Analysis of stochastic errors and Gaussianity of random vector process simulated by dimension-reduction representation method

The simulation of random vector process is of great importance, particularly for the nonlinear dynamic analysis and reliability evaluation of large-scale structures. Combining the probability density evolution method, which provides an efficient way to perform the structural dynamic reliability analysis, reducing the randomness degree of the simulation method becomes a new challenge task. In order to address this issue, dimension-reduction representation methods for simulating stationary vector processes were proposed recently. However, the features regarding the stochastic error of power spectral density functions and the Gaussianity of vector processes generated by the dimension-reduction representation methods remain unclear. In this paper, the analytical stochastic errors of power spectral density functions for dimension-reduction representation methods will be given, and be proved through the numerical example. Moreover, the Gaussianity of the generated vector processes will be discussed from the view of the first four statistical moments.(c) 2022 Elsevier Inc. All rights reserved.