Multilayered piezo-tribo hybrid nanogenerator integrated with machine learning for advanced sign language to speech system
摘要
Multilayer structured piezo-triboelectric hybrid nanogenerators (m-PT-HNG) are emerging as promising candidates for next-generation wearable sensors owing to their ability to harvest energy with high sensitivity and enhanced output. In this work, we report a reliable and sensitive multilayered intrinsic piezo-tribo hybrid nanogenerator (m-PT-HNG) based on a multilayer piezoelectric composite nanogenerator (m-PCNG) architecture combined with triboelectric functionality. The m-PCNG fabricated via parallelly connected multilayers demonstrate significant enhancement of output performance compared to single-layer PCNG. The ferroelectric, piezoelectric performance of Cu2O-doped 0.3Ba0.7Ca0.3TiO3-0.7BaSn0.12Ti0.88O3 (BCST-0.01Cu2O) ceramic fillers was systematically optimized by applying various piston loads (10 to 50 kN) and an electric field of 25 kV/cm. The resulting intrinsically coupled m-PT-HNG produces an instantaneous power density of 85.36 mW/m2 at 200 MΩ. To demonstrate practical utility, a sign language recognition smart glove (SLR-SG) was developed integrating the five m-PT-HNGs, enabling accurate sign language classification through a machine learning algorithm and real-time sign language to speech conversion via a mobile application.