Response of an Oscillator with Shape Memory Alloy Element Under Gaussian White Noise and Colored Noise Excitation Based on Radial Basis Function Neural Network

The vibration response of an oscillator containing a shape memory alloy (SMA) element under Gaussian noise excitation is investigated in this paper. The constitutive model polynomial can be used to characterize the SMA element’s behavior. Both Gaussian white noise and colored noise excitations are considered. This study applies a semi-analytic approach called RBFNN method to address this problem. In this research, the radial basis function(RBF) is identified as the Gaussian basis function, and a Gaussian basis functions’ weighted sum is used to depict the system’s approximate stationary probability density function (PDF). Simulations with numbers using Monte Carlo simulation (MCS) are conducted, and their results demonstrate excellent agreement with those obtained from RBFNN method.