ABSTRACT
White noise processes are used in random vibration extensively to model approximately broad band inputs or inputs with very short memory. The chapter shows that there are several types of white noise processes of interest in random vibration: the Gaussian, the Poisson, and the a-stable white noises. The use of the white noise model simplifies the analysis significantly. Moreover, filtered white noise processes can provide satisfactory approximations to inputs with almost any spectral characteristics. The theory of systems subject to white noise can be applied in this case because the augmented vector process including the system and the filter states is driven by white noise. Generalized versions of these calculi or alternative formulations are needed when dealing with Poisson and a-stable white noises. The emphasis is on the response of linear and nonlinear systems to Poisson white noise. Several methods are developed for solutions. Examples are presented to illustrate theoretical results.
