricity generation, because photovoltaic generating system is by illumination
The influence of the climatic factor such as intensity and environment temperature, the change of its power output have uncertainty, and the disturbance of power output will
It is possible to influence the stabilization of power network, therefore, it is necessary to strengthen the research of photovoltaic power generation power prediction, photovoltaic generating system is obtained ahead of time
Daily generation curve, so as to coordinate power system formulate generation schedule, reduce the Randomization of photovoltaic generation to power train
The influence of system.The prediction of earth's surface irradiation level is the primary link of photovoltaic generation power substep prediction, and its accuracy is to ensure that power is pre-
Survey the key of precision.But the earth's surface irradiation level under cloudy weather by cloud cluster life disappear with motion influenceed, its change sometimes in
Machine, it is quick, violent the features such as, had a strong impact on the precision of Classical forecast algorithm.Therefore, in order to improve earth's surface spoke under cloudy weather
The precision of prediction of illumination, cloud cluster that must be aerial to day carry out directly observation to obtain corresponding data, research cloud cluster displacement identification with
And the method for motion prediction.
The content of the invention
In view of this, the present invention provides a kind of particle clouds motion Forecasting Methodology, to be carried out according to sky image to particle clouds motion
More accurate prediction.
A kind of particle clouds motion Forecasting Methodology of the present invention includes:
The first sky image is obtained at the first moment, the second sky image, the second moment phase are obtained at the second moment
Lagged for first scheduled time for first moment;
Obtain the first point set for characterizing the first sky image medium cloud group image border and characterize second sky
Image medium cloud rolls into a ball the second point set of image border;
First point set is calculated to the distance matrix of second point set, the element D of the distance matrixi,jFor
I-th point of first point set arrives j-th point of distance of second point set;
To minimize the shift length sum of all pixels of cloud cluster image border point and minimize cloud cluster image border
The shift length variance of all pixels point is optimization aim, and transposed matrix is asked for based on optimization algorithm;
The transposed matrix is used for the corresponding relation for characterizing first point set and second point set, is patrolled for two-value
Collect matrix;The element E of the transposed matrixi,jBe worth for 1 when show i-th point of first point set and second point set
J-th point of corresponding, the element E of the transposed matrix closedi,jBe worth for 0 when show and institute of first point set at i-th point
J-th point of the second point set is stated without corresponding relation;
Obtained according to first point set, second point set and the transposed matrix and thirdly gathered, described the
Three point sets are used for the predicted position for characterizing the 3rd moment cloud cluster image border, and the 3rd moment is relative to second moment
Lagged for second scheduled time.
Preferably, obtaining the first point set of sign the first sky image medium cloud group image border includes:
First sky image binaryzation is obtained to the first sky image of binaryzation;
The first smoothed image of acquisition is smoothed to the first sky image of binaryzation;
For the first smoothed image, travel through in a predefined manner image edge pixels point by obtain edge pixel point coordinates and in terms of
Enter the first point set, until all pixels point be traversed or the first point set in point form closed figures, the edge pixel
Point is that gray value is 0 and surrounding at least has the pixel that a gray value is 1.
Preferably, obtaining the second point set of sign the second sky image medium cloud group image border includes:
Second sky image binaryzation is obtained to the second sky image of binaryzation;
The second smoothed image of acquisition is smoothed to the second sky image of binaryzation;
For the second smoothed image, travel through in a predefined manner image edge pixels point by obtain edge pixel point coordinates and in terms of
Enter the second point set, until all pixels point is traversed or the second point set forms closed figures, the edge pixel point is ash
Angle value is 0 and surrounding at least has the pixel that a gray value is 1.
Preferably, the optimization algorithm is genetic algorithm, and its object function is:
Wherein, E is the transposed matrix, Ei,jFor the element of the transposed matrix, D is the distance matrix, Di,jFor institute
The element of distance matrix is stated,For { Ei,j·Di,j| i=1,2 ..., m;J=1,2 ..., n in own
It is not the variance of 0 number being worth, α and β are weight coefficient, and m is the quantity of pixel in the first point set, and n is in the second point set
The quantity of pixel.
Preferably, it is described transposed matrix is asked for based on optimization algorithm to include:
If with Ei,j=1, then Ei+1,j+Ei,j+1+Ei+1,j+1=1, transposed matrix is asked for based on genetic algorithm for constraints,
Wherein, i+1=1 is made if i=m, j+1=1 is made if j=n.
It is preferably based on first point set, second point set and the transposed matrix and obtains and thirdly gathers
Including:
The corresponding relation of the pixel in first point set and the second point set is obtained according to transposed matrix;
The pixel in second point set is traveled through, calculates the position of prediction pixel point corresponding to obtaining, and by described in
Prediction pixel point, which is included in, thirdly to be gathered;
Wherein, if the pixel of the second point set corresponds only to the pixel of first point set, based on as follows
Formula calculates prediction pixel point position:
Wherein,For the pixel point coordinates in the second point set,For withPixel in corresponding first point set
Coordinate,For the prediction pixel point coordinates in the thirdly set, i, j, k are each pixel point coordinates in corresponding point set
In sequence number;Δt1For first scheduled time, Δ t2For second scheduled time;
If the pixel that the pixel of second point set corresponds at least two first point sets, based on as follows
Formula calculates prediction pixel point position:
Wherein, L be withThe quantity of the pixel of corresponding first point set,For withCorresponding L the first point sets
The coordinate of ith pixel point in pixel is closed, round () is round function.
Preferably, methods described also includes:
According to the 3rd sky image of the thirdly set generation binaryzation to characterize the cloud cluster edge of prediction.
By first obtaining the pixel point set at cloud cluster edge in sky image at different moments, based on predetermined optimization aim
The corresponding relation of the pixel at cloud cluster edge in sky image at different moments is solved by optimization algorithm, it is corresponding based on this
The position of relation and corresponding pixel, it is predicted for the subsequent motion position of cloud cluster image border.It is thus, it is possible to right
More accurate prediction is carried out in particle clouds motion, providing data for the prediction of photovoltaic generation power minute level supports.
Brief description of the drawings
By the description to the embodiment of the present invention referring to the drawings, above-mentioned and other purpose of the invention, feature and
Advantage will be apparent from, in the accompanying drawings:
Fig. 1 is the flow chart of the particle clouds motion Forecasting Methodology of the embodiment of the present invention;
Fig. 2 is the first sky image of the embodiment of the present invention and the schematic diagram of the second sky image;
Fig. 3 is the first smoothed image of the embodiment of the present invention and the schematic diagram of the second smoothed image;
Fig. 4 is the 3rd sky image obtained based on the embodiment of the present invention and the sky image wheel mutually actually measured in the same time
Wide comparison diagram.
Embodiment
Below based on embodiment, present invention is described, but the present invention is not restricted to these embodiments.Under
It is detailed to describe some specific detail sections in the literary detailed description to the present invention.Do not have for a person skilled in the art
The description of these detail sections can also understand the present invention completely.In order to avoid obscuring the essence of the present invention, known method, mistake
The not narration in detail of journey, flow, element and circuit.
In addition, it should be understood by one skilled in the art that provided herein accompanying drawing be provided to explanation purpose, and
What accompanying drawing was not necessarily drawn to scale.
Unless the context clearly requires otherwise, otherwise entire disclosure is similar with the " comprising " in claims, "comprising" etc.
Word should be construed to the implication included rather than exclusive or exhaustive implication;That is, it is containing for " including but is not limited to "
Justice.
In the description of the invention, it is to be understood that term " first ", " second " etc. are only used for describing purpose, without
It is understood that to indicate or implying relative importance.In addition, in the description of the invention, unless otherwise indicated, the implication of " multiple "
It is two or more.
In the present invention, the overall nonlinear motion of cloud cluster is decomposed into the linear movement for each point for forming cloud cluster set
Combination, by the corresponding relation of the Algorithm for Solving cloud cluster edge each point of optimization, based on the corresponding relation and existing fortune
Emotionally condition is moved and changed in situation predict cloud cluster next step.
Fig. 1 is the flow chart of the particle clouds motion Forecasting Methodology of the embodiment of the present invention.
As shown in figure 1, the particle clouds motion Forecasting Methodology includes:
Step 100, in the first moment t1The first sky image is obtained, in the second moment t2The second sky image is obtained, it is described
Second moment t2Relative to first moment t1Lag the first scheduled time Δ t1。
It should be understood that the sky image obtained at different moments is to be obtained based on identical position and parameter by ground facilities for observation
, be so possible to carry out subsequent treatment.
First sky image and the second sky image are gray level image.Preferably, when the second moment is relative to first
The lag time Δ t at quarter1It is limited within 10 minutes.
Step 200, obtain the first point set B for characterizing the first sky image medium cloud group image border1With sign institute
State the second point set B of the second sky image medium cloud group image border2。
Described first point can be obtained respectively by carrying out image procossing to the first sky image and the second sky image
Set B1With the second point set B2。
Specifically, processing of the step 200 to the first sky image can include:
Step 210, the first sky image by the first sky image binaryzation acquisition binaryzation.
Preferably, binary conversion treatment is carried out by the following method.
If image resolution ratio is M × N, gray scale value matrix is F, and image binaryzation parameter is T, the bianry image square after conversion
Battle array is F ', then
In the present embodiment, binaryzation parameter T=100 is set, by the sky image of binaryzation, cloud cluster part is black
(gray scale 0), and sky portion is white (gray scale 1).
Step 220, the first sky image to binaryzation, which are smoothed, obtains the first smoothed image.
Preferably, the smoothing processing can be carried out in the following way:
Opening operation in morphological image is first carried out to image with identical structural element, then carries out closed operation again, is tied
The dimension of constitutive element prime matrix determines that shape is generally circular according to sky image resolution ratio.In the present embodiment, setting structure member
Element is
First smoothed image is binary image, as shown in Figure 3.
It should be understood that binaryzation and smoothing processing can also be carried out using mode known to other skilled in the art.
Step 230, for the first smoothed image, travel through image edge pixels point in a predefined manner to obtain edge pixel point
Coordinate is simultaneously included in the first point set B1, until all pixels point is traversed or the first point set B1In point form closed figures, institute
It is that gray value is 0 and surrounding at least has the pixel that a gray value is 1 to state edge pixel point.
Specifically, since the matrix upper left corner of the first smoothed image, first in rows from left to right sequentially, then press and arrange from upper
To lower order, judged according to the value of each coordinate points on image, confirm its whether be cloud cluster edge pixel.
An if pixel (x0,y0) meet:AndThen should
Pixel is labeled as edge pixel point, is included in the first point set B1.That is, it is 0 and surrounding that edge pixel point, which is gray value,
The pixel that a gray value is 1 at least be present.
In the case of coordinate value exceeds image range, the pixel gray value at the coordinate can be ignored when calculating.
Preferably, (first marginal point on the left of cloud cluster the top) can be lighted from first labeled edge, with this
On the basis of edge pixel point, if its coordinate is (xa,ya), then press (xa-1,ya)、(xa-1,ya+1)、(xa,ya+1)、(xa+1,ya+1)、
(xa+1,ya)、(xa+1,ya-1)、(xa,ya-1)、(xa-1,ya-1) sequentially judge whether to meet successively:A, the point is marginal point;B, should
Pixel point coordinates is not counted into the first point set also.The pixel point coordinates is extracted if the condition of a, b two is met and enters the first point set
Close, and using the point as new datum mark, repeat the above steps, until all pixels point is traversed or the first point set is formed and closed
Close figure.
Similarly, processing of the step 200 to the second sky image can include:
Step 240, the second sky image by the second sky image binaryzation acquisition binaryzation.
Step 250, the second sky image to binaryzation, which are smoothed, obtains the second smoothed image.
Second smoothed image is as shown in Figure 3.
Step 260, for the second smoothed image, travel through image edge pixels point in a predefined manner to obtain edge pixel point
And it is included in the second point set B2, until all pixels point is traversed or the second point set B2In point form closed figures, the side
Edge pixel is that gray value is 0 and surrounding at least has the pixel that a gray value is 1.
Above-mentioned steps 210-230 and step 240-260 can be carried out simultaneously, can also be carried out in a predetermined order.
Thus, by step 210-230 and step 240-260 can obtain the first smoothed image medium cloud roll into a ball image border and
Second smoothed image medium cloud rolls into a ball all pixels point coordinates of image border, namely obtains the first point set B1With the second point set B2。
Step 300, calculate the first point set B1To the second point set B2Distance matrix D, the member of the distance matrix
Plain Di,jFor the first point set B1I-th point arrive the second point set B2J-th point of distance.
Specifically, initial pictures and displacement diagram are as cloud cluster edge pixel point coordinates set is respectively B1、B2, wherein B1Middle member
Plain number is m, B2Middle element number is n, then
Wherein
Then
Thus, the line number of Distance matrix D is m, columns n.
It should be understood that Distance matrix D is not limited to use Euclidean distance, other well known distance calculating side can also be used
Formula.
Step 400, with minimize the shift length sum of all pixels of cloud cluster image border point and minimize cloud cluster figure
As the shift length variance of all pixels point at edge is optimization aim, transposed matrix, institute's rheme are asked for based on optimization algorithm
Matrix is moved to be used to characterize the first point set B1With the second point set B2Corresponding relation.
The corresponding relation computation model of the embodiment of the present invention is based on following consideration and established, i.e. for cloud cluster edge pixel point
Transposed matrix E, because inter-picture temporal interval is extremely short, be not enough to allow cloud cluster to carry out complicated motion, therefore final mask institute
The transposed matrix tried to achieve should correspond to particle clouds motion process the most simple and quick, that is, require cloud cluster edge each point shift length sum
It is minimum.On the other hand, the motion of cloud cluster is influenceed by the wind field and cloud cluster self inertia in its residing region, in level time minute
Under yardstick, the possibility range of movement of cloud cluster is also restricted between image, thus is believed that the direction of wind field and size are smaller at this
The time and space in the range of be consistent substantially, now the motion of each point should keep one as far as possible on cloud cluster edge in image
Cause, that is, require that the variance of cloud cluster edge each point shift length is minimum.
Wherein, transposed matrix E dimension is identical with Distance matrix D, and wherein element is 0 or 1, and
I=1,2 ..., m;J=1,2 ..., n
Determine that optimization aim is based on above-mentioned consideration:A, minimize cloud cluster image border shift length a little it
With;B, minimize cloud cluster image border shift length variance a little.
Preferably for the different weight coefficient of above-mentioned two optimization aim, building object function in a linear fashion is:
Wherein, E is the transposed matrix, Ei,jFor the element of the transposed matrix, D is the distance matrix, Di,jFor institute
The element of distance matrix is stated,For { Ei,j·Di,j| i=1,2 ..., m;J=1,2 ..., n in own
It is not the variance of 0 number being worth, α and β are weight coefficient.In the present embodiment, α=1, β=10 are set.
In embodiments of the present invention, the optimization algorithm can be genetic algorithm.(Genetic Algorithm) is mould
Intend the computation model of the natural selection of Darwinian evolutionism and the biological evolution process of genetic mechanisms, be that one kind passes through mould
Intend the method for natural evolution process searches optimal solution.Genetic algorithm is a population from the possible potential disaggregation of the problem that represents
(population) start, and a population is then by the individual of the certain amount by gene (gene) coding
(individual) form.Each individual is actually the entity that chromosome (chromosome) carries feature.Chromosome conduct
The main carriers of inhereditary material, i.e., the set of multiple genes, its internal performance (i.e. genotype) is certain assortment of genes, and it is determined
The external presentation of the shape of individual, as dark hair is characterized in by controlling certain assortment of genes of this feature to determine in chromosome
Fixed.Therefore, needing to realize the mapping i.e. coding work from phenotype to genotype at the beginning.Due to copying gene code
Work very complicated, we are often simplified, such as binary coding, after primary population produces, according to the survival of the fittest and winning
The bad principle eliminated, develop by generation (generation) and produce the approximate solution become better and better, in every generation, according in Problem Areas
Fitness (fitness) size selection (selection) individual of body, and by means of the genetic operator of natural genetics
(genetic operators) is combined intersection (crossover) and variation (mutation), produces and represents new disaggregation
Population.This process will cause the same rear life of kind of images of a group of characters natural evolution to be adaptive to environment, the last reign of a dynasty than former generation for population
Optimum individual in population can be used as problem approximate optimal solution by decoding (decoding).Genetic algorithm can be in quantity
The solution for make it that optimization aim is optimal is selected in more possibility.
The transposed matrix E obtained based on optimization algorithm can characterize the first point set B1To the second point set B2Mapping,
That is, corresponding relation, each second point set B2In pixel point coordinatesThe first point set B of one or more can be corresponded to1
Pixel point coordinates
Preferably, constraints can also be added in the solution procedure of optimization algorithm, to cause solving result more
Close to the actual movement rule of cloud cluster.
Because particle clouds motion is a continuous deformation process, its edge wheel profile should keep continuous in motion process, it is impossible to
Situations such as in the presence of twisting and breaking is intersected, because cloud cluster edge wheel profile by respectively pressing permanent order on edge connects to obtain,
Therefore require that the relative position of each point on edge remains constant.Therefore, constraints can be arranged to:If Ei,j=1, then
Ei+1,j+Ei,j+1+Ei+1,j+1=1.Wherein, the 1st index is returned to if i+1 or j+1 exceeds index range.
Thus, based on this displacement relation, it is in a short time linearly moving premise based on each marginal point, can carries out
The prediction of cloud cluster movement.
Step 500, according to the first point set B1, the second point set B2Obtained thirdly with the transposed matrix E
Set B3, the thirdly set B3For characterizing the 3rd moment t3The predicted position of cloud cluster image border.The t3Moment is relative
In the t2Moment lags the second scheduled time Δ t2。
Specifically, step 500 can include:
Step 510, the first point set B obtained according to transposed matrix1With the second point set B2In pixel correspondence
Relation.
Step 520, traversal the second point set B2In pixel, calculate obtain corresponding to prediction pixel point position
Put, and the prediction pixel point is included in thirdly set B3。
Wherein, if the second point set B2Pixel correspond only to a first point set B1Pixel, then based on such as
Lower formula calculates prediction pixel point position:
Wherein,For the pixel point coordinates of the second point set,For withThe pixel of corresponding first point set is sat
Mark,For thirdly set B3In prediction pixel point coordinates, i, j, k are sequence number of each coordinate value in corresponding point set.
By taking rectangular coordinate system as an example, ifThen
If second point set B2Pixel correspond at least two first point set B1Pixel, then based on such as
Lower formula calculates prediction pixel point position:
Wherein, L be withCorresponding first point set B1Pixel quantity,For withCorresponding L first point
Gather the coordinate of ith pixel point in pixel, round () is round function.
Thus, the second point set B is being traveled through2In all pixels point after, can calculate obtain it is all corresponding to prediction pictures
Vegetarian refreshments.The thirdly set B being made up of prediction pixel point3Form the prediction for particle clouds motion image.
Preferably, the method for the embodiment of the present invention can also include step 600, that is, according to the thirdly set B3
The 3rd sky image of binaryzation is generated to characterize the cloud cluster edge of prediction.The binaryzation of generation the 3rd sky image (that is,
3rd moment t3For the prognostic chart picture of sky image) and in the 3rd moment t3The comparison diagram at the sky image edge actually measured
As shown in Figure 4.
Thus, by first obtaining the pixel point set at cloud cluster edge in sky image at different moments, based on predetermined excellent
Change target based on optimization algorithm to solve the corresponding relation of the pixel at cloud cluster edge in sky image at different moments, be based on
The position of the corresponding relation and corresponding pixel, it is predicted for cloud cluster image border subsequent motion position.Thus, may be used
To carry out more accurate prediction for particle clouds motion, provide data for the prediction of photovoltaic generation power minute level and support.
It is predicted according to the particle clouds motion Forecasting Methodology of the embodiment of the present invention, in step 100, can obtains such as Fig. 2 institutes
The first sky image and the second sky image shown
By step 200, the first sky image and the second sky image progress edge acquisition can be obtained following
First point set B1With the second point set B2:
B1For (odd-numbered line corresponds to the abscissa value of each point, and even number line corresponds to the ordinate value of each point):
B2For (odd-numbered line corresponds to the abscissa value of each point, and even number line corresponds to the ordinate value of each point):
By step 300-500, the thirdly set B of acquisition is predicted3As follows (odd-numbered line corresponds to the abscissa value of each point,
Even number line corresponds to the ordinate value of each point):
Comparison diagram based on Fig. 4 understands that the detection of the embodiment of the present invention has preferable accuracy.Sky in the present invention
Image particle clouds motion Forecasting Methodology is by the way that the overall nonlinear motion of cloud cluster to be decomposed into the group of the linear movement of each marginal point
Close, realize sky image particle clouds motion nonlinear prediction, compared with the existing forecast model based on linear extrapolation, Neng Gougeng
Change in shape of the cloud cluster in motion process is identified well.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the invention, for those skilled in the art
For, the present invention can have various changes and change.All any modifications made within spirit and principles of the present invention, it is equal
Replace, improve etc., it should be included in the scope of the protection.
Claims (7)
1. a kind of particle clouds motion Forecasting Methodology, including:
The first moment obtain the first sky image, the second moment obtain the second sky image, second moment relative to
First moment lagged for first scheduled time;
Obtain the first point set for characterizing the first sky image medium cloud group image border and characterize second sky image
Second point set of medium cloud group image border;
First point set is calculated to the distance matrix of second point set, the element D of the distance matrixi,jTo be described
I-th point of first point set arrives j-th point of distance of second point set;
To minimize the shift length sum of all pixels of cloud cluster image border point and minimize all of cloud cluster image border
The shift length variance of pixel is optimization aim, and transposed matrix is asked for based on optimization algorithm;
The transposed matrix is used for the corresponding relation for characterizing first point set and second point set, is two-valued function square
Battle array;The element E of the transposed matrixi,jBe worth for 1 when show i-th point of first point set and second point set
J-th point corresponding, the element E of the transposed matrixi,jBe worth for 0 when show i-th point of first point set and described the
J-th point of two point sets is without corresponding relation;
According to first point set, second point set and the transposed matrix obtain thirdly gather, it is described thirdly
Gather the predicted position for characterizing the 3rd moment cloud cluster image border, the 3rd moment lags relative to second moment
Second scheduled time.
2. particle clouds motion Forecasting Methodology according to claim 1, it is characterised in that obtain and characterize first sky image
First point set of medium cloud group image border includes:
First sky image binaryzation is obtained to the first sky image of binaryzation;
The first smoothed image of acquisition is smoothed to the first sky image of binaryzation;
For the first smoothed image, image edge pixels point is traveled through in a predefined manner to obtain edge pixel point coordinates and be included in the
One point set, until all pixels point be traversed or the first point set in point form closed figures, the edge pixel point is
Gray value is 0 and surrounding at least has the pixel that a gray value is 1.
3. particle clouds motion Forecasting Methodology according to claim 2, it is characterised in that obtain and characterize second sky image
Second point set of medium cloud group image border includes:
Second sky image binaryzation is obtained to the second sky image of binaryzation;
The second smoothed image of acquisition is smoothed to the second sky image of binaryzation;
For the second smoothed image, image edge pixels point is traveled through in a predefined manner to obtain edge pixel point coordinates and be included in the
Two point sets, until all pixels point is traversed or the second point set forms closed figures, the edge pixel point is gray value
The pixel that a gray value is 1 at least be present for 0 and surrounding.
4. particle clouds motion Forecasting Methodology according to claim 1, it is characterised in that the optimization algorithm is calculated for heredity
Method, its object function are:
<mrow>
<mi>min</mi>
<mo>=</mo>
<mi>&alpha;</mi>
<mrow>
<mo>(</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>m</mi>
</munderover>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>j</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msub>
<mi>E</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>D</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mo>+</mo>
<mi>&beta;</mi>
<mo>&CenterDot;</mo>
<mi>v</mi>
<mi>a</mi>
<mi>r</mi>
<mrow>
<mo>(</mo>
<mi>E</mi>
<mo>&CenterDot;</mo>
<mi>D</mi>
<msub>
<mo>|</mo>
<mrow>
<msub>
<mi>E</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>D</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
</mrow>
</msub>
<mo>&NotEqual;</mo>
<mn>0</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mrow>
Wherein, E is the transposed matrix, Ei,jFor the element of the transposed matrix, D is the distance matrix, Di,jFor it is described away from
From the element of matrix,For { Ei,j·Di,j| i=1,2 ..., m;J=1,2 ..., n in it is all be 0
The variance of the number of value, α and β are weight coefficient, and m is the quantity of pixel in the first point set, and n is pixel in the second point set
Quantity.
5. particle clouds motion Forecasting Methodology according to claim 4, it is characterised in that described that position is asked for based on optimization algorithm
Moving matrix includes:
If with Ei,j=1, then Ei+1,j+Ei,j+1+Ei+1,j+1=1, transposed matrix is asked for based on genetic algorithm for constraints, wherein,
I+1=1 is made if i=m, j+1=1 is made if j=n.
6. particle clouds motion Forecasting Methodology according to claim 1, it is characterised in that based on first point set, described
Second point set and the transposed matrix, which obtain thirdly set, to be included:
The corresponding relation of the pixel in first point set and the second point set is obtained according to transposed matrix;
Travel through the pixel in second point set, calculate the position of prediction pixel point corresponding to obtaining, and by the prediction
Pixel, which is included in, thirdly to be gathered;
Wherein, if the pixel of the second point set corresponds only to the pixel of first point set, based on equation below
Calculate prediction pixel point position:
<mrow>
<msubsup>
<mi>B</mi>
<mi>k</mi>
<mn>3</mn>
</msubsup>
<mo>=</mo>
<msubsup>
<mi>B</mi>
<mi>j</mi>
<mn>2</mn>
</msubsup>
<mo>+</mo>
<mrow>
<mo>(</mo>
<msubsup>
<mi>B</mi>
<mi>j</mi>
<mn>2</mn>
</msubsup>
<mo>-</mo>
<msubsup>
<mi>B</mi>
<mi>i</mi>
<mn>1</mn>
</msubsup>
<mo>)</mo>
</mrow>
<mfrac>
<mrow>
<msub>
<mi>&Delta;t</mi>
<mn>2</mn>
</msub>
</mrow>
<mrow>
<msub>
<mi>&Delta;t</mi>
<mn>1</mn>
</msub>
</mrow>
</mfrac>
<mo>;</mo>
</mrow>
Wherein,For the pixel point coordinates in the second point set,For withPixel in corresponding first point set is sat
Mark,For the prediction pixel point coordinates in the thirdly set, i, j, k are each pixel point coordinates in corresponding point set
Sequence number;Δt1For first scheduled time, Δ t2For second scheduled time;
If the pixel that the pixel of second point set corresponds at least two first point sets, based on equation below
Calculate prediction pixel point position:
<mrow>
<msubsup>
<mi>B</mi>
<mi>k</mi>
<mn>3</mn>
</msubsup>
<mo>=</mo>
<mi>r</mi>
<mi>o</mi>
<mi>u</mi>
<mi>n</mi>
<mi>d</mi>
<mo>&lsqb;</mo>
<msubsup>
<mi>B</mi>
<mi>j</mi>
<mn>2</mn>
</msubsup>
<mo>+</mo>
<mrow>
<mo>(</mo>
<mrow>
<msubsup>
<mi>B</mi>
<mi>j</mi>
<mn>2</mn>
</msubsup>
<mo>-</mo>
<mfrac>
<mrow>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>l</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>L</mi>
</munderover>
<msubsup>
<mi>B</mi>
<mrow>
<mi>i</mi>
<mo>,</mo>
<mi>l</mi>
</mrow>
<mn>1</mn>
</msubsup>
</mrow>
<mi>L</mi>
</mfrac>
</mrow>
<mo>)</mo>
</mrow>
<mfrac>
<mrow>
<msub>
<mi>&Delta;t</mi>
<mn>2</mn>
</msub>
</mrow>
<mrow>
<msub>
<mi>&Delta;t</mi>
<mn>1</mn>
</msub>
</mrow>
</mfrac>
<mo>&rsqb;</mo>
<mo>;</mo>
</mrow>
Wherein, L be withThe quantity of the pixel of corresponding first point set,For withCorresponding L the first point set pictures
The coordinate of ith pixel point in vegetarian refreshments, round () are round function.
7. particle clouds motion Forecasting Methodology according to claim 1, it is characterised in that methods described also includes:
According to the 3rd sky image of the thirdly set generation binaryzation to characterize the cloud cluster edge of prediction.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510603236.9A CN105225252B (en) | 2015-09-21 | 2015-09-21 | Particle clouds motion Forecasting Methodology |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510603236.9A CN105225252B (en) | 2015-09-21 | 2015-09-21 | Particle clouds motion Forecasting Methodology |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105225252A CN105225252A (en) | 2016-01-06 |
CN105225252B true CN105225252B (en) | 2018-03-30 |
Family
ID=54994201
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510603236.9A Active CN105225252B (en) | 2015-09-21 | 2015-09-21 | Particle clouds motion Forecasting Methodology |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105225252B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108846854B (en) * | 2018-05-07 | 2021-03-16 | 中国科学院声学研究所 | Vehicle tracking method based on motion prediction and multi-feature fusion |
CN110414420B (en) * | 2019-07-25 | 2021-08-06 | 中国人民解放军国防科技大学 | Mesoscale convection system identification and tracking method based on infrared cloud picture of stationary satellite |
CN111460196A (en) * | 2020-05-21 | 2020-07-28 | 南京大学 | Cloud layer change trend prediction method based on deep learning |
CN112419407B (en) * | 2020-11-04 | 2024-06-11 | 清华大学 | Cloud cluster displacement vector calculation method and device based on cloud cluster edge recognition |
CN112782700B (en) * | 2020-12-30 | 2024-04-16 | 北京墨迹风云科技股份有限公司 | Precipitation prediction method and device based on radar map |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6438253B1 (en) * | 1998-06-05 | 2002-08-20 | Thomson-Csf | Process for dynamic monitoring of changes to deformable media, and prediction of changes thereof |
CN103886616A (en) * | 2014-02-25 | 2014-06-25 | 国家电网公司 | Method for short-term prediction of photovoltaic power generation power based on whole-sky imaging data |
CN104778728A (en) * | 2015-05-04 | 2015-07-15 | 华北电力大学(保定) | Sky image cloud cluster movement velocity computing method based on phase correlation principle |
-
2015
- 2015-09-21 CN CN201510603236.9A patent/CN105225252B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6438253B1 (en) * | 1998-06-05 | 2002-08-20 | Thomson-Csf | Process for dynamic monitoring of changes to deformable media, and prediction of changes thereof |
CN103886616A (en) * | 2014-02-25 | 2014-06-25 | 国家电网公司 | Method for short-term prediction of photovoltaic power generation power based on whole-sky imaging data |
CN104778728A (en) * | 2015-05-04 | 2015-07-15 | 华北电力大学(保定) | Sky image cloud cluster movement velocity computing method based on phase correlation principle |
Also Published As
Publication number | Publication date |
---|---|
CN105225252A (en) | 2016-01-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105225252B (en) | Particle clouds motion Forecasting Methodology | |
Mensour et al. | A geographical information system-based multi-criteria method for the evaluation of solar farms locations: A case study in Souss-Massa area, southern Morocco | |
CN110738252A (en) | Space autocorrelation machine learning satellite precipitation data downscaling method and system | |
CN108364097B (en) | Based on the typhoon cloud system prediction technique for generating confrontation network | |
CN104200484B (en) | Distributed photovoltaic system ultra-short-term power output prediction method based on cloud cluster characteristic analysis | |
CN109493599A (en) | A kind of Short-time Traffic Flow Forecasting Methods based on production confrontation network | |
CN110991497B (en) | BSVC (binary sequence video coding) -method-based urban land utilization change simulation cellular automaton method | |
CN110163187A (en) | Remote road traffic sign detection recognition methods based on F-RCNN | |
CN108985965A (en) | A kind of photovoltaic power interval prediction method of combination neural network and parameter Estimation | |
CN109543721A (en) | A kind of solar irradiance ultra-short term forecasting procedure under fine with occasional clouds weather condition | |
CN103116865A (en) | Multidimensional collaborative power grid planning method | |
CN106779130A (en) | A kind of photovoltaic plant radiation Forecasting Methodology based on all-sky cloud atlas | |
CN106372749A (en) | Ultrashort-term photovoltaic power prediction method based on cloud change analysis | |
CN107194495A (en) | A kind of longitudinal Forecasting Methodology of photovoltaic power excavated based on historical data | |
CN105608271A (en) | Decomposition and optimization based short-term wind speed time series prediction method | |
CN108805345A (en) | A kind of crime space-time Risk Forecast Method based on depth convolutional neural networks model | |
CN113837499A (en) | Ultra-short-term wind power prediction method and system | |
CN106991666A (en) | A kind of disease geo-radar image recognition methods suitable for many size pictorial informations | |
CN104778728A (en) | Sky image cloud cluster movement velocity computing method based on phase correlation principle | |
CN114693064A (en) | Building group scheme generation performance evaluation method | |
CN112819237A (en) | Precipitation multi-scale fusion forecasting method and device | |
CN105930633A (en) | Method for forecasting urban heat island effect | |
Fan | Research on deep learning energy consumption prediction based on generating confrontation network | |
Feng et al. | An optimised cellular automata model based on adaptive genetic algorithm for urban growth simulation | |
CN105205560B (en) | Photovoltaic power supply power prediction method based on positive and negative error variable weights |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
TA01 | Transfer of patent application right | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20170306 Address after: Box No. 18 No. 619 071003 North China Electric Power University Hebei province Baoding Yonghua North Street Applicant after: North China Electric Power University (Baoding) Applicant after: Guo Wang Xinjiang power company Address before: Box No. 18 No. 619 071003 North China Electric Power University Hebei province Baoding Yonghua North Street Applicant before: North China Electric Power University (Baoding) |
|
GR01 | Patent grant | ||
GR01 | Patent grant |