Machine learning-driven optimization of locally resonant metamaterials for simultaneous vibration control and effective triboelectric sensing
- 문서 유형
- 연구출판물 · 학술지 논문
- 등재
- SCIEScopus
- 저널 순위
- JCR Q1Scopus Q1
- 저널
- NANO ENERGY142 · 2025
- 피인용
- Web of Science 9Scopus 9
- FWCI (Scopus)
- 1.31
- JCR IF
- JCR 2024 17.1
- ISSN
- 2211-2855
저자
- 박지교제1저자Pusan Natl Univ, Sch Mech Engn, 2,Busandaehak Ro 63 Beon Gil, Busan 46241, South Korea
- 김백규Pusan Natl Univ, Sch Mech Engn, 2,Busandaehak Ro 63 Beon Gil, Busan 46241, South Korea
- 송진영Pusan Natl Univ, Sch Mech Engn, 2,Busandaehak Ro 63 Beon Gil, Busan 46241, South Korea
- 고건Pusan Natl Univ, Sch Mech Engn, 2,Busandaehak Ro 63 Beon Gil, Busan 46241, South Korea
- 이호경Pusan Natl Univ, Sch Mech Engn, 2,Busandaehak Ro 63 Beon Gil, Busan 46241, South Korea
- Choi, Dongwhi교신저자Kyung Hee Univ, Dept Mech Engn, Integrated Engn Program, 1732 Deogyeong Daero, Yongin 17104, Gyeonggi Do, South Korea
- 백승훈교신저자Pusan Natl Univ, Sch Mech Engn, 2,Busandaehak Ro 63 Beon Gil, Busan 46241, South Korea
- 박상민교신저자Pusan Natl Univ, Sch Mech Engn, 2,Busandaehak Ro 63 Beon Gil, Busan 46241, South Korea
초록
Metamaterial-based multifunctional devices capable of both vibration attenuation and monitoring, have shown great potential to increase system stability and performance. However, designing these systems remains challenging due to complex dynamic behavior, narrow operational frequency ranges, and difficulties in simultaneously optimizing both functionalities using conventional analytical methods. This study presents a multi-metamaterial device with a self-powered sensing capability and corresponding its machine learning-driven design framework. The proposed device simultaneously achieves low-frequency vibration suppression and electricity generation based on ferroelectric-assisted triboelectric nanogenerators (f-TENG). The mechanism of f-TENG synergistically combines triboelectric charge generation via contact electrification, ferroelectric polarization and piezoelectric activation for output enhancement. The integrated device exhibited a 6.5 % vibration reduction and a peak voltage output of 192.2 V. To address the complexity of designing multi-metamaterial device for multifunctionality for tailored-frequency ranges, a machine learning-driven multi-objective optimization is introduced, significantly reducing computational cost by more than 100 times compared to conventional finite element method (FEM)-based exploration. The optimized structures outperformed dataset-based designs, achieving up to 1.7 dB greater vibration attenuation and up to 5.4-fold voltage enhancement, depending on the target frequency range. The optimized designs are integrated into a digital twin platform for application-tailored vibration control and monitoring. This work establishes an automated design methodology for multifunctional metamaterials, offering a versatile solution for advanced vibration control and self-powered sensing across various operational environments.
키워드
Machine learning-drivenMulti-objective optimizationTriboelectric nanogeneratorVibration AttenuationVibration MetamaterialPREDICTION
식별자
- Scopus EID
- 2-s2.0-105007842811
- WoS UID
- 001511030300003



로그인
연구물 목록으로