TY - JOUR
T1 - Cyclostationary gaussian and non-gaussian linearization on analyzing double-well nonlinear oscillators
AU - Chang, R. J.
N1 - Funding Information:
The author would like to thank the Ministry of Science and Technology ( MOST ), Taiwan, ROC for the partial support under contract No. 108-2221-E-006-188.
Publisher Copyright:
© 2020 Elsevier Ltd
PY - 2020/8
Y1 - 2020/8
N2 - The first order and second order bistable Duffing oscillator are selected as typical models for analyzing the stochastic characteristics of double-well nonlinear oscillators excited by sinusoidal force and Gaussian white noise. By employing cyclostationary Gaussian analysis, two solution modes of the almost low-energy and high-energy oscillations are identified. A criterion is proposed to classify two energy modes in oscillation. Cyclostationary numerical Gaussian linearization, analytical Gaussian linearization, and non-Gaussian linearization are developed for analyzing the stochastic response of the bistable oscillator. The time-averaged cyclostationary responses of probability density, mean, and second moment by employing the Gaussian and non-Gaussian linearization are compared and validated with Monte Carlo method. The present study contributes a significant and novel engineering method for analyzing double-well oscillators under white noise and sinusoidal excitations. In addition, a classification criterion, which is important in operation, to ensure high energy orbit is proposed and validated.
AB - The first order and second order bistable Duffing oscillator are selected as typical models for analyzing the stochastic characteristics of double-well nonlinear oscillators excited by sinusoidal force and Gaussian white noise. By employing cyclostationary Gaussian analysis, two solution modes of the almost low-energy and high-energy oscillations are identified. A criterion is proposed to classify two energy modes in oscillation. Cyclostationary numerical Gaussian linearization, analytical Gaussian linearization, and non-Gaussian linearization are developed for analyzing the stochastic response of the bistable oscillator. The time-averaged cyclostationary responses of probability density, mean, and second moment by employing the Gaussian and non-Gaussian linearization are compared and validated with Monte Carlo method. The present study contributes a significant and novel engineering method for analyzing double-well oscillators under white noise and sinusoidal excitations. In addition, a classification criterion, which is important in operation, to ensure high energy orbit is proposed and validated.
UR - https://www.scopus.com/pages/publications/85080131568
UR - https://www.scopus.com/pages/publications/85080131568#tab=citedBy
U2 - 10.1016/j.ymssp.2020.106726
DO - 10.1016/j.ymssp.2020.106726
M3 - Article
AN - SCOPUS:85080131568
SN - 0888-3270
VL - 142
JO - Mechanical Systems and Signal Processing
JF - Mechanical Systems and Signal Processing
M1 - 106726
ER -