Reliability-based Robust Design Optimization of Non-contact Piezoelectric Energy Harvester Based on Rotational Mechanism

Abstract

This study performs reliability-based robust design optimization (RBRDO) of a non-contact piezoelectric energy harvester (N-PEH) using magnetic force and considers the reliability of fatigue loading design conditions and the robustness of output voltage. Piezoelectric energy harvesting accumulates kinetic energy from vibrations and loads and converts it into electrical energy for power generation. Unlike previous studies, this design considers optimization processes for aleatory uncertainty of N-PEH using a rotational mechanism of magnetic plucking to enhance the output voltage. A magneto-mechanical analysis was performed to determine the performance factors. The design variables affecting the output voltage of the N-PEH were selected based on analysis of means. A surrogate model was generated using the response surface method, and its accuracy was evaluated using the coefficient of determination (R2). To investigate the changes in the output voltage, an RBRDO case study was conducted based on the fatigue failure safety factors, target reliability indices, and RBRDO weighting factors. Cases 3, 7, and 8 were recommended in terms of the output voltage when considering both the reliability and robustness of the N-PEH design. The mean and standard deviation of the output voltage decreased by 2.16% and 1.3%, respectively, compared to the DDO Case 1. Consequently, a conservative optimal solution that satisfied the reliability of the constraint function and ensured robustness of the objective function was obtained. Furthermore, this study proposes a design optimization framework for piezoelectric systems with guaranteed structural stability and durability of the piezoelectric bimorph.