Armine Karami, Dimitri Galayko
UPMC, Paris

Near-limits electrostatic kinetic energy harvesting from arbitrary input vibrations

Kinetic energy harvesting (KEH) is viewed as a potential candidate to complement or replace batteries as power supplies for autonomous microsystems. In some applications, the system is submitted to external vibrations from its environment, which at its scale constitute a large reserve of energy. KEH is interested in the conception of microsystems allowing to convert part of this energy in electrical energy.
For about fifteen years, the research community on KEH has explored different techniques so as to implement such systems. Among these techniques, the electrostatic transduction mechanism has major advantages, such as compatibility with batch fabrication processes. However, electrostatic KEHs present some challenges with regard to their optimization and to the systematization of their conception. These challenges are linked with the complexity of the electrical interfaces required to optimize the conversion, combined with the nonlinear dynamics inherent to electrostatic KEH.
In this talk, we formally pose the problem kinetic energy harvesting from arbitrary types of external vibration inputs, with performances approaching the physical limits. On this subject, we present the control problems arising from the design of such near-limits kinetic energy harvesters. We then show some early results on the conception of such kinetic energy harvesters, using the electrostatic transduction mechanism.
In particular, the electrical interfaces needed to implement the needed controls, and their impact on the system's performances, are discussed. We finish by discussing, as directions for future research, how more sophisticated controls as well as techniques coming from data science can possibly enhance the system's performances.