UNCERTAINTY PROPAGATION: AN ALTERNATIVE TO THE MONTE CARLO METHOD

JORGE BALLABEN; MARTA B. ROSALES; HÉCTOR E. GOICOECHEA

Rio Cuarto

Congreso; VII Congreso de Matemática Aplicada, Computacional e Industrial - MACI 2019; 2019

ASAMACI

Uncertainties are present in every prediction model, since the exact definition of all the relevant parameters is rarely possible. The Monte Carlo method and its variants are undisputedly the most employed in the literature, in order to perform uncertainty propagation. Complex systems, where uncertainty propagation is particularly interesting, require time expensive computations, despite the use of state-of-the-art solvers and parallelization techniques. In this work, a method with the aim of reducing the number of simulations is proposed: the idea is to perform a parametric sweep for a certain parameter X to be considered stochastic, then assign probabilities (according to a previously selected cumulative probability density function) to the values of X, and finally map the corresponding probability values to the target variables. Hence, the probability density function or cumulative density function of the target variables are estimated. The theory, implementation and application examples of the proposed method are herein included.