A novel hybrid method for the stationary non-Gaussian wind pressures simulation based on maximum entropy method and moment-based translation function

The simulation of stationary multivariate wind pressures with non-Gaussianity is of great important to the structural wind-resistant design. The moment-based translation process methods such as Johnson transformation model (JTM) have attracted more attentions and been widely used in the non-Gaussian simulation due to their conceptual straightforwardness. However, they perform poorly for the wind pressures with strong non-Gaussian characteristics. Moreover, the simulation efficiency is still constrained by the iterative calculation of obtaining the cross-correlation function (CCF) of the underlying Gaussian wind pressures. This study proposes a novel hybrid approach for simulating stationary multivariate non-Gaussian wind pressures based on maximum entropy method and moment-based translation function. In the developed approach, a series of closed-form expressions to determine the CCF of the underlying Gaussian wind pressures are developed for very large areas in the Pearson diagram, and an interpolation-based scheme is proposed to estimate the Gaussian CCF for the remaining areas in the Pearson diagram. The numerical example, in which the wind pressures from the wind tunnel tests were used, demonstrates that the proposed hybrid approach has higher simulation efficiency and better simulation accuracy as compared with the conventional moment-based simulation methods such as JTM.