Strong and weak error estimates for the solutions of elliptic partial differential equations with random coefficients
نویسنده
چکیده
We consider the problem of numerically approximating the solution of an elliptic partial di erential equation with random coe cients and homogeneous Dirichlet boundary conditions. We focus on the case of a lognormal coe cient, we have then to deal with the lack of uniform coercivity and uniform boundedness with respect to the randomness. This model is frequently used in hydrogeology. We approximate this coe cient by a nite dimensional noise using a truncated Karhunen-Loève expansion. We give then estimates of the corresponding error on the solution, both a strong error estimate and a weak error estimate, that is to say an estimate of the error commited on the law of the solution. We obtain a weak rate of convergence wich is twice the strong one. Besides this, we give a complete error estimate for the stochastic collocation method in this case, where neither coercivity nor boundedness are stochastically uniform. To conclude, we apply these results of s trong and weak convergence to two classical cases of covariance kernel choices: the case of an exponential covariance kernel on a box and the case of an analytic covariance kernel, yielding explicit weak and strong convergence rates. Key-words: uncertainty quanti cation, elliptic PDE with random coe cients, KarhunenLoève expansion, strong error estimate, weak error estimate, lognormal distribution ∗ Inria Rennes, Irisa † Ens Cachan antenne de Bretagne in ria -0 04 90 04 5, v er si on 3 5 Ap r 2 01 1 Estimations d'erreurs forte et faible pour des équations aux dérivées partielles elliptiques à coe cients aléatoires Résumé : On s'intéresse à l'approximation numérique de la solution d'une équation aux dérivées partielles elliptique à coe cients aléatoires, avec des conditions de Dirichlet homogènes au bord. On se concentre sur le cas d'un coe cient lognormal, on est ainsi confronté au fait que ce coe cient n'est ni uniformément borné, ni uniformément coercif par rapport à l'aléatoire. Ce modèle est fréquemment utilisé en hydrogéologie. On approche ce coe cient dans un espace aléatoire de dimension nie, en utilisant un développement de Karhunen-Loève. On donne alors des estimations pour l'erreur qui en découle sur la solution, une estimation d'erreur forte mais également une estimation d'erreur faible, c'est à dire une estimation de l'erreur commise sur la loi de la solution. On obtient alors un taux de convergence faible double du taux de convergence forte. De plus, on donne une estimation d'erreur complète pour la méthode de collocation appliquée dans ce cas où le coe cient n'est ni uniformément borné, ni uniformément coercif par rapport à l'aléatoire. Pour conclure, on applique ces résultats à deux choix particuliers de noyaux de covariance: le cas d'une covariance exponentielle sur un pavé et le cas d'une covariance analytique, donnant des taux de convergence forte et faible explicites dans chaque cas. Mots-clés : quanti cation des incertitudes, EDP elliptique à coe cients aléatoires, développement de Karhunen-Loève, estimation d'erreur forte, estimation d'erreur faible, distribution lognormale in ria -0 04 90 04 5, v er si on 3 5 Ap r 2 01 1 ERROR ESTIMATES FOR PDEs WITH RANDOM COEFFICIENTS 3
منابع مشابه
On The Simulation of Partial Differential Equations Using the Hybrid of Fourier Transform and Homotopy Perturbation Method
In the present work, a hybrid of Fourier transform and homotopy perturbation method is developed for solving the non-homogeneous partial differential equations with variable coefficients. The Fourier transform is employed with combination of homotopy perturbation method (HPM), the so called Fourier transform homotopy perturbation method (FTHPM) to solve the partial differential equations. The c...
متن کاملA numerical method for solving nonlinear partial differential equations based on Sinc-Galerkin method
In this paper, we consider two dimensional nonlinear elliptic equations of the form $ -{rm div}(a(u,nabla u)) = f $. Then, in order to solve these equations on rectangular domains, we propose a numerical method based on Sinc-Galerkin method. Finally, the presented method is tested on some examples. Numerical results show the accuracy and reliability of the proposed method.
متن کاملA qMC-spectral method for elliptic PDEs with random coefficients on the unit sphere
We present a quasi-Monte Carlo spectral method for a class of elliptic partial differential equations (PDEs) with random coefficients defined on the unit sphere. The random coefficients are parametrised by the Karhunen-Loève expansion, while the exact solution is approximated by the spherical harmonics. The expectation of the solution is approximated by a quasi-Monte Carlo integration rule. A m...
متن کاملExistence of at least three weak solutions for a quasilinear elliptic system
In this paper, applying two theorems of Ricceri and Bonanno, we will establish the existence of three weak solutions for a quasilinear elliptic system. Indeed, we will assign a differentiable nonlinear operator to a differential equation system such that the critical points of this operator are weak solutions of the system. In this paper, applying two theorems of R...
متن کاملA Kernel-based Collocation Method for Elliptic Partial Differential Equations with Random Coefficients
This paper is an extension of previous work where we laid the foundation for the kernel-based collocation solution of stochastic partial differential equations (SPDEs), but dealt only with the simpler problem of right-hand-side Gaussian noises. In the present paper we show that kernel-based collocation methods can be used to approximate the solutions of high-dimensional elliptic partial differe...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2010