Uncertainty propagation via multi-element grid

In this paper, a new method named multi-element grid (MEG) is proposed to solve the uncertainty propagation problem. The MEG can be used for arbitrary probability distributions. In addition, the number of points and accuracy can be flexibly controlled. The proposed method is compared with the Monte Carlo method, the Gauss-Hermite quadrature (GHQ), and the sparse-grid quadrature (SGQ) via the Lorenz problem. The simulation results show that the new proposed MEG method provides more accurate results than the GHQ and the SGQ, and is computationally more efficient than the Monte Carlo method.