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
|
.. _nataf_transformation:
Generalized Nataf Transformation
--------------------------------
The Generalized Nataf transformation is an :ref:`isoprobabilistic transformation <isoprobabilistic_transformation>` which is
used under the following context: the input random vector is :math:`\vect{X}` with marginal
cumulative density functions :math:`F_i` and copula :math:`C`. The copula is assumed to be
elliptical.
Introduction
~~~~~~~~~~~~
Let :math:`\vect{d}` be a deterministic vector, let :math:`g(\vect{X}\,,\,\vect{d})` be the
limit state function of the model and let :math:`\cD_f = \{\vect{X} \in \Rset^n \,/ \,
g(\vect{X}\,,\,\vect{d}) \le 0\}` be an event whose probability :math:`P_f` is defined as:
.. math::
:label: PfX
P_f = \Prob{g(\vect{X}\,,\,\vect{d})\leq 0}= \int_{\cD_f} \pdf\, d\vect{x}
The Generalized Nataf transformation :math:`T` allows one to calculate :math:`P_f`. This mapping
is a diffeomorphism from the support of :math:`f_{\vect{X}}` into the standard space
:math:`\Rset^n`, where distributions are spherical, with zero mean,
unit variance and unit correlation matrix. The type of the spherical
distribution is the type of the elliptical copula :math:`C`.
The Generalized Nataf transformation presented here is a generalization
of the traditional Nataf transformation (see [nataf1962]_): the reference
[lebrun2009a]_ shows that the Nataf transformation can be used
only if the copula of :math:`\vect{X}` is normal. The Generalized Nataf
transformation (see [lebrun2009b]_) extends the Nataf
transformation to elliptical copulas.
Let us recall some definitions.
A random vector :math:`\vect{X}` in :math:`\Rset^n` has an *elliptical
distribution* if and only if there exists a deterministic vector
:math:`\vect{\mu}` such that the characteristic function of
:math:`\vect{X} - \vect{\mu}` is a scalar function of the quadratic
form :math:`\vect{u}^t\mat{\Sigma}\, \vect{u}`:
.. math::
\varphi_{\vect{X}-\vect{\mu}}(\vect{u})=\psi(\vect{u}^t\,\mat{\Sigma}\, \vect{u})
with :math:`\mat{\Sigma}` a symmetric positive definite matrix of
rank :math:`p`. As :math:`\mat{\Sigma}` is symmetric positive, it can
be written in the form :math:`\mat{\Sigma}=\mat{D}\,\mat{R}\,\mat{D}`,
where :math:`\mat{D}` is the diagonal matrix
:math:`\mat{\diag{\sigma_i}}` with :math:`\sigma_i=\sqrt{\Sigma_{ii}}`
and :math:`R_{ij}=\frac{\Sigma_{ij}}{\sqrt{\Sigma_{ii}\Sigma_{jj}}}`.
With a specific choice of normalization for :math:`\psi`, in the case
of finite second moment, the covariance matrix of :math:`\vect{X}` is
:math:`\mat{\Sigma}` and :math:`\mat{R}` is then its linear
correlation matrix. The matrix :math:`\mat{R}` is always well-defined,
even if the distribution has no finite second moment: even in this
case, we call it the correlation matrix of the distribution. We note
:math:`\vect{\sigma}=(\sigma_1,\dots,\sigma_n)`.
We denote by :math:`E_{\vect{\mu},\vect{\sigma},\mat{R},\psi}` the
cumulative distribution function of the elliptical distribution
:math:`\cE_{\vect{\mu},\vect{\sigma}, \mat{R},\psi}`.
- An *elliptical copula* :math:`C^E_{\mat{R},\psi}` is the copula of an
elliptical distribution :math:`\cE_{\vect{\mu},\vect{\sigma},\mat{R},\psi}`.
- The *generic elliptical representative* of an elliptical distribution
family :math:`\cE_{\vect{\mu},\vect{\sigma},\mat{R},\psi}` is the
elliptical distribution whose cumulative distribution function is
:math:`E_{\vect{0},\vect{1},\mat{R},\psi}`.
- The *standard spherical representative* of an elliptical distribution
family :math:`\cE_{\vect{\mu},\vect{\sigma},\mat{R},\psi}` is the
spherical distribution whose cumulative distribution function is
:math:`E_{\vect{0},\vect{1},\mat{I}_n,\psi}`.
- The family of distributions with marginal cumulative distribution
functions :math:`F_1,\dots,F_n` and any elliptical copula
:math:`C^E_{\mat{R},\psi}` is denoted by
:math:`{\cD}_{F_1,\dots,F_n,C^E_{\mat{R},\psi}}`. The cumulative
distribution function of this distribution is noted
:math:`D_{F_1,\dots,F_n,C^E_{\mat{R},\psi}}`.
The random vector :math:`\vect{X}` is supposed to be continuous and
with full rank. It is also supposed that its cumulative marginal
distribution functions :math:`F_i` are strictly increasing (so they
are bijective) and that the matrix :math:`\mat{R}` of its elliptical
copula is symmetric positive definite.
Generalized Nataf transformation
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Let :math:`\vect{X}` in
:math:`\Rset^n` be a continuous random vector following the
distribution :math:`D_{F_1,\dots,F_n,C^E_{\mat{R},\psi}}`. The
*Generalized Nataf transformation* :math:`T_{Nataf}^{gen}` is defined
by:
.. math::
\vect{u} = T_{Nataf}^{gen}(\vect{X})=T_3\circ T_2\circ T_1(\vect{X})
where the three transformations :math:`T_1`, :math:`T_2` and :math:`T_3` are given by:
.. math::
T_1 : \Rset^n & \rightarrow \Rset^n\\
\vect{x} & \mapsto \vect{w}=\Tr{(F_1(x_1),\dots,F_n(x_n))}\\
T_2 : \Rset^n & \rightarrow \Rset^n \\
\vect{w} & \mapsto \vect{v}=\Tr{(E^{-1}(w_1),\dots,E^{-1}(w_n))}\\
T_3 : \Rset^n & \rightarrow \Rset^n\\
\vect{v} & \mapsto \vect{u}=\mat{\Gamma}\,\vect{v}
where :math:`E` is the cumulative distribution function of the
standard 1-dimensional elliptical distribution with characteristic
generator :math:`\psi` and :math:`\mat{\Gamma}` is the inverse of the
Cholesky factor of :math:`\mat{R}`.
The distribution of :math:`\vect{W}=T_2\circ T_1(\vect{X})` is the
generic elliptical representative associated to the copula of
:math:`\vect{X}`. The step :math:`T_3` maps this distribution into its
standard representative, following exactly the same algebra as the
normal copula. Thus, in the Generalized Nataf standard space, the
random vector :math:`\vect{U}` follows the standard representative
distribution of the copula of the physical random vector
:math:`\vect{X}`.
If the copula of :math:`\vect{X}` is normal, :math:`\vect{U}` follows
the standard normal distribution with independent components.
.. topic:: API:
- See the available :ref:`Isoprobabilistic transformations <isoprobabilistic_transformation>`.
.. topic:: References:
- O. Ditlevsen and H.O. Madsen, 2004, "Structural reliability methods,"
Department of mechanical engineering technical university of Denmark -
Maritime engineering, internet publication.
- J. Goyet, 1998, "Sécurité probabiliste des structures -
Fiabilité d'un élément de structure," Collège de Polytechnique.
- A. Der Kiureghian, P.L. Liu, 1986,"Structural Reliability Under
Incomplete Probabilistic Information", Journal of Engineering Mechanics,
vol 112, no. 1, pp85-104.
- [lebrun2009a]_
- [lebrun2009b]_
- [lebrun2009c]_
- H.O. Madsen, Krenk, S., Lind, N. C., 1986, "Methods of Structural
Safety," Prentice Hall.
- [nataf1962]_
|
