Prototyping of vibration testing machine that is generated non-gaussian random vibration which is determined by probability density function and power spectrum

Abstract

Analysis methods for non-Gaussian stochastic processes are required for various applications in the areas of reliability analysis, signal processing, system identification, and so on. On the other hand, most of the conventional studies in terms of non-Gaussian random excitation have been limited to theoretical vibration response analysis, and non-Gaussian vibration experiments using actual systems are still in the developing stage. In this paper, a non-Gaussian random exciter using a waveform generation method based on the Cai and Lin model is developed, and the probability density shapes and power spectrum of the waveforms output from the actual shaker were evaluated. From the evaluation results, it was found that the developed experimental setup can output waveforms that accurately follow the given probability density function and power spectrum, confirming the basic operation as a non-Gaussian random exciter.