A new direct statistical analysis of nonlinear systems: CAUTDET

摘要

This paper investigates the nonlinear system analysis problem for achieving higher computational efficiency and accuracy. A novel and efficient technique, covariance analysis unscented transform describing function technique (CAUTDET), is proposed. This method combines linear system covariance analysis algorithm with describing function method based on unscented transform. Furthermore, the quasi-linearization of the nonlinear system are derived to approximate the nonlinearity, while there is no requirement of the probability density function of the nonlinear system input and needs only one time calculation. Moreover, the proposed technique can be viewed as a generalization method of the nonlinear system design and analysis. Finally, simulations are presented to illustrate the effectiveness of the presented analysis method.