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
|
/***********************************************/
/**
* @file preprocessingVariationalEquationSstFit.cpp
*
* @brief Fit variational equations to sst/orbit observations.
*
* @author Torsten Mayer-Guerr
* @date 2015-07-06
*
*/
/***********************************************/
// Latex documentation
#define DOCSTRING docstring
static const char *docstring = R"(
This program fits two \configFile{inputfileVariational1/2}{variationalEquation} to satellite-to-satellite-tracking (SST) and orbit
observations in a GRACE like configuration. It works similar to \program{PreprocessingVariationalEquationOrbitFit},
see there for details.
As the relative weighting of the observation types is important complex description of the covariances can be set with
\configClass{covarianceSst}{covarianceSstType}, \configClass{covariancePod1}{covariancePodType}, \configClass{covariancePod2}{covariancePodType}.
)";
/***********************************************/
#include "programs/program.h"
#include "files/fileMatrix.h"
#include "files/fileInstrument.h"
#include "files/fileVariationalEquation.h"
#include "files/fileParameterName.h"
#include "misc/varianceComponentEstimation.h"
#include "misc/observation/observationMiscSstVariational.h"
#include "misc/observation/covariancePod.h"
#include "misc/observation/covarianceSst.h"
#include "misc/observation/variationalEquation.h"
#include "misc/observation/variationalEquationFromFile.h"
/***** CLASS ***********************************/
/** @brief Fit variational equations to sst/orbit observations.
* @ingroup programsGroup */
class PreprocessingVariationalEquationSstFit
{
public:
ObservationMiscSstVariationalPtr observationMisc;
CovarianceSstPtr covSst;
CovariancePodPtr covPod1, covPod2;
std::vector<ObservationMiscSst::Arc> observationArc;
// normal equations
// ----------------
Matrix N_Sst, N_Pod; // =A'PA, Normal matrix
Vector n_Sst, n_Pod; // =A'Pl, right hand side
Double lPl_Sst, lPl_Pod; // =l'Pl, weighted norm of the observations
UInt obsCountSst, obsCountPod; // number of observations
UInt outlierCountSst, outlierCountPod;
Double sigma0Sst, sigma0Pod;
Vector x, x1, x2;
void run(Config &config, Parallel::CommunicatorPtr comm);
void buildNormals(UInt arcNo);
};
GROOPS_REGISTER_PROGRAM(PreprocessingVariationalEquationSstFit, PARALLEL, "fit variational equations to sst/orbit observations", Preprocessing, VariationalEquation)
/***********************************************/
void PreprocessingVariationalEquationSstFit::run(Config &config, Parallel::CommunicatorPtr comm)
{
try
{
FileName fileNameOutVariational1, fileNameOutVariational2;
FileName fileNameOutOrbit1, fileNameOutOrbit2;
FileName fileNameOutSolution1, fileNameOutSolution2;
UInt iterCount;
readConfig(config, "outputfileVariational1", fileNameOutVariational1, Config::MUSTSET, "", "approximate position and integrated state matrix");
readConfig(config, "outputfileVariational2", fileNameOutVariational2, Config::MUSTSET, "", "approximate position and integrated state matrix");
readConfig(config, "outputfileOrbit1", fileNameOutOrbit1, Config::OPTIONAL, "", "integrated orbit");
readConfig(config, "outputfileOrbit2", fileNameOutOrbit2, Config::OPTIONAL, "", "integrated orbit");
readConfig(config, "outputfileSolution1", fileNameOutSolution1, Config::OPTIONAL, "", "estimated calibration and state parameters");
readConfig(config, "outputfileSolution2", fileNameOutSolution2, Config::OPTIONAL, "", "estimated calibration and state parameters");
observationMisc = ObservationMiscSstVariationalPtr(new ObservationMiscSstVariational(config));
readConfig(config, "covarianceSst", covSst, Config::MUSTSET, "", "covariance matrix of satellite to satellite tracking observations");
readConfig(config, "covariancePod1", covPod1, Config::MUSTSET, "", "covariance matrix of kinematic orbits (satellite 1)");
readConfig(config, "covariancePod2", covPod2, Config::MUSTSET, "", "covariance matrix of kinematic orbits (satellite 2)");
readConfig(config, "iterationCount", iterCount, Config::DEFAULT, "10", "for the estimation of calibration parameter and error PSD");
if(isCreateSchema(config)) return;
// =============================================
logStatus<<"set up observation equations"<<Log::endl;
observationArc.resize(observationMisc->arcCount());
std::vector<UInt> processNo = Parallel::forEach(observationMisc->arcCount(), [this](UInt arcNo) {observationArc.at(arcNo) = observationMisc->computeArc(arcNo, covSst, covPod1, covPod2);}, comm);
x = Vector(observationMisc->parameterCount());
sigma0Sst = 1;
sigma0Pod = 1;
for(UInt iter=0; iter<iterCount; iter++)
{
// build normals
// -------------
logStatus<<"accumulate normal equations"<<Log::endl;
logInfo<<" parameter count = "<<observationMisc->parameterCount()<<Log::endl;
N_Sst = Matrix(observationMisc->parameterCount(), Matrix::SYMMETRIC);
N_Pod = Matrix(observationMisc->parameterCount(), Matrix::SYMMETRIC);
n_Sst = Vector(observationMisc->parameterCount());
n_Pod = Vector(observationMisc->parameterCount());
lPl_Sst = 0;
lPl_Pod = 0;
obsCountSst = outlierCountSst = 0;
obsCountPod = outlierCountPod = 0;
Parallel::forEachProcess(observationMisc->arcCount(), [this](UInt arcNo) {buildNormals(arcNo);}, processNo, comm);
Parallel::reduceSum(N_Sst, 0, comm);
Parallel::reduceSum(N_Pod, 0, comm);
Parallel::reduceSum(n_Sst, 0, comm);
Parallel::reduceSum(n_Pod, 0, comm);
Parallel::reduceSum(lPl_Sst, 0, comm);
Parallel::reduceSum(lPl_Pod, 0, comm);
Parallel::reduceSum(obsCountSst, 0, comm);
Parallel::reduceSum(obsCountPod, 0, comm);
Parallel::reduceSum(outlierCountSst, 0, comm);
Parallel::reduceSum(outlierCountPod, 0, comm);
UInt outlierCount = outlierCountSst + outlierCountPod;
// Estimate parameters
// -------------------
if(Parallel::isMaster(comm))
{
Matrix N = pow(1./sigma0Sst,2) * N_Sst + pow(1./sigma0Pod,2) * N_Pod;
Vector n = pow(1./sigma0Sst,2) * n_Sst + pow(1./sigma0Pod,2) * n_Pod;
// Regularize unused parameters
for(UInt i=0; i<N.rows(); i++)
if(N(i,i) == 0)
N(i,i) = 1.0;
logStatus<<"solve system of normal equations"<<Log::endl;
x = solve(N,n);
// variance component estimation
cholesky2Inverse(N);
fillSymmetric(N);
fillSymmetric(N_Pod);
fillSymmetric(N_Sst);
const Double ePe_Sst = lPl_Sst - 2*inner(x, n_Sst) + inner(x, N_Sst*x);
const Double ePe_Pod = lPl_Pod - 2*inner(x, n_Pod) + inner(x, N_Pod*x);
const Double r_Sst = obsCountSst - pow(1./sigma0Sst,2)*inner(N, N_Sst);
const Double r_Pod = obsCountPod - pow(1./sigma0Pod,2)*inner(N, N_Pod);
sigma0Sst = Vce::standardDeviation(ePe_Sst, r_Sst, 2.5/*huber*/, 1./*huberPower*/);
sigma0Pod = Vce::standardDeviation(ePe_Pod, r_Pod, 2.5/*huber*/, 1./*huberPower*/);
logInfo<<" aposteriori sigma (SST): "<<sigma0Sst<<Log::endl;
logInfo<<" aposteriori sigma (POD): "<<sigma0Pod<<Log::endl;
logInfo<<" outlier (SST) "<<outlierCountSst<<" of "<<obsCountSst<<" ("<<100.*outlierCountSst/obsCountSst<<"%)"<<Log::endl;
logInfo<<" outlier (POD) "<<outlierCountPod<<" of "<<obsCountPod<<" ("<<100.*outlierCountPod/obsCountPod<<"%)"<<Log::endl;
}
Parallel::broadCast(outlierCount, 0, comm);
if((iter>0) && (outlierCount==0))
break;
Parallel::broadCast(x, 0, comm);
Parallel::broadCast(sigma0Sst, 0, comm);
Parallel::broadCast(sigma0Pod, 0, comm);
} // for(iter)
// Split solution
// --------------
const UInt gravityCount = observationMisc->variationalEquation1.parameterCountGravity();
const UInt state1Count = observationMisc->variationalEquation1.parameterCount() - gravityCount;
const UInt state2Count = observationMisc->variationalEquation2.parameterCount() - gravityCount;
UInt countAParameter = 0;
const UInt idxGravity = countAParameter; countAParameter += gravityCount;
const UInt idxState1 = countAParameter; countAParameter += state1Count;
const UInt idxState2 = countAParameter; countAParameter += state2Count;
x1 = Vector(gravityCount+state1Count);
if(gravityCount) copy(x.row(idxGravity, gravityCount), x1.row(0, gravityCount));
if(state1Count) copy(x.row(idxState1, state1Count), x1.row(gravityCount, state1Count));
x2 = Vector(gravityCount+state2Count);
if(gravityCount) copy(x.row(idxGravity, gravityCount), x2.row(0, gravityCount));
if(state2Count) copy(x.row(idxState2, state2Count), x2.row(gravityCount, state2Count));
if(Parallel::isMaster(comm) && !fileNameOutSolution1.empty())
{
logStatus<<"write solution to <"<<fileNameOutSolution1<<">"<<Log::endl;
writeFileMatrix(fileNameOutSolution1, x1);
std::vector<ParameterName> parameterName;
observationMisc->variationalEquation1.parameterName(parameterName);
writeFileParameterName(fileNameOutSolution1.replaceFullExtension(".parameterName.txt"), parameterName);
}
if(Parallel::isMaster(comm) && !fileNameOutSolution2.empty())
{
logStatus<<"write solution to <"<<fileNameOutSolution2<<">"<<Log::endl;
writeFileMatrix(fileNameOutSolution2, x2);
std::vector<ParameterName> parameterName;
observationMisc->variationalEquation2.parameterName(parameterName);
writeFileParameterName(fileNameOutSolution2.replaceFullExtension(".parameterName.txt"), parameterName);
}
// =============================================================================
// Improve orbits with estimated parameters
// ----------------------------------------
logStatus<<"Improve orbits with estimated parameters"<<Log::endl;
std::vector<VariationalEquationArc> arcs1(observationMisc->variationalEquation1.arcCount());
std::vector<VariationalEquationArc> arcs2(observationMisc->variationalEquation2.arcCount());
Parallel::forEach(arcs1, [&](UInt arcNo) {return observationMisc->variationalEquation1.refineVariationalEquationArc(arcNo, x1);}, comm);
Parallel::forEach(arcs2, [&](UInt arcNo) {return observationMisc->variationalEquation2.refineVariationalEquationArc(arcNo, x2);}, comm);
// =============================================================================
if(Parallel::isMaster(comm) && !fileNameOutVariational1.empty())
{
logStatus<<"write variational equation to file <"<<fileNameOutVariational1<<">"<<Log::endl;
writeFileVariationalEquation(fileNameOutVariational2, observationMisc->variationalEquation1.satellite(), arcs1);
}
if(Parallel::isMaster(comm) && !fileNameOutVariational2.empty())
{
logStatus<<"write variational equation to file <"<<fileNameOutVariational2<<">"<<Log::endl;
writeFileVariationalEquation(fileNameOutVariational2, observationMisc->variationalEquation2.satellite(), arcs2);
}
if(Parallel::isMaster(comm) && !fileNameOutOrbit1.empty())
{
logStatus<<"write orbit to file <"<<fileNameOutOrbit1<<">"<<Log::endl;
std::list<Arc> arcList;
for(UInt arcNo=0; arcNo<arcs1.size(); arcNo++)
arcList.push_back( arcs1.at(arcNo).orbitArc() );
InstrumentFile::write(fileNameOutOrbit1, arcList);
}
if(Parallel::isMaster(comm) && !fileNameOutOrbit2.empty())
{
logStatus<<"write orbit to file <"<<fileNameOutOrbit2<<">"<<Log::endl;
std::list<Arc> arcList;
for(UInt arcNo=0; arcNo<arcs2.size(); arcNo++)
arcList.push_back( arcs2.at(arcNo).orbitArc() );
InstrumentFile::write(fileNameOutOrbit2, arcList);
}
}
catch(std::exception &e)
{
GROOPS_RETHROW(e)
}
}
/***********************************************/
void PreprocessingVariationalEquationSstFit::buildNormals(UInt arcNo)
{
try
{
Vector l = observationArc.at(arcNo).l;
Matrix A = observationArc.at(arcNo).A;
// count observations and calculate index
// --------------------------------------
const UInt countSst = observationArc.at(arcNo).times.at(0).size();
const UInt countPod1 = observationArc.at(arcNo).times.at(1).size();
const UInt countPod2 = observationArc.at(arcNo).times.at(2).size();
UInt obsCount = 0;
const UInt idxSst = obsCount; obsCount += countSst;
const UInt idxPod1 = obsCount; obsCount += 3*countPod1;
const UInt idxPod2 = obsCount; obsCount += 3*countPod2;
// downweight outliers
if(quadsum(x))
{
const Double huber = 2.5;
Vector e = l;
matMult(-1, A, x, e);
for(UInt k=0; k<countSst; k++)
{
const Double s = fabs(e(idxSst+k));
if(s>huber*sigma0Sst)
{
l.row(idxSst+k) *= huber*sigma0Sst/s;
A.row(idxSst+k) *= huber*sigma0Sst/s;
outlierCountSst++;
}
}
for(UInt k=0; k<countPod1; k++)
{
const Double s = sqrt(quadsum(e.row(idxPod1+3*k,3))/3);
if(s>huber*sigma0Pod)
{
l.row(idxPod1+3*k,3) *= huber*sigma0Pod/s;
A.row(idxPod1+3*k,3) *= huber*sigma0Pod/s;
outlierCountPod += 3;
}
}
for(UInt k=0; k<countPod2; k++)
{
const Double s = sqrt(quadsum(e.row(idxPod2+3*k,3))/3);
if(s>huber*sigma0Pod)
{
l.row(idxPod2+3*k,3) *= huber*sigma0Pod/s;
A.row(idxPod2+3*k,3) *= huber*sigma0Pod/s;
outlierCountPod += 3;
}
}
} // if(quadsum(x))
// Sst
rankKUpdate(1., A.row(idxSst, countSst), N_Sst);
matMult(1., A.row(idxSst, countSst).trans(), l.row(idxSst, countSst), n_Sst);
lPl_Sst += quadsum(l.row(idxSst, countSst));
obsCountSst += countSst;
// Pod
rankKUpdate(1., A.row(idxPod1, 3*(countPod1+countPod2)), N_Pod);
matMult(1., A.row(idxPod1, 3*(countPod1+countPod2)).trans(), l.row(idxPod1, 3*(countPod1+countPod2)), n_Pod);
lPl_Pod += quadsum(l.row(idxPod1, 3*(countPod1+countPod2)));
obsCountPod += 3*(countPod1+countPod2);
}
catch(std::exception &e)
{
GROOPS_RETHROW(e)
}
}
/***********************************************/
|
