Performance Investigation of a Nonlinear Energy Harvester With Random Vibrations and Subthreshold Deterministic Signals

In the field of energy harvesting, it has been demonstrated that under the appropriate conditions, nonlinear configurations of the harvester can provide better performance, compared with linear resonant oscillators; the performance is quantified in terms of the amount of energy extracted from environmental mechanical or seismic vibrations. In this paper, the results of investigations on a system for energy harvesting from wideband vibrations, using a nonlinear snap-through-buckling configuration and two piezoelectric actuators, placed at the stable minima of the potential energy function that underpins the dynamics of the flexible beam are presented. A nonlinear model is proposed that can describe the device behavior when it is driven by a suprathreshold deterministic signal, a wideband noise, or a subthreshold signal superimposed onto a (usually bandlimited) noise background. In the latter case, the system is seen to exhibit the so-called stochastic resonance behavior. The bandwidth of the nonlinear energy harvester is about 15 Hz, which is compatible with vibrational energy sources at low frequencies, e.g., a walking human. The device is seen to generate power up to 160 mu W when subject to a noise limited at 15 Hz. The power is sufficient to operate a standard wireless sensor node and the conversion efficiency of the harvester is about 12%.