Stochastic description of a matched-load mechanical energy harvester

Mechanical energy harvesting is one of the most promising solutions to the renewable powering of dispersed Internet of Things devices. The design of such powering systems, however, is a challenging task, not only because a stochastic description is required to represent the very mechanical energy source which is random in nature, but also because a significant mismatch is often present between the electrical load and the equivalent circuit representing the harvester. In this contribution we propose a sophisticated solution technique allowing for the evaluation of the first two order moments of the output voltage for the stochastic differential equations model representing the device. The approach applies also to nonlinear harvesters ex-ploiting a moment closure technique. We take into consideration also the presence of a reactive matching network aiming at the optimization of the energy flow between the harvesting device and the load. Results, besides validating the stochastic analysis technique, show an important improvement in the output power delivery and in the conversion efficiency.