A Comparative Study of Meta-Modeling for Response Estimation of Stochastic Nonlinear MDOF Systems Using MIMO-NARX Models

Complex dynamic behavior of nonlinear structures makes it challenging for uncertainty analysis through Monte Carlo simulations (MCS). Surrogate modeling presents an efficient and accurate computational alternative a large number MCS. The previous study has demonstrated that the multi-input multi-output autoregressive with exogenous input (MIMO-NARX) model provides good discrete-time representations deterministic multi-degree-of-freedom (MDOF) structural systems. Model order reduction (MOR) is executed to eliminate insignificant modes reduce burden due too many degrees freedom. In this study, MIMO-NARX strategy integrated different meta-modeling techniques analysis. Different meta-models including Kriging, polynomial chaos expansion (PCE), arbitrary (APC) are used surrogate NARX coefficients system uncertainties. A nine-DOF structure as MDOF evaluate MIMO-NARX. Good fitness statistical responses observed between MCS results original all surrogated predictions. It APC-NARX advantage being data-driven most tool quantification dynamics.