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
|
/***********************************************/
/**
* @file gnssAntennaNormalsConstraint.cpp
*
* @brief Apply constraints to normals of antenna parametrization.
*
* @author Torsten Mayer-Guerr
* @date 2019-09-14
*/
/***********************************************/
// Latex documentation
#define DOCSTRING docstring
static const char *docstring = R"(
Apply constraints to \file{normal equations}{normalEquation}
containing \configClass{antennaCenterVariations}{parametrizationGnssAntennaType}.
Usually the antenna center variations are estimated together with other parameters
like station coordinates, signal biases and slant TEC in \program{GnssProcessing}.
This results in a rank deficient matrix as not all parameters can be separated.
The deficient can be solved by adding pseudo observation equations as constraints.
To separate antenna center variations and signal biases
apply \config{constraint:mean} for each GNSS \configClass{type}{gnssType}.
The observation equation for the integral mean of antenna center variations (ACV)
in all azimuth~$A$ and elevation~$E$ dependent directions
\begin{equation}
0 = \iint ACV(A,E)\, d\Phi \approx \sum_i ACV(A_i,E_i)\, \Delta\Phi_i
\end{equation}
is approximated by a grid defined by
\config{deltaAzimuth}, \config{deltaZenith}, and \config{maxZenith}.
To separate from station coordinates use \config{constraint:centerMean}
and from slant TEC parameters use \config{constraint:TEC}.
The constraints are applied separately to all antennas matching
the wildcard patterns of \config{name}, \config{serial}, \config{radome}.
See also \program{ParameterVector2GnssAntennaDefinition}.
)";
/***********************************************/
#include "programs/program.h"
#include "base/string.h"
#include "files/fileMatrix.h"
#include "files/fileGnssAntennaDefinition.h"
#include "files/fileParameterName.h"
#include "files/fileNormalEquation.h"
#include "classes/parametrizationGnssAntenna/parametrizationGnssAntenna.h"
/***** CLASS ***********************************/
/** @brief Apply constraints to normals of antenna parametrization.
* @ingroup programsGroup */
class GnssAntennaNormalsConstraint
{
public:
class Constraint
{
public:
enum Type {CENTER, CENTERMEAN, CONSTANT, CONSTANTMEAN, TEC};
Type type;
Bool applyWeight;
GnssType gnssType;
Double sigma;
};
void run(Config &config, Parallel::CommunicatorPtr comm);
};
GROOPS_REGISTER_PROGRAM(GnssAntennaNormalsConstraint, SINGLEPROCESS, "Apply constraints to normals of antenna parametrization", Gnss)
GROOPS_RENAMED_PROGRAM(GnssAntennaRepresentationConstraint, GnssAntennaNormalsConstraint, date2time(2020, 6, 26))
/***********************************************/
template<> Bool readConfig(Config &config, const std::string &name, GnssAntennaNormalsConstraint::Constraint &var, Config::Appearance mustSet, const std::string &defaultValue, const std::string &annotation)
{
std::string choice;
if(!readConfigChoice(config, name, choice, mustSet, defaultValue, annotation))
return FALSE;
if(readConfigChoiceElement(config, "center", choice, "zero center (x,y,z) of a single pattern"))
{
var.type = GnssAntennaNormalsConstraint::Constraint::CENTER;
readConfig(config, "type", var.gnssType, Config::MUSTSET, "", "applied for each matching types");
readConfig(config, "applyWeight", var.applyWeight, Config::DEFAULT, "1", "from normal equations");
readConfig(config, "sigma", var.sigma, Config::DEFAULT, "1e-5", "[m]");
}
if(readConfigChoiceElement(config, "centerMean", choice, "zero center (x,y,z) as (weighted) mean of all patterns"))
{
var.type = GnssAntennaNormalsConstraint::Constraint::CENTERMEAN;
readConfig(config, "applyWeight", var.applyWeight, Config::DEFAULT, "1", "from normal equations");
readConfig(config, "sigma", var.sigma, Config::DEFAULT, "1e-5", "[m]");
}
if(readConfigChoiceElement(config, "constant", choice, "zero constant (mean of all directions) of a single pattern"))
{
var.type = GnssAntennaNormalsConstraint::Constraint::CONSTANT;
readConfig(config, "type", var.gnssType, Config::MUSTSET, "", "applied for each matching types");
readConfig(config, "applyWeight", var.applyWeight, Config::DEFAULT, "1", "from normal equations");
readConfig(config, "sigma", var.sigma, Config::DEFAULT, "1e-5", "[m]");
}
if(readConfigChoiceElement(config, "constantMean", choice, "zero constant (mean of all directions) as (weighted) mean of all patterns"))
{
var.type = GnssAntennaNormalsConstraint::Constraint::CONSTANTMEAN;
readConfig(config, "applyWeight", var.applyWeight, Config::DEFAULT, "1", "from normal equations");
readConfig(config, "sigma", var.sigma, Config::DEFAULT, "1e-5", "[m]");
}
if(readConfigChoiceElement(config, "TEC", choice, "zero TEC computed as (weighetd) least squares from all types"))
{
var.type = GnssAntennaNormalsConstraint::Constraint::TEC;
readConfig(config, "type", var.gnssType, Config::DEFAULT, "*", "applied for combination of matching types");
readConfig(config, "applyWeight", var.applyWeight, Config::DEFAULT, "1", "from normal equations");
readConfig(config, "sigma", var.sigma, Config::DEFAULT, "1e-4", "[TECU]");
}
endChoice(config);
return TRUE;
}
/***********************************************/
void GnssAntennaNormalsConstraint::run(Config &config, Parallel::CommunicatorPtr /*comm*/)
{
try
{
FileName fileNameNormalsOut;
FileName fileNameNormalsIn;
std::vector<Constraint> constraints;
ParametrizationGnssAntennaPtr parametrization;
std::string name, serial, radome;
Angle dAzimuth, dZenith, maxZenith;
renameDeprecatedConfig(config, "outputfileNormalequation", "outputfileNormalEquation", date2time(2020, 6, 3));
renameDeprecatedConfig(config, "inputfileNormalequation", "inputfileNormalEquation", date2time(2020, 6, 3));
readConfig(config, "outputfileNormalEquation", fileNameNormalsOut, Config::MUSTSET, "", "with applied constraints");
readConfig(config, "inputfileNormalEquation", fileNameNormalsIn, Config::MUSTSET, "", "");
readConfig(config, "constraint", constraints, Config::MUSTSET, "", "");
readConfig(config, "antennaCenterVariations", parametrization, Config::MUSTSET, "", "");
readConfig(config, "antennaName", name, Config::OPTIONAL, "*", "apply constraints to all machting antennas");
readConfig(config, "antennaSerial", serial, Config::OPTIONAL, "*", "apply constraints to all machting antennas");
readConfig(config, "antennaRadome", radome, Config::OPTIONAL, "*", "apply constraints to all machting antennas");
readConfig(config, "deltaAzimuth", dAzimuth, Config::DEFAULT, "1", "[degree] sampling of pattern to estimate center/constant");
readConfig(config, "deltaZenith", dZenith, Config::DEFAULT, "1", "[degree] sampling of pattern to estimate center/constant");
readConfig(config, "maxZenith", maxZenith, Config::DEFAULT, "90", "[degree] sampling of pattern to estimate center/constant");
if(isCreateSchema(config)) return;
// ============================
logStatus<<"read normal equations <"<<fileNameNormalsIn<<">"<<Log::endl;
Matrix normals, rhs;
NormalEquationInfo info;
readFileNormalEquation(fileNameNormalsIn, info, normals, rhs);
// ============================
// find all antennas in normals/parameterNames
// -------------------------------------------
std::vector<std::string> antennaName;
std::vector<std::vector<GnssType>> types; // for each antenna
std::vector<std::vector<UInt>> index; // for each antenna, type: index in normals
std::vector<ParameterName> parametrizationNames;
parametrization->parameterName(parametrizationNames);
std::regex wildcard = String::wildcard2regex(GnssAntennaDefinition::str(name, serial, radome));
for(UInt i=0; i<info.parameterName.size(); i++)
if((info.parameterName.at(i).type.find(parametrizationNames.at(0).type) == 0) && std::regex_match(info.parameterName.at(i).object, wildcard))
{
// check antenna name
// ------------------
const UInt idAnt = std::distance(antennaName.begin(), std::find(antennaName.begin(), antennaName.end(), info.parameterName.at(i).object));
if(idAnt >= antennaName.size())
{
antennaName.push_back(info.parameterName.at(i).object);
types.push_back(std::vector<GnssType>());
index.push_back(std::vector<UInt>());
}
// extract GnssType
auto pos = info.parameterName.at(i).type.rfind('.');
if((pos == std::string::npos) || (pos+1 == info.parameterName.at(i).type.size()))
throw(Exception(info.parameterName.at(i).str() + ": GnssType not found in parameter name"));
types.at(idAnt).push_back(GnssType(info.parameterName.at(i).type.substr(pos+1)));
index.at(idAnt).push_back(i);
}
// ============================
// partial derivatives of parametrization coeffcients with respect to center (x,y,z)
// performed by numerical integration and least squares adjustment
Matrix dacv_dcenter;
{
logStatus<<"compute center constraint"<<Log::endl;
const UInt rows = static_cast<UInt>(std::round(2*PI/dAzimuth));
const UInt cols = static_cast<UInt>(std::round(maxZenith/dZenith));
Matrix A(rows*cols, parametrization->parameterCount());
Matrix L(rows*cols, 3);
UInt idx = 0;
for(UInt i=0; i<rows; i++)
for(UInt k=0; k<cols; k++)
{
const Angle azimuth (i*2*PI/rows);
const Angle elevation(PI/2-(k+0.5)*Double(dZenith));
copy(parametrization->designMatrix(azimuth, elevation), A.row(idx));
L(idx, 0) = -std::cos(elevation) * std::cos(azimuth); // x
L(idx, 1) = -std::cos(elevation) * std::sin(azimuth); // y
L(idx, 2) = -std::sin(elevation); // z
A.row(idx) *= std::sqrt(std::cos(elevation)); // area weights
L.row(idx) *= std::sqrt(std::cos(elevation));
idx++;
}
dacv_dcenter = leastSquares(A, L);
}
// ============================
// partial derivatives of parametrization coeffcients with respect to a constant
// performed by numerical integration and least squares adjustment
Matrix dacv_dconst;
{
logStatus<<"compute constant constraint"<<Log::endl;
const UInt rows = static_cast<UInt>(std::round(2*PI/dAzimuth));
const UInt cols = static_cast<UInt>(std::round(maxZenith/dZenith));
Matrix A(rows*cols, parametrization->parameterCount());
Vector l(rows*cols);
UInt idx = 0;
for(UInt i=0; i<rows; i++)
for(UInt k=0; k<cols; k++)
{
const Angle azimuth (i*2*PI/rows);
const Angle elevation(PI/2-(k+0.5)*Double(dZenith));
copy(parametrization->designMatrix(azimuth, elevation), A.row(idx));
l(idx) = 1; // constant
A.row(idx) *= std::sqrt(std::cos(elevation)); // area weights
l.row(idx) *= std::sqrt(std::cos(elevation));
idx++;
}
dacv_dconst = leastSquares(A, l);
}
// ============================
logStatus<<"apply constraints to antennas"<<Log::endl;
const UInt parameterCount = parametrization->parameterCount();
for(UInt idAnt=0; idAnt<antennaName.size(); idAnt++)
{
// antenna name
std::string str;
for(GnssType type : types.at(idAnt))
str += " "+type.str();
logInfo<<" "<<antennaName.at(idAnt)<<str<<Log::endl;
// weight matrices of each pattern: taken from normal equations
std::vector<Matrix> P(types.at(idAnt).size());
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
{
P.at(idType) = normals.slice(index.at(idAnt).at(idType), index.at(idAnt).at(idType), parameterCount, parameterCount);
fillSymmetric(P.at(idType));
P.at(idType).setType(Matrix::GENERAL);
}
Matrix I = identityMatrix(parameterCount);
// --- lambda ---------------------
auto accumulateNormals = [&](Double weight, const std::vector<Matrix> &A)
{
// accumulate constraint normals
for(UInt idType1=0; idType1<types.at(idAnt).size(); idType1++)
if(A.at(idType1).size())
{
rankKUpdate(weight, A.at(idType1), normals.slice(index.at(idAnt).at(idType1), index.at(idAnt).at(idType1), parameterCount, parameterCount));
for(UInt idType2=idType1+1; idType2<types.at(idAnt).size(); idType2++)
if(A.at(idType2).size())
matMult(weight, A.at(idType1).trans(), A.at(idType2), normals.slice(index.at(idAnt).at(idType1), index.at(idAnt).at(idType2), parameterCount, parameterCount));
}
info.observationCount += A.at(0).rows();
};
// --------------------------------
for(const auto &constraint : constraints)
{
switch(constraint.type)
{
case Constraint::CENTER:
{
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
if(types.at(idAnt).at(idType) == constraint.gnssType)
{
logInfo<<" - center "<<types.at(idAnt).at(idType).str()<<Log::endl;
Matrix N = dacv_dcenter.trans() * ((constraint.applyWeight ? P.at(idType) : I) * dacv_dcenter);
N.setType(Matrix::SYMMETRIC);
cholesky(N);
Matrix A = dacv_dcenter.trans() * (constraint.applyWeight ? P.at(idType) : I);
triangularSolve(1., N.trans(), A);
triangularSolve(1., N, A);
rankKUpdate(1./std::pow(constraint.sigma, 2), A, normals.slice(index.at(idAnt).at(idType), index.at(idAnt).at(idType), parameterCount, parameterCount));
info.observationCount += 3;
}
break;
}
case Constraint::CENTERMEAN:
{
logInfo<<" - zero mean center"<<Log::endl;
// normals: estimate antenna center
Matrix N(3, Matrix::SYMMETRIC);
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
N += dacv_dcenter.trans() * ((constraint.applyWeight ? P.at(idType) : I) * dacv_dcenter);
cholesky(N);
std::vector<Matrix> A(types.at(idAnt).size());
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
{
A.at(idType) = (constraint.applyWeight ? (dacv_dcenter.trans()*P.at(idType)) : dacv_dcenter.trans());
triangularSolve(1., N.trans(), A.at(idType));
triangularSolve(1., N, A.at(idType));
}
accumulateNormals(1./std::pow(constraint.sigma, 2), A);
break;
}
case Constraint::CONSTANT:
{
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
if(types.at(idAnt).at(idType) == constraint.gnssType)
{
logInfo<<" - constant "<<types.at(idAnt).at(idType).str()<<Log::endl;
// normals: estimate antenna constant
Double N = inner(dacv_dconst, (constraint.applyWeight ? (P.at(idType)*dacv_dconst) : dacv_dconst));
Matrix A = (1./N) * (constraint.applyWeight ? (dacv_dconst.trans()*P.at(idType)) : dacv_dconst.trans());
// accumulate constraint normals
rankKUpdate(1./std::pow(constraint.sigma, 2), A, normals.slice(index.at(idAnt).at(idType), index.at(idAnt).at(idType), parameterCount, parameterCount));
info.observationCount += 1;
}
break;
}
case Constraint::CONSTANTMEAN:
{
// normals: estimate antenna center
Double N = 0;
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
N += inner(dacv_dconst, (constraint.applyWeight ? (P.at(idType)*dacv_dconst) : dacv_dconst));
std::vector<Matrix> A(types.at(idAnt).size());
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
A.at(idType) = (1./N) * (constraint.applyWeight ? (dacv_dconst.trans()*P.at(idType)) : dacv_dconst.trans());
accumulateNormals(1./std::pow(constraint.sigma, 2), A);
break;
}
case Constraint::TEC:
{
std::string str;
for(GnssType type : types.at(idAnt))
if(type == constraint.gnssType)
str += " "+type.str();
logInfo<<" - zero mean TEC"<<str<<Log::endl;
Matrix N(parametrization->parameterCount(), Matrix::SYMMETRIC);
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
if(types.at(idAnt).at(idType) == constraint.gnssType)
axpy(std::pow(types.at(idAnt).at(idType).ionosphericFactor(), 2), (constraint.applyWeight ? P.at(idType) : I), N);
cholesky(N);
// constraint as pseudo observation equations
std::vector<Matrix> A(types.at(idAnt).size());
for(UInt idType=0; idType<types.at(idAnt).size(); idType++)
if(types.at(idAnt).at(idType) == constraint.gnssType)
{
A.at(idType) = types.at(idAnt).at(idType).ionosphericFactor() * (constraint.applyWeight ? P.at(idType) : I);
triangularSolve(1., N.trans(), A.at(idType));
triangularSolve(1., N, A.at(idType));
}
accumulateNormals(1./std::pow(constraint.sigma, 2), A);
break;
}
}
}
} // for(idAnt)
// ============================
logStatus<<"write constrained normal equations to <"<<fileNameNormalsOut<<">"<<Log::endl;
writeFileNormalEquation(fileNameNormalsOut, info, normals, rhs);
}
catch(std::exception &e)
{
GROOPS_RETHROW(e)
}
}
/***********************************************/
|
