Search

The stationary probability density function (PDF) solution of the stochastic responses is derived for nonlinear oscillators subjected to both additive and multiplicative Poisson white noises. The PDF solution is governed by the generalized Fokker--Planck--Kolmogorov (FPK) equation and obtained with the exponential-polynomial closure (EPC) method, which was originally proposed for solving the FPK equation. The extended EPC solution procedure is presented for the case of Poisson pulses in this paper. In order to evaluate the effectiveness of the solution procedure, nonlinear oscillators are investigated under multiplicative Poisson white noise excitation on velocity and additive Poisson white noise excitation. Both weakly and strongly nonlinear oscillators are considered, respectively. In the numerical analysis, both the unimodal and bimodal stationary PDFs of oscillator responses are obtained with the EPC method and Monte Carlo simulation. Compared with the simulation results, good agreement is achieved with the presented solution procedure in the case of the polynomial degree being 6, especially in the tail regions of the PDFs of the system responses. [DOI: 10.1115/1.4000385] Keywords: nonlinear, oscillator, generalized FPK equation, probability density function, Poisson white noise

This paper deals with the stationary probability density function solution of the dynamic response for nonlinear oscillators excited by Poisson impulses. The probability density function solution is governed by the generalized Fokker--Planck--Kolmogorov equation. The exponential-polynomial closure method is used to obtain an approximate solution to the generalized Fokker--Planck--Kolmogorov equation. To evaluate the effectiveness of the exponential-polynomial closure method in the case of Poisson excitations, Duffing oscillators, van der Pol oscillators, and an additional nonlinear oscillator excited by external and parametric Poisson impulses with different levels of nonlinearity in stiffness and damping are studied. The numerical results show good agreement with the probability distribution obtained with Monte Carlo simulations including the tail regions, which is of significance for reliability analyses. DOI: 10.2514/1.36556