Wind-Induced Piezoelectric Energy Harvester Optimization and ANN-Based Performance Prediction
Iranian Journal of Science and Technology, Transactions of Mechanical Engineering · 16.06.2026
DOI: 10.1007/s40997-026-00988-1 ↗Kurz gefasst
Fachartikel in der Zeitschrift Iranian Journal of Science and Technology, Transactions of Mechanical Engineering, online erschienen am 16.06.2026. Verfasst von Muhammad Mohamed Salleh und 7 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 50 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Universiti Tun Hussein Onn Malaysia.
Abstract (Originalsprache)
In the past decade, piezoelectric wind energy harvesters have proven to be an effective method for generating electrical energy from fluid flow in small-scale applications. However, their performance is often limited by low power output, which necessitates the exploration of bluff body shape designs. The bluff body with angles and semicircles was adapted from previous research on opening angle designs to enhance aerodynamic forces, thereby increasing energy collection efficiency. This study was conducted to develop a piezoelectric energy harvester (PEH) using various bluff body configurations, evaluate its performance through experimental methods, and validate the results using an Artificial Neural Network (ANN). The experiment was conducted using a wind tunnel structure equipped with a wind source and integrating piezoelectric elements with beams to optimize power output, resulting in a 2-DOF indirectly driven (PEH) with high reliability and wide bandwidth. Experimental research was conducted on eight bluff body geometries, namely (100°, 120°, 140°, 160°, R5, R6, R7, and R8) with varying cutting angles and semicircular profiles across wind speeds from 1 to 10 m/s. Among the tested configurations, the R8 geometry achieved the highest mean power output of 74.4 mW, corresponding to an observed increase of approximately 25.9% relative to the 160° configuration. Prediction using an Artificial Neural Network (ANN) was carried out with a Radial Basis Function Network (RBFN), which was trained on the experimental dataset to predict the output for the new 110° and R5.5 bluff body shapes. The RFBN results show an R 2 value of 0.9675 for the 110° bluff shape and 0.9974. The results from the predictions were validated with experimental data, and it was found that the new R5.5 and 110° bluff body configurations showed a strong correlation for the proposed prediction model. This finding represents a significant advancement in piezoelectric systems to enable self-powered wireless sensor nodes for structural health monitoring, the Internet of Things (IoT), and smart agriculture.
Förderung
- Universiti Tun Hussein Onn Malaysia
Zitieren
APA
Salleh, M. M., Zaman, I., Wagiman, K. R., Mokhtar, M. S., Ibrahim, A. N., Chumari, M. Z., Sambu, M., & Badrin, M. I. (2026). Wind-Induced Piezoelectric Energy Harvester Optimization and ANN-Based Performance Prediction. Iranian Journal of Science and Technology, Transactions of Mechanical Engineering. https://doi.org/10.1007/s40997-026-00988-1
Vancouver
Salleh MM, Zaman I, Wagiman KR, Mokhtar MS, Ibrahim AN, Chumari MZ, et al. Wind-Induced Piezoelectric Energy Harvester Optimization and ANN-Based Performance Prediction. Iranian Journal of Science and Technology, Transactions of Mechanical Engineering. 2026. doi:10.1007/s40997-026-00988-1
BibTeX
@article{Salleh2026,
author = {Salleh, Muhammad Mohamed and Zaman, Izzuddin and Wagiman, Khairul Rijal and Mokhtar, Mohd Shukri and Ibrahim, Anis Nabihah and Chumari, Mohamad Zaky and Sambu, Mathan and Badrin, Muhammad Izzat},
title = {Wind-Induced Piezoelectric Energy Harvester Optimization and ANN-Based Performance Prediction},
journal = {Iranian Journal of Science and Technology, Transactions of Mechanical Engineering},
year = {2026},
doi = {10.1007/s40997-026-00988-1}
}
EUVIMED