Maximizing the Harvested Energy from Mechanical Random Vibrations with a Matching Network: A Stochastic Analysis

We consider the problem of modeling and analyzing nonlinear piezoelectric energy harvesters for ambient mechanical vibrations. The equations of motion are derived from the mechanical properties, the characterization of piezoelectric materials, and the circuit description of the electrical load. For random ambient vibrations, modeled as white Gaussian noise, the describing equations become stochastic. The harvester performances are analyzed through time-domain Monte-Carlo simulations. Recently proposed solutions inspired by circuit theory, aimed at improving the power performances of energy harvesters are discussed in presence of random vibrations. Our results show that, even in this case, matching network-based approaches improve significantly the energy harvester performance.