Font Size: 

The advent of the Internet of Things brings with it a technological challenge associated with the development of new technologies to supply energy to micro-sensors of the most diverse types. Energy harvesting devices are essential in this context, being the search for strategies to maximize the power recovered by these devices a topic of great interest in engineering. This work deals with the study of the efficiency of a bi-stable harvesting device excited by a random signal. For this purpose, it presents the construction of a consistent stochastic model of uncertainties to describe the non-linear dynamic behaviour of the bi-stable system. The physical system of interest consists of a energy harvesting device based on a piezo-magneto-elastic beam, subject to effects of large displacements, modeled by a system of 3 nonlinear differential equations. Uncertainties in the external loading are take into account through a parametric probabilistic approach, where the external excitation is represented by Karhunen-Loève decomposition and and Monte Carlo method is employed to compute the propagation of uncertainties through the stochastic model. The effect of random forcing on system efficiency is analysed by the power spectral density of the output voltage.