forked from johannesgerer/jburkardt-f
-
Notifications
You must be signed in to change notification settings - Fork 2
/
pod_basis_flow.html
576 lines (536 loc) · 18.4 KB
/
pod_basis_flow.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
<html>
<head>
<title>
POD_BASIS_FLOW - PDE Model Reduction
</title>
</head>
<body bgcolor="#EEEEEE" link="#CC0000" alink="#FF3300" vlink="#000055">
<h1 align = "center">
POD_BASIS_FLOW <br> PDE Model Reduction
</h1>
<hr>
<p>
<b>POD_BASIS_FLOW</b>
is a FORTRAN90 program which
applies the principal
orthogonal direction (POD) analysis to a set of solutions of a
PDE that models two dimensional time dependent fluid flow.
</p>
<p>
This procedure, originally devised by Karl Pearson, has arisen
repeatedly in a variety of fields, and hence is known under
various names, including:
<ul>
<li>
the Hotelling transform;
</li>
<li>
the discrete Karhunen-Loeve transform (KLT)
</li>
<li>
Principal Component Analysis (PCA)
</li>
<li>
Principal Orthogonal Direction (POD)
</li>
<li>
Proper Orthogonal Decomposition (POD)
</li>
<li>
Singular Value Decomposition (SVD)
</li>
</ul>
</p>
<p>
<i>
You almost certainly should rather use the simpler program
called
<a href = "../../f_src/svd_basis/svd_basis.html">SVD_BASIS</a>,
which does not assume that the data comes from a particular
fluid flow problem!
</i>
</p>
<p>
The purpose of the algorithm is to extract the dominant modes
of behavior of the system; these modes could then be used, for
instance, in a reduced order model of the physical system.
</p>
<p>
For the computations considered here, a partial differential
equation (PDE) has been defined, specifying the time-dependent
flow of a fluid through a region. The PDE specification includes
a parameter <b>alpha</b> whose value strongly affects the behavior of
the flow. The steady state solution X0 is computed for a particular
value of <b>alpha</b>. Then the time-dependent problem is solved over a
fixed time interval, with <b>alpha</b> varying from time to time.
A set of several hundred solutions X(T(I),<b>alpha</b>(I)) are saved.
</p>
<p>
The need is to try to extract from this solution data the
typical modes of behavior of the solution, that is, a small set
of orthogonal vectors <b>V</b> such that "most" of the solution
vectors <b>X</b> can be well represented by a linear combination
of elements of <b>V</b>. Such a set of modes may then be used as
a finite element basis that is highly tuned to the physics of the
problem, so that a very small set of basis functions can be used
to closely approximate the behavior of the solution over a range
of values of <b>alpha</b>.
</p>
<p>
The method of extracting information from the solution data
begins by constructing an <b>M</b> by <b>N</b> matrix <b>A</b>, each
of whose columns is one of the solution vectors <b>X</b>. Thus,
<blockquote><b>
A = [ X1 | X2 | ... | XN ]
</b></blockquote>
</p>
<p>
Then the singular value decomposition of <b>A</b>:
<blockquote><b>
A = U * S * V'
</b></blockquote>
is determined using the DGESVD routine from the linear algebra package
<a href = "../lapack/lapack.html">LAPACK</a>.
A subset of the columns of the orthogonal <b>M</b> by <b>M</b>
matrix <b>U</b>, associated with the largest singular values <b>S</b>,
is chosen to form the POD basis.
</p>
<p>
Because the data comes from a finite element computation, and
the results may be used as a new reduced basis, it may be
desirable to carry out mass matrix preconditioning of the data,
so that output POD vectors are orthogonal
in the L2 inner product (integration of the product of the finite
element functions over the domain).
</p>
<p>
The current version of the program assumes that a steady state
solution <b>SS</b> of the PDE is known, and that a multiple
of SS is to be subtracted from each solution vector before processing.
</p>
<p>
<b>FILES</b>: the program assumes the existence of the following files:
(the actual names of the files are specified by the user at run time.
The names used here are just suggestions.)
<ul>
<li>
<i>xy.txt</i>, contains the (x,y) coordinates of each node, with
one pair of coordinates per line of the file;
</li>
<li>
<i>ss.txt</i>, contains the (u,v) values at each node for the
steady state solution, with one pair of values per line of the file;
</li>
<li>
<i>uv01.txt</i>, <i>uv02.txt</i>, ..., contains the (u,v) values
at each node for solution 1, 2, and so on;
</li>
<li>
<i>element.txt</i>, contains the indices of the six nodes that
make up each element, with one set of six indices per line of
the file <i>(only needed if mass matrix
preconditioning is used)</i>;
</li>
</ul>
</p>
<p>
<b>INPUT</b>: at run time, the user specifies:
<ul>
<li>
<i>run_type</i> describes how we subtract off the steady state,
whether we drop some data, and other options. The current
values range from 1 to 8. The most common value is 6, used
with the TCELL data:
<ol>
<li>
no steady state file, no preprocessing;
</li>
<li>
no steady state file, no preprocessing;
</li>
<li>
subtract 1/3 SS from solution 1, 5/3 SS from solutions
2 to 201, and 1/3 SS from solutions 202 through 401.
</li>
<li>
subtract 1/3 SS from solution 1, 5/3 SS from solutions
2 to 201, and 1/3 SS from solutions 202 through 401,
and drop the even-numbered data.
</li>
<li>
subtract 1/3 SS from solution 1, 5/3 SS from solutions
2 to 201, and 1/3 SS from solutions 202 through 401,
and skip half the data and normalize it.
</li>
<li>
subtract 5/3 SS from solutions
1 to 250, and 1/3 SS from solutions 251 through 500, do not
normalize.
</li>
<li>
subtract 5/3 SS from solutions
1 to 250, and 1/3 SS from solutions 251 through 500,
normalize the data.
</li>
<li>
subtract 5/3 SS from solutions
1 to 250, and 1/3 SS from solutions 251 through 500, then
drop the odd-numbered data, do not
normalize.
</li>
</ol>
</li>
<li>
<i>basis_num</i>, the requested basis size.
</li>
<li>
<i>xy_file</i>, the name of the xy file.
</li>
<li>
<i>steady_file</i>, the name of the steady state solution file
(only if <b>run_type</b> is 1 or 2).
</li>
<li>
<i>uv0_file</i>, the name of the first solution file (the program
will assume all the files are numbered consecutively).
</li>
<li>
<i>element_file</i>, the name of the element file, if mass matrix
preconditioning is desired, or else "none".
</li>
</ul>
</p>
<p>
<b>OUTPUT</b>: the program computes <b>basis_num</b> basis vectors.
The first vector is written to the file <i>pod_001.txt</i>; again,
the output vectors are written with two values per line, since
this represents the two components of velocity at a particular
node.
</p>
<h3 align = "center">
Licensing:
</h3>
<p>
The computer code and data files described and made available on this web page
are distributed under
<a href = "../../txt/gnu_lgpl.txt">the GNU LGPL license.</a>
</p>
<h3 align = "center">
Related Data and Programs:
</h3>
<p>
<a href = "../../m_src/brain_sensor_pod/brain_sensor_pod.html">
BRAIN_SENSOR_POD</a>,
a MATLAB program which
applies the method of Proper Orthogonal Decomposition
to seek underlying patterns in sets of 40 sensor readings of
brain activity.
</p>
<p>
<a href = "../../f_src/cvt_basis_flow/cvt_basis_flow.html">
CVT_BASIS_FLOW</a>,
a FORTRAN90 program which
is similar to POD_BASIS_FLOW but uses CVT techniques to
do the analysis of the data.
</p>
<p>
<a href = "../../f_src/lapack_examples/lapack_examples.html">
LAPACK_EXAMPLES</a>,
a FORTRAN90 program which
demonstrates the use of the LAPACK linear algebra library.
</p>
<p>
<a href = "../../f_src/svd_basis/svd_basis.html">
SVD_BASIS</a>,
a FORTRAN90 program which
is a simpler version of this program; it does not assume that
the underlying data represents solutions of a fluid flow problem.
</p>
<h3 align = "center">
Reference:
</h3>
<p>
<ol>
<li>
Edward Anderson, Zhaojun Bai, Christian Bischof, Susan Blackford,
James Demmel, Jack Dongarra, Jeremy Du Croz, Anne Greenbaum,
Sven Hammarling, Alan McKenney, Danny Sorensen,<br>
LAPACK User's Guide,<br>
Third Edition,<br>
SIAM, 1999,<br>
LC: QA76.73.F25L36
</li>
<li>
John Burkardt, Max Gunzburger, Hyung-Chun Lee,<br>
Centroidal Voronoi Tessellation-Based Reduced-Order
Modelling of Complex Systems,<br>
SIAM Journal on Scientific Computing,<br>
Volume 28, Number 2, 2006, pages 459-484.
</li>
<li>
Gal Berkooz, Philip Holmes, John Lumley,<br>
The proper orthogonal decomposition in the analysis
of turbulent flows,<br>
Annual Review of Fluid Mechanics,<br>
Volume 25, 1993, pages 539-575.
</li>
<li>
Lawrence Sirovitch,<br>
Turbulence and the dynamics of coherent structures, Parts I-III,<br>
Quarterly of Applied Mathematics,<br>
Volume XLV, Number 3, 1987, pages 561-590.
</li>
</ol>
</p>
<h3 align = "center">
Source Code:
</h3>
<p>
<ul>
<li>
<a href = "pod_basis_flow.f90">pod_basis_flow.f90</a>, the source code.
</li>
<li>
<a href = "pod_basis_flow.sh">pod_basis_flow.sh</a>,
commands to compile and load the source code.
</li>
</ul>
</p>
<h3 align = "center">
Examples and Tests:
</h3>
<p>
PDE solution datasets you may copy include:
<ul>
<li>
<a href = "../../datasets/cavity_flow/cavity_flow.html">
CAVITY_FLOW</a>, the driven cavity;
</li>
<li>
<a href = "../../datasets/inout_flow/inout_flow.html">
INOUT_FLOW</a>, flow in and out of a chamber;
</li>
<li>
<a href = "../../datasets/tcell_flow/tcell_flow.html">
TCELL_FLOW</a>, flow through a T-cell;
</li>
</ul>
</p>
<p>
Once the reduced basis set is computed, it is possible to
set up new finite element problems in which the reduced basis
set is used as the finite element basis. To see an example
of such a calculation, refer to
<ul>
<li>
<a href = "../../f77_src/tcell/tcell.html">
TCELL</a>, the program which generated the TCELL data;
</li>
<li>
<a href = "../../f77_src/tcell_mass/tcell_mass.html">
TCELL_MASS</a>, used to calculate the mass matrix;
</li>
<li>
<a href = "../../m_src/tcell_rom/tcell_rom.html">
TCELL_ROM</a>, the TCELL reduced order modeling program;
</li>
</ul>
</p>
<p>
This program has been run with a number of different datasets,
and with various requirements as to normalization and so on.
The purpose of most of the runs is to find a generator set of
given size. The input and output of each run is stored in
a separate subdirectory.
</p>
<p>
Each run of the code is stored in a separate subdirectory.
Available runs include:
<ul>
<li>
<a href = "run_01/run_01.html">run_01</a>,
getting 16 POD vectors from the CAVITY_FLOW data.
</li>
<li>
<a href = "run_02/run_02.html">run_02</a>,
getting 16 POD vectors from the INOUT_FLOW data.
</li>
<li>
<a href = "run_03/run_03.html">run_03</a>,
getting 16 POD vectors from the TCELL_FLOW data.
</li>
<li>
<a href = "run_04/run_04.html">run_04</a>,
getting 16 POD vectors from the CAVITY_FLOW data,
with mass matrix preconditioning.
</li>
<li>
<a href = "run_05/run_05.html">run_05</a>,
getting 16 POD vectors from the INOUT_FLOW data,
with mass matrix preconditioning.
</li>
<li>
<a href = "run_06/run_06.html">run_06</a>,
getting 16 POD vectors from the TCELL_FLOW data,
with mass matrix preconditioning.
</li>
<li>
<a href = "run_07/run_07.html">run_07</a>,
getting 16 POD vectors from the CAVITY_FLOW data,
with mass matrix preconditioning. We drop the odd
numbered data vectors.
</li>
<li>
<a href = "run_08/run_08.html">run_08</a>,
getting 16 POD vectors from the INOUT_FLOW data,
with mass matrix preconditioning. We drop the odd
numbered data vectors.
</li>
<li>
<a href = "run_09/run_09.html">run_09</a>,
getting 16 POD vectors from the TCELL_FLOW data,
with mass matrix preconditioning. We drop the odd
numbered data vectors.
</li>
<li>
<a href = "run_10/run_10.html">run_10</a>,
getting 16 POD vectors from the INOUT_FLOW2 data.
</li>
<li>
<a href = "run_11/run_11.html">run_11</a>,
getting 16 POD vectors from the INOUT_FLOW2 data,
with mass matrix preconditioning.
</li>
</ul>
</p>
<h3 align = "center">
List of Routines:
</h3>
<p>
<ul>
<li>
<b>MAIN</b> is the main program for POD_BASIS_FLOW.
</li>
<li>
<b>BANDWIDTH_DETERMINE</b> computes the lower bandwidth of a finite element matrix.
</li>
<li>
<b>BASIS_WRITE</b> writes a basis vector to a file.
</li>
<li>
<b>CH_CAP</b> capitalizes a single character.
</li>
<li>
<b>CH_EQI</b> is a case insensitive comparison of two characters for equality.
</li>
<li>
<b>CH_IS_DIGIT</b> is TRUE if a character is a decimal digit.
</li>
<li>
<b>CH_TO_DIGIT</b> returns the integer value of a base 10 digit.
</li>
<li>
<b>DATA_IVEC_READ</b> reads an dataset of integer vectors stored in a file.
</li>
<li>
<b>DATA_D2_READ</b> reads pairs of double precision numbers stored in a file.
</li>
<li>
<b>DATA_SIZE</b> counts the size of a data set stored in a file.
</li>
<li>
<b>DBLT_CHECK</b> checks the dimensions of a banded lower triangular matrix.
</li>
<li>
<b>DBLT_PRINT</b> prints a band lower triangular matrix.
</li>
<li>
<b>DBLT_PRINT_SOME</b> prints some of a band lower triangular matrix.
</li>
<li>
<b>DIGIT_INC</b> increments a decimal digit.
</li>
<li>
<b>DIGIT_TO_CH</b> returns the character representation of a decimal digit.
</li>
<li>
<b>DPBL_CHECK</b> checks the dimensions of a positive definite symmetric band matrix.
</li>
<li>
<b>DPBL_PRINT</b> prints a symmetric banded matrix.
</li>
<li>
<b>DPBL_PRINT_SOME</b> prints some of a symmetric banded matrix.
</li>
<li>
<b>FILE_EXIST</b> reports whether a file exists.
</li>
<li>
<b>FILE_NAME_INC</b> generates the next filename in a series.
</li>
<li>
<b>GET_UNIT</b> returns a free FORTRAN unit number.
</li>
<li>
<b>I4_INPUT</b> prints a prompt string and reads an integer from the user.
</li>
<li>
<b>I4_RANGE_INPUT</b> reads a pair of integers from the user, representing a range.
</li>
<li>
<b>MASS_MATRIX</b> computes the mass matrix.
</li>
<li>
<b>NODE_T6</b> returns the basis nodes for a 6 node triangle.
</li>
<li>
<b>R8VEC_PRINT</b> prints a double precision vector.
</li>
<li>
<b>REFQBF</b> evaluates a reference element quadratic basis function.
</li>
<li>
<b>S_EQI</b> is a case insensitive comparison of two strings for equality.
</li>
<li>
<b>S_INPUT</b> prints a prompt string and reads a string from the user.
</li>
<li>
<b>S_OF_I4</b> converts an integer to a left-justified string.
</li>
<li>
<b>S_REP_CH</b> replaces all occurrences of one character by another.
</li>
<li>
<b>S_TO_I4</b> reads an I4 from a string.
</li>
<li>
<b>S_TO_R8</b> reads an R8 from a string.
</li>
<li>
<b>S_WORD_COUNT</b> counts the number of "words" in a string.
</li>
<li>
<b>SINGULAR_VECTORS</b> computes the desired singular values.
</li>
<li>
<b>TIMESTAMP</b> prints the current YMDHMS date as a time stamp.
</li>
<li>
<b>TIMESTRING</b> writes the current YMDHMS date into a string.
</li>
<li>
<b>TRIANGLE_UNIT_SET</b> sets weights and abscissas for quadrature in a unit triangle.
</li>
</ul>
</p>
<p>
You can go up one level to <a href = "../f_src.html">
the FORTRAN90 source codes</a>.
</p>
<hr>
<i>
Last revised on 17 July 2004.
</i>
</body>
</html>