High-power-density energy harvester via optimized negative Poisson’s ratio metastructure with nonlinear gradient wall thickness

With the advancement of the Internet of Things (IoT), miniature low-power electronic devices have become prevalent in vibration-detection applications. However, miniaturized cantilever piezoelectric energy harvesters often exhibit insufficient power at low frequencies. To address this limitation, this study proposes a negative Poisson’s ratio energy harvester with nonlinear gradient supercell wall thickness (NEH). The proposed harvester is compared with a plain energy harvester (PEH) and a negative Poisson’s ratio energy harvester with uniform wall thickness (UEH) using parametric 3D modeling, finite element analysis, and experimental evaluations. The findings indicate that the NEH has a lower fundamental resonant frequency, as well as more uniform stress–strain distribution that leads to higher output voltage and power. Subsequently, multi-objective optimization aimed at maximizing power and minimizing stress is conducted using an optimal Latin hypercube design, neural network, and non-dominated sorting genetic algorithm II. This process results in an optimized energy harvester (OEH), which reduces the maximum stress by 11.85% compared with NEH and achieves a power density of 124.89 μ W/g, representing increases of 94.1%, 186.8%, and 348.56% compared with NEH, UEH, and PEH, respectively. The OEH achieves a normalized power density of 374.11 μ W cm − 3 Hz − 1 that surpasses the previous works. Finally, a vibration-driven wireless sensing system (VPSS) is constructed to validate the performance of the proposed energy harvesters, enabling data transmission over approximately 200 m. Under an excitation of 5 m/s 2 , the OEH facilitates temperature signal transmission with 37.5% higher efficiency compared to the NEH, demonstrating significant potential for self-powered wireless monitoring of vibrating equipment. • Propose a negative Poisson’s ratio energy harvester with nonlinear gradient supercell wall thickness. • Apply a machine learning-based multi-objective optimization framework for performance enhancement. • Validate its high power density and mechanical durability under low-frequency excitations experimentally. • Enable self-powered wireless sensing over 200 m in IoT systems.