Nonlinear dynamical behavior and energy harvesting analyses of flexoelectric MEMS under residual stresses: Application of machine learning for simulating the system

Abstract

Over the last several decades, there has been a lot of research interest in the micro/nano-scale energy harvester, a technology that offers sustainable energy solutions for different micro/nano-electromechanical systems. Investigate various geometries, topologies, and material options for energy receiver designs, as well as interesting nonlinear responses that may be used to push the boundaries of energy receiver design, energy efficiency, and energy density. In the current work, the residual stresses in the system after fabrication, are modeled as the in-plane loading in the system. Although the majority of designs are restricted to MEMS alone, research shows that in-plane loading in the system greatly affects the non-linear vibration of the energy collector and illustrates the potential of increasing its performance by controlling the in-plane load. The impacts of in-plane loading in the MEMS at the microscale tissues were quantified by creating a stress-based piezoelectric-flexoelectric MEMS model and analyzing the non-linear frequency response using a time integrator and a Newtonian iterator. By contrasting them with the outcomes of mathematically modeling the current system and contrasting the outcomes of the current technique with the outcomes of the earlier study, the findings in a Python environment known as the XGBoost methodology are re-validated. This strategy is founded on the concept of “reinforcement,” which integrates many training methodologies with the detection of underperforming individuals to cultivate proficient learners. Finally, numerous suggestions for raising the nonlinear vibration of the energy harvester are carefully considered.