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
|
.. _strong_maximum_test:
Strong Maximum Test
-------------------
The Strong Maximum Test is used under the following context: let :math:`\inputRV` be a
probabilistic input
vector with joint density probability :math:`\pdf`, let :math:`\model` be the limit state function of
the model and let :math:`\cD_f = \{\vect{x} \in \Rset^\inputDim \,
/ \, \model(\vect{x}) \le 0\}` be an event whose probability
:math:`P_f` is defined as:
.. math::
:label: PfX12
P_f = \int_{{\model(\inputRV) \le 0}} \pdf\, d\vect{x}
The probability :math:`P_f` is evaluated with the :ref:`form_approximation` and :ref:`sorm_approximation`
methods. These methods use the Nataf
isoprobabilistic transformation which maps the
probabilistic model in terms of :math:`\inputRV` onto an equivalent model in terms of :math:`n`
independent standard normal random :math:`\vect{U}` (refer to :ref:`isoprobabilistic_transformation`). In that new
:math:`\vect{u}`-space,
the event has the new expression defined from the transformed limit state function of the model :math:`G`:
:math:`\cD_f = \{\vect{U} \in \Rset^d \, / \, h(\vect{U}) \le 0\}`.
These analytical methods rely on the assumption that most of the contribution to :math:`P_f`
comes from points located in the vicinity of a particular point :math:`P^*`, the design point, defined in the
:math:`\vect{u}`-space as the point located on the limit state surface
and of maximal likelihood. Given the probabilistic characteristics of the
:math:`\vect{u}`-space,
:math:`P^*` has a geometrical interpretation : it is the point located on the event boundary and at minimal distance
from the center of the :math:`\vect{u}`-space. Thus, the design point :math:`P^*` is the result of a constrained
optimization problem.
Both FORM and SORM methods assume that :math:`P^*` is unique. One important difficulty comes
from the fact that numerical methods involved in the determination of :math:`P^*` gives no
guaranty of a global optimum: the point to which they converge might be a local optimum only.
In that case, the contribution of the points in the vicinity of the real design point is not
taken into account, and this contribution is the most important one.
Furthermore, even in the case where the global optimum has really been found, there may exist
another local optimum :math:`\tilde{P}^*` with
likelihood only slightly inferior to the likelihood of the design point one, which means that it is only slightly
further apart from the
center of the :math:`\vect{u}`-space than the design point. Thus, points in the vicinity of :math:`\tilde{P}^*` may
contribute significantly to the probability
:math:`P_f` and are not taken into account in the FORM and SORM approximations.
In both cases, the FORM and SORM approximations are of bad quality because they neglect important contributions to
:math:`P_f`.
The Strong Maximum Test helps to evaluate the quality of the design point resulting from the optimization algorithm. It checks whether the
design point computed is:
- the *true* design point, which means a global maximum point,
- a *strong* design point, which means that there is no other local
maximum located on the event boundary and with likelihood
only slightly inferior to the likelihood of the design point one.
This verification is very important in order to give sense to the FORM
and SORM approximations.
The principle of the Strong Maximum Test relies on the geometrical definition
of the design point.
The objective is to detect all the points :math:`\tilde{P}^*` in the
ball of radius :math:`R_{\varepsilon} = \beta(1+\delta_{\varepsilon})`
which are potentially the real design point (case of
:math:`\tilde{P}_2^*`) or whose contribution to :math:`P_f` is not
negligible as regards the approximations Form and SORM (case of
:math:`\tilde{P}_1^*`). The contribution of a point is considered as
negligible when its likelihood in the :math:`\vect{u}`-space is less
than :math:`\varepsilon` times the likelihood of the design point. The
radius :math:`R_{\varepsilon}` is the distance to the
:math:`\vect{u}`-space center upon which points are considered as
negligible :math:`P_f`.
In order to catch the potential points located on the sphere of radius
:math:`R_{\varepsilon}` (frontier of the zone of prospection), it is
necessary to go a little further : this is why the test samples
the sphere of radius :math:`R = \beta(1+\tau \delta_{\varepsilon})`,
with :math:`\tau >0`.
Points on the sampled sphere which are in the vicinity of the design
point :math:`P^*` are less interesting than those verifying the event
and located *far* from the design point : the latter might reveal
a potential :math:`\tilde{P}^*` whose contribution to :math:`P_f` has
to be taken into account. The vicinity of the design point is defined
with the angular parameter :math:`\alpha` as the cone centered on
:math:`P^*` and of half-angle :math:`\alpha`.
The number :math:`N` of the simulations sampling the sphere of radius
:math:`R` is determined to ensure that the test detects with a
probability greater than :math:`(1 - q)` any point verifying the event
and outside the design point vicinity.
.. image:: FigureStrongMaxTest.svg
:align: center
The vicinity of the design point is the arc of the sampled sphere which
is inside the half space whose frontier is the linearized limit state
function at the Design Point: the vicinity is
the arc included in the half space :math:`D_1`.
.. image:: StrongMaxTest_vicinity.png
:align: center
:scale: 50
The Strong Maximum Test proceeds as follows. The User selects the
parameters:
- the importance level :math:`\varepsilon`, where
:math:`0 < \varepsilon < 1`,
- the accuracy level :math:`\tau`, where :math:`\tau >0`,
- the confidence level :math:`(1 - q)` where :math:`0<q<1` or the
number of points :math:`N` used to sample the sphere. The parameters
are deductible from one other.
The Strong Maximum Test will sample the sphere of radius
:math:`\beta(1+\tau \delta_{\varepsilon})`, where
:math:`\delta_{\varepsilon} = \sqrt{1 - 2 \frac{\ln(\varepsilon)}{\beta^2}}- 1`.
The test will detect with probability greater than :math:`(1 - q)`
any point of :math:`\cD_f` whose contribution to :math:`P_f` is not
negligible (i.e. with density value in the :math:`\vect{u}`-space
greater than :math:`\varepsilon` times the density value at the design
point).
The Strong Maximum Test provides:
- set 1: all the points detected on the sampled sphere that are in
:math:`\cD_f` and outside the design point vicinity, with the
corresponding value of the limit state function,
- set 2: all the points detected on the sampled sphere that are in
:math:`\cD_f` and in the design point vicinity, with the
corresponding value of the limit state function,
- set 3: all the points detected on the sampled sphere that are
outside :math:`\cD_f` and outside the design point vicinity, with the
corresponding value of the limit state function,
- set 4: all the points detected on the sampled sphere that are
outside :math:`\cD_f` but in the vicinity of the design point, with
the corresponding value of the limit state function.
Points are described by their coordinates in the :math:`\vect{x}`-space.
The parameter :math:`\tau` is directly linked to the hypothesis
according to which the boundary of the space :math:`\cD_f` is supposed
to be well approximated by a plane near the design point, which is
primordial for a FORM approximation of the probability content of
:math:`\cD_f`. Increasing :math:`\tau` is increasing the area where
the approximation FORM is applied.
The parameter :math:`\tau` also serves as a measure of distance from
the design point :math:`P^*` for a hypothetical local maximum:
the larger it is, the further we search for another local maximum.
Numerical experiments show that it is recommended to take
:math:`\tau \leq 4` (see the given reference below).
The following table helps to quantify the parameters of the test for a
problem of dimension 5.
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| :math:`\beta_g` | :math:`\varepsilon` | :math:`\tau` | :math:`1-q` | :math:`\delta_{\varepsilon}` | :math:`N` |
+===================+=======================+================+===============+================================+=============+
| 3.0 | 0.01 | 2.0 | 0.9 | :math:`4.224 e^{-1}` | 62 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 3.0 | 0.01 | 2.0 | 0.99 | :math:`4.224 e^{-1}` | 124 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 3.0 | 0.01 | 4.0 | 0.9 | :math:`4.224 e^{-1}` | 15 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 3.0 | 0.01 | 4.0 | 0.99 | :math:`4.224 e^{-1}` | 30 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 3.0 | 0.1 | 2.0 | 0.9 | :math:`2.295 e^{-1}` | 130 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 3.0 | 0.1 | 2.0 | 0.99 | :math:`2.295 e^{-1}` | 260 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 3.0 | 0.1 | 4.0 | 0.9 | :math:`2.295 e^{-1}` | 26 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 3.0 | 0.1 | 4.0 | 0.99 | :math:`2.295 e^{-1}` | 52 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.01 | 2.0 | 0.9 | :math:`1.698 e^{-1}` | 198 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.01 | 2.0 | 0.99 | :math:`1.698 e^{-1}` | 397 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.01 | 4.0 | 0.9 | :math:`1.698 e^{-1}` | 36 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.01 | 4.0 | 0.99 | :math:`1.698 e^{-1}` | 72 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.1 | 2.0 | 0.9 | :math:`8.821 e^{-2}` | 559 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.1 | 2.0 | 0.99 | :math:`8.821 e^{-2}` | 1118 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.1 | 4.0 | 0.9 | :math:`8.821 e^{-2}` | 85 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
| 5.0 | 0.1 | 4.0 | 0.99 | :math:`8.821 e^{-2}` | 169 |
+-------------------+-----------------------+----------------+---------------+--------------------------------+-------------+
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| :math:`\beta_g` | :math:`\varepsilon` | :math:`\tau` | :math:`N` | :math:`\delta_{\varepsilon}` | :math:`1-q` |
+===================+=======================+================+=============+================================+===============+
| 3.0 | 0.01 | 2.0 | 100 | :math:`4.224e^{-1}` | 0.97 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 3.0 | 0.01 | 2.0 | 1000 | :math:`4.224e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 3.0 | 0.01 | 4.0 | 100 | :math:`4.224e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 3.0 | 0.01 | 4.0 | 1000 | :math:`4.224e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 3.0 | 0.1 | 2.0 | 100 | :math:`2.295e^{-1}` | 0.83 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 3.0 | 0.1 | 2.0 | 1000 | :math:`2.295e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 3.0 | 0.1 | 4.0 | 100 | :math:`2.295e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 3.0 | 0.1 | 4.0 | 1000 | :math:`2.295e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.01 | 2.0 | 100 | :math:`1.698e^{-1}` | 0.69 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.01 | 2.0 | 1000 | :math:`1.698e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.01 | 4.0 | 100 | :math:`1.698e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.01 | 4.0 | 1000 | :math:`1.698e^{-1}` | 1.0 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.1 | 2.0 | 100 | :math:`8.821 e^{-2}` | 0.34 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.1 | 2.0 | 1000 | :math:`8.821 e^{-2}` | 0.98 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.1 | 4.0 | 100 | :math:`8.821 e^{-2}` | 0.93 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
| 5.0 | 0.1 | 4.0 | 1000 | :math:`8.821 e^{-2}` | 0.99 |
+-------------------+-----------------------+----------------+-------------+--------------------------------+---------------+
As the Strong Maximum Test involves the computation of :math:`N` values
of the limit state function, which is computationally intensive, it is
interesting to have more than just an indication about the quality of
:math:`P^*`. In fact, the test gives some information about the
trace of the limit state function on the sphere of radius
:math:`\beta(1+\tau \delta_{\varepsilon})` centered on the origin of the
:math:`\vect{u}`-space. Two cases can be distinguished:
- | Case 1: set 1 is empty. We are confident on the fact that
:math:`P^*` is a design point verifying the hypothesis
according to which most of the contribution of :math:`P_f` is
concentrated in the vicinity of :math:`P^*`. By using the
value of the limit state function on the sample
:math:`(\vect{U}_1, \dots, \vect{U}_N)`, we can check if the limit
state function is reasonably linear in the vicinity of
:math:`P^*`, which can validate the second hypothesis of
FORM.
| If the behavior of the limit state function is not linear, we can
decide to use an importance sampling version of the Monte Carlo
method to compute the probability of failure.
However, the information obtained through the Strong Max Test,
according to which :math:`P^*` is the actual design point,
is quite essential : it allows one to construct an effective importance
sampling density, e.g. a multidimensional Gaussian distribution
centered on :math:`P^*`.
- Case 2: set 1 is not empty. There are two possibilities:
#. We have found some points that suggest that :math:`P^*` is
not a strong maximum, because for some points of the sampled
sphere, the value taken by the limit state function is slightly
negative;
#. We have found some points that suggest that :math:`P^*`
is not even the global maximum, because for some points of the
sampled sphere, the value taken by the limit state function is
very negative.
In the first case, we can decide to use an importance sampling
version of the Monte Carlo method for computing the probability
of failure, but with a mixture of e.g. multidimensional gaussian
distributions centered on the :math:`U_i` in :math:`\cD_f`
(refer to ). In the second case, we can restart the search of
the design point by starting at the detected :math:`U_i`.
.. topic:: API:
- See :class:`~openturns.StrongMaximumTest`
- See :class:`~openturns.FORM`
.. topic:: Examples:
- See :doc:`/auto_reliability_sensitivity/reliability/plot_strong_maximum_test`
.. topic:: References:
- A. Dutfoy, R. Lebrun, 2006, "The Strong Maximum Test: an
efficient way to assess the quality of a design point", PSAM8, New Orleans.
|
