期刊論文
| 學年 | 115 |
|---|---|
| 學期 | 1 |
| 出版（發表）日期 | 2026-09-01 |
| 作品名稱 | Acoustic black hole-enabled piezoelectric energy harvesting with AI-assisted parameter mapping |
| 作品名稱（其他語言） | |
| 著者 | Yi-Ren Wang; Chen-Yen Yu |
| 單位 | |
| 出版者 | |
| 著錄名稱、卷期、頁數 | MRS Bulletin; Volume 51; August 2026. Available online 1 September. |
| 摘要 | This study proposes a vibration energy-harvesting system that integrates an acoustic black hole (ABH) structure, piezoelectric materials, and a nonlinear coupled-beam configuration. A double elastic steel sheet (DESS) model is developed to describe the interaction between a primary beam and an ABH beam connected through a position-adjustable elastic coupling. The ABH mechanism induces wave slowing and energy localization, providing a favorable region for converting localized mechanical deformation into electrical output. Both theoretical predictions and experimental measurements indicate that the coupling location plays an important role in governing energy transfer. Under the excitation condition investigated in this study, the highest voltage output among the tested locations is obtained when the coupling element is placed near the mid-span of the beam. To accelerate response prediction and parameter mapping, machine-learning surrogate models, including DNN, LSTM, and XGBoost, are trained using a dataset generated from nonlinear RK4 simulations. Among them, XGBoost provides the best balance between prediction accuracy and computational efficiency. The surrogate-assisted parameter map identifies a favorable coupling region that is consistent with the experimental trend. This work suggests that combining ABH-based structural design with physics-based simulation and data-driven surrogate modeling can support efficient design exploration of vibration energy-harvesting systems. |
| 關鍵字 | Acoustic black hole (ABH); Piezoelectric energy harvesting; Energy localization; Surrogate modeling; Machine learning. |
| 語言 | en |
| ISSN | |
| 期刊性質 | 國外 |
| 收錄於 | SCI EI |
| 產學合作 | |
| 通訊作者 | Yi-Ren Wang |
| 審稿制度 | 是 |
| 國別 | USA |
| 公開徵稿 | |
| 出版型式 | ,電子版 |
| 相關連結 |
機構典藏連結 ( http://tkuir.lib.tku.edu.tw:8080/dspace/handle/987654321/129870 ) |