CN107545118A - Improved parabolic type Lagrange coherent structure automatically extracts technology - Google Patents
Improved parabolic type Lagrange coherent structure automatically extracts technology Download PDFInfo
- Publication number
- CN107545118A CN107545118A CN201710827687.XA CN201710827687A CN107545118A CN 107545118 A CN107545118 A CN 107545118A CN 201710827687 A CN201710827687 A CN 201710827687A CN 107545118 A CN107545118 A CN 107545118A
- Authority
- CN
- China
- Prior art keywords
- flow field
- lcs
- singular
- point
- singular point
- 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.)
- Pending
Links
Abstract
Improved parabolic type Lagrange coherent structure automatically extracts technology, the fluid-dynamic analysis method of the invention based on foreign latest, a kind of technology that can be automatically analyzed to the structure of Sea Current big data of proposition.The purpose is to extract the most important feature in flow field-stream axle, and its changing rule is probed into, find new oceanographic phenomena, so as to be that the research such as physical oceanography, Marine Sciences provides help and inspiration.Parabola shaped LCS is the flow field main shaft that a kind of flow field change by a period of time calculates in the present invention, the mean state in the flow field for the previous period that it can be represented, it is also predicted that after flow field a period of time differentiation.
Description
Technical field
The present invention carries out big data analysis based on the correlative study of ocean surface remotely-sensed data combination hydrodynamics, belongs to fluid
Analysis of Topological Structure field.
Background technology
LCS (Lagrangian coherent structure) is that Lagrangian coherent structure is in recent years by Haller
A series of a kind of method of the data analysis fluidic structures of being formatted based on flow surface or in-house network proposed etc. scientists.At present
It is transoceanic near Agulhas rings through being applied to the multiple fields of ocean research, such as Gulf of Mexico oil spill accident diffusion analysis
Matter transportation phenomenon etc..Wherein parabolic type LCS is its chief component, and it mainly describes the stream axle i.e. mobility in fluid
Most strong position.The method for extracting parabolic type LCS before this is extremely complex, and its key step is as follows：First, fluid speed is initialized
Grid is spent, velocity field is established in flow surface, because the universal deficiency of source data resolution ratio to data, it is necessary to carry out at interpolation encryption
Reason, first have to be interpolated to regular grid data if source data is irregular grid data.Thereafter, according on each mesh point
Data calculate tensor field, and Cauchy's Green strain tensor matrix is drawn, if the tensor matrix that certain in flow field is put is by chance positive definite
Matrix, then it is flagged as singular point.By analyzing the topological structure of singular point, singular point can be divided into two
Class --- wedge shape and triangle.Finally, different types of singular point, selection are connected by the complicated integral operation in tensor field
It is parabolic type LCS to go out a minimum alternating chain of corresponding total shearing tensor sum stretch tensor, i.e. our the stream axles to be found.
The content of the invention
The present invention seeks to propose an improved automation generation parabolic type LCS algorithm, with suitable for a wide range of multiple
The structural analysis of miscellaneous practical flow field needs.It can be realized with full automation and have objectivity, substantially independent of being manually set
Threshold value and parameter.The technical scheme that the improved parabolic type Lagrange coherent structure of the present invention automatically extracts technology is included such as
Lower step（It can refer to Fig. 1）.
1. the extra large surface flow field data obtained according to satellite remote sensing, formation gridding flow field data network after interpolation refinement, one
As by legacy data resolution ratio improve 5-10 times, concrete condition depending on original resolution ratio, should not it is excessively sparse also need not be too
Refine and waste amount of calculation.
2. basis（1）In the velocity amplitude of each grid node storage calculate tensor matrix, the corresponding tensor of each node
Matrix.The value of tensor matrix represents the movement tendency of this mesh point position fluid（Including direction and intensity）；Utilize tensor
Cauchy's Green strain tensor of each mesh point of matrix computations, and obtain the characteristic value and feature of Cauchy's Green strain tensor matrix
Vector, further extract the variation characteristic of fluid.
3. with（2）Middle characteristic value local maximum is that starting point calculates hyperbolic-repulsion type LCS, while according to singular point in flow field
Tensor matrix go out singular point all in the flow field of region for positive definite matrix this conditional search.Thereafter, hyperbolic-repulsion is only chosen
Singular point around type hyperbola LCS is standby, and other singular points are given up to simplify calculating process.
4. all singular points of selection are divided into by two kinds-triangle singular point and wedge-shaped singular point according to the topological structure of singular point, will
Two kinds of singular points are used as seed point in two kinds of tensor fields respectively（Two kinds of tensor fields are respectively the category feature of tensor matrix two vector
Set）, stop integrating if the singular point of another species is run into integral process, it is respectively variety classes to form a head and the tail
The integration chain of singular point.
5. the integration chain of two kinds of singular points alternating compositions is found to both sides since the maximum singular point of characteristic value in region, and together
When ensure：When LCS stops when locally turnover is excessive；Given threshold makes parabolic type LCS will not be too adjacent and retained as far as possible special
Value indicative is larger；Removing intersection or LCS arranged side by side ensures its Local Steady.
Brief description of the drawings
The improved parabolic type Lagrange coherent structure automatic Extraction Algorithm flow charts of Fig. 1.
Fig. 2 triangle critical-point topological structures.
Fig. 3 wedge shape critical-point topological structures.
Embodiment
（1）Initial data is refined, initializes grid node：According to demand to original fluid surface（Such as ocean surface）Speed
Degrees of data is encrypted by interpolation algorithm, obtains the regular grid speed data of plane.If should be noted that during interpolation, data time is differentiated
Rate is too low（Such as oceanographic data has Zhou Pingjun or monthly average evidence）Encryption is equally answered, with the analysis knot after optimization
Fruit.
（2）Calculate the Cauchy's Green strain tensor matrix and its characteristic value characteristic vector of each mesh point after inserting：This part
Calculating had been described in the method for video memory, specially assume there is a particle on each node, allow its insert after velocity field
The middle motion a bit of time, calculate its movement tendency composition tensor matrix in all directions（Plane is two-dimensional matrix）, use
The tensor matrix dot product transposed matrix of itself, you can obtain secondary and mono- Ei Cauchy's Green strain tensor matrix（Plane is same
For two-dimensional matrix）, this is its characteristic vector characteristic value（Two groups）Also can ask.
（3）Calculate repulsion-hyperbolic type LCS：Take that composition plane larger in two characteristic values corresponding to each node
Data set, traversal whole region find all Local modulus maximas, are formed from these points in its corresponding characteristic vector
Integrated in vector field by a fixed step size, stop integration when running into edges of regions or can not readvance, obtain repulsion-hyperbolic
Type LCS.
（4）Choose singular point and classify：Cauchy's Green strain tensor matrix according to corresponding to defining singular point is positive definite matrix, therefore
Singular point not necessarily falls on existing grid node, it is necessary to by corresponding to tensor matrix in interpolation search region as positive definite matrix
Point.Thereafter, the distance for arriving nearest repulsion-hyperbolic type LCS a little is calculated, is given up if hypertelorism.We typically take
0.5 ° of distance, the singular point that can so ensure to choose are respectively positioned on the distinct region of Field Characteristics.Finally, to all selected
SINGULAR POINT processing, the foundation of classification is the topological structure of character pair vector vector field around it（It can refer to Fig. 2, Fig. 3）, will
Two kinds of singular point-triangles and wedge shape separately mark.
（5）Connect each singular point and integrate and obtain parabola shaped LCS：Formed using each singular point as seed point in characteristic vector
Vector field in integrated forward with a fixed step size（Triangle pair answers that group of characteristic vector of larger characteristic value）If run into another kind
Singular point or zone boundary then stop integrating.After all singular points have been traveled through, it is respectively difference from beginning to end only to choose in all integration chains
The line of species singular point is standby, and parabola shaped LCS will be combined by them.The maximum singular point of character pair value is found as parabolic
Line LCS starting point, go out parabola shaped LCS along its integration chain link, taken in way if integration chain fracture is run into nearest
Character pair value is maximum in several integration chains one continues to connect, until a complete parabola shaped LCS is produced.
Claims (4)
1. improved parabolic type Lagrange coherent structure automatically extracts technology, its principal character is as follows:
A. obtain and after all singular points, excluded in region apart from the repulsion-singular points of hyperbolic type LCS farther out, only retained remaining strange
Point;
B. to SINGULAR POINT processing, and retain with two class singular points be respectively beginning and end integration chain;
C. the head and the tail for integrating to obtain are the integration chain of different singular points, obtain parabola shaped LCS flow field structures figure.
2. the Time Continuous flow field structure analytical technology according to claim 1 based on objective Euler's coherent structure, it is special
Sign is, the step（1）In, the repulsion to having obtained --- hyperbolical LCS makes the buffering area that a radius is 0.5 °
（For ocean remote sensing data）, only retain to fall singular point in the buffer, remove other singular points, so as to ensure to extract parabolic
Speed-up computation process on the premise of linear LCS.
3. the Time Continuous flow field structure analytical technology according to claim 1 based on objective Euler's coherent structure, it is special
Sign is, the step（2）In, method is during to SINGULAR POINT：Characteristic vector where taking singular point on grid corner（In two groups
Any one）If the direction of four characteristic vectors is in side（I.e. maximum angle is less than 180 ° or minimum angle is more than 180 °）,
It is then wedge-shaped singular point, it is on the contrary then be triangle singular point;（Legend is shown in specification）In addition, step-length is unsuitable excessive when obtaining integration chain,
If it can be approximately considered when being integrated to another kind of near its singularity and run into the singular point, you can stop integration.
4. the Time Continuous flow field structure analytical technology according to claim 1 based on objective Euler's coherent structure, it is special
Sign is, the step（3）In, choose the appropriate parabola shaped LCS of starting point integration, general selection pair by observing flow field structure
The singular point for answering characteristic value maximum starts, and can so ensure the complete extraction of flow field structure;Integration is substantially to obtaining before
Integration chain spliced, a unique parabola shaped LCS is only retained in the range of agreement, this need splicing when select characteristic value compared with
Big integration chain is attached, and smooth treatment is carried out with five point Linear fitting algorithms when finally obtaining parabola shaped LCS.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710827687.XA CN107545118A (en) | 2017-09-14 | 2017-09-14 | Improved parabolic type Lagrange coherent structure automatically extracts technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710827687.XA CN107545118A (en) | 2017-09-14 | 2017-09-14 | Improved parabolic type Lagrange coherent structure automatically extracts technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107545118A true CN107545118A (en) | 2018-01-05 |
Family
ID=60963796
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710827687.XA Pending CN107545118A (en) | 2017-09-14 | 2017-09-14 | Improved parabolic type Lagrange coherent structure automatically extracts technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107545118A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108875117A (en) * | 2018-04-11 | 2018-11-23 | 中国海洋大学 | Multi-level flow field structure extraction method for ocean complex ocean current data based on Lagrange's pseudo-sequence structure |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101898644A (en) * | 2009-05-25 | 2010-12-01 | 丹麦技术大学 | System and method for designing airfoil |
| WO2014099204A1 (en) * | 2012-12-20 | 2014-06-26 | Exxonmobil Upstream Research Company | Method and system for geophysical modeling of subsurface volumes based on computed vectors |
| CN105913484A (en) * | 2016-04-05 | 2016-08-31 | 中国海洋大学 | Three-dimensional ocean current Lagrange coherent structure analysis algorithm |
| CN105975437A (en) * | 2016-03-29 | 2016-09-28 | 中国海洋大学 | Two-dimensional ocean current Lagrangian coherent structure analytical algorithm |
-
2017
- 2017-09-14 CN CN201710827687.XA patent/CN107545118A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101898644A (en) * | 2009-05-25 | 2010-12-01 | 丹麦技术大学 | System and method for designing airfoil |
| WO2014099204A1 (en) * | 2012-12-20 | 2014-06-26 | Exxonmobil Upstream Research Company | Method and system for geophysical modeling of subsurface volumes based on computed vectors |
| CN105975437A (en) * | 2016-03-29 | 2016-09-28 | 中国海洋大学 | Two-dimensional ocean current Lagrangian coherent structure analytical algorithm |
| CN105913484A (en) * | 2016-04-05 | 2016-08-31 | 中国海洋大学 | Three-dimensional ocean current Lagrange coherent structure analysis algorithm |
Non-Patent Citations (3)
| Title |
|---|
| MOHAMMAD FARAZMAND ET AL: "Computing Lagrangian coherent structures from their variational theory", 《PUBMED》 * |
| MOHAMMAD FARAZMAND ET AL: "How coherent are the vortices of two-dimensional turbulence?", 《ARXIV》 * |
| MOHAMMAD FARAZMAND ET AL: "Shearless transport barriers in unsteady two-dimensional flows and maps", 《ELSEVIER》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108875117A (en) * | 2018-04-11 | 2018-11-23 | 中国海洋大学 | Multi-level flow field structure extraction method for ocean complex ocean current data based on Lagrange's pseudo-sequence structure |
| CN108875117B (en) * | 2018-04-11 | 2022-05-13 | 中国海洋大学 | Method for extracting multilayer flow field structure based on Lagrangian order-fitting structure |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112070779B (en) | Remote sensing image road segmentation method based on convolutional neural network weak supervised learning | |
| CN110210551B (en) | Visual target tracking method based on adaptive subject sensitivity | |
| CN109493346B (en) | Stomach cancer pathological section image segmentation method and device based on multiple losses | |
| White | The basics of least cost analysis for archaeological applications | |
| CN112906830B (en) | Automatic generation method of ship optimal route based on AIS big data | |
| WO2022252398A1 (en) | Ship trajectory feature point extraction-based spatio-temporal dp method | |
| CN105701508A (en) | Global-local optimization model based on multistage convolution neural network and significant detection algorithm | |
| CN103389076B (en) | Submarine topography change detection and analysis method based on mesh reconstruction | |
| CN108121991A (en) | A kind of deep learning Ship Target Detection method based on the extraction of edge candidate region | |
| CN104680187B (en) | Naval vessel track Hierarchical Cluster Analysis based on buffering area similarity measurement | |
| CN104200507A (en) | Estimating method for normal vectors of points of three-dimensional point clouds | |
| CN103745191B (en) | Landform analysis based method for automatically identifying tablelands, ridges and loess hills in loess region | |
| CN108305266A (en) | Semantic image dividing method based on the study of condition random field graph structure | |
| CN111753677A (en) | Multi-angle remote sensing ship image target detection method based on characteristic pyramid structure | |
| CN110580388B (en) | Channel network extraction method based on crowd-sourced trajectory data | |
| Ai et al. | A simplification of ria coastline with geomorphologic characteristics preserved | |
| Dang et al. | Application of deep learning models to detect coastlines and shorelines | |
| CN104182771A (en) | Time series data graphics analysis method based on automatic coding technology with packet loss | |
| CN114092697A (en) | Building facade semantic segmentation method with attention fused with global and local depth features | |
| CN112257496A (en) | Deep learning-based power transmission channel surrounding environment classification method and system | |
| Bahmanikashkooli et al. | Application of particle swarm optimization algorithm for computing critical depth of horseshoe cross section tunnel | |
| CN107545118A (en) | Improved parabolic type Lagrange coherent structure automatically extracts technology | |
| CN116678432B (en) | Vehicle real-time path planning method based on edge service | |
| Oga et al. | River water quality estimation based on convolutional neural network | |
| CN110853058B (en) | High-resolution remote sensing image road extraction method based on visual saliency detection |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180105 |