Preconditioners based on Voronoi quantizers of random variable coefficients for stochastic elliptic partial differential equations

Abstract

A preconditioning strategy is proposed for the iterative solve of large numbers of linear systems with parameter-dependent matrix and right-hand side which arise during the computation of solution statistics of stochastic elliptic partial differential equations with random and spatially variable coefficients sampled by Monte Carlo. Building on the assumption that a truncated Karhunen-Lo\`{e}ve expansion of a known transform of the random coefficient is available, we introduce a compact approximation of the random coefficient in the form of a Voronoi quantizer. The number of Voronoi cells, each of which is represented by a centroidal coefficient, is set to the prescribed number of preconditioners. Upon sampling the random coefficient, the linear system assembled with a given realization of the coefficient is solved using a Krylov subspace iterative solver with the preconditioner whose centroidal coefficient is the closest to the realization. We consider different ways to define and obtain the centroidal coefficients, and we investigate the properties of the induced preconditioning strategies in terms of average number of solver iterations for sequential simulations, and of load balancing for parallel simulations. Another approach, which is based on deterministic grids on the system of stochastic coordinates of the truncated representation of the random coefficient, is proposed with a stochastic dimension that increases with the number of preconditioners. This approach allows to bypass the need for preliminary computations in order to determine the optimal stochastic dimension of the truncated approximation of the random coefficient for a given number of preconditioners.