Neural-Based Ensemble AI Meta-Model for Enhanced Motor Design Optimization
IEEE Transactions on Transportation Electrification · 2026-08 · 12 (4) · 5708-5719
DOI: 10.1109/tte.2026.3676834 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Transportation Electrification, 2026-08. Verfasst von Shivanshu Thapliyal und 4 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 40 Literaturangaben. Gefördert von: Kumoh National Institute of Technology.
Abstract (Originalsprache)
Permanent magnet synchronous machines (PMSMs) are vital to electrified transportation due to their superior torque density, efficiency, and reliability. However, traditional workflows that rely on analytical models and finite element analysis (FEAs) are highly computationally intensive, limiting their practicality for large-scale, multiobjective optimization. To address this challenge, we propose an Artificial Intelligence (AI)-driven ensemble metamodel for optimizing integrated electronic brake (IEB) motors. The framework combines gradient boosting regressor (GBR), Gaussian process regressor (GPR), and artificial neural networks (ANNs), leveraging the strengths of boosting for nonlinear regression, probabilistic modeling for uncertainty quantification, and hierarchical learning for capturing complex mappings. When embedded in the nondominated sorting genetic algorithm III (NSGA-III), the ensemble enables simultaneous optimization of torque ripple, cogging torque, and total harmonic distortion (THD) of the back-electromotive force (back EMF), while satisfying torque constraints. The methodology was validated on both extensive (1000-sample) and reduced (25–50 sample) datasets, confirming robust accuracy and efficiency even under data-scarce conditions. A prototype optimized motor demonstrated strong alignment between predicted and experimental performance trends.
Förderung
- Kumoh National Institute of Technology
Zitieren
APA
Thapliyal, S., Gupta, S., Song, B., Oh, S., & Hwang, K. (2026). Neural-Based Ensemble AI Meta-Model for Enhanced Motor Design Optimization. IEEE Transactions on Transportation Electrification, 12(4), 5708-5719. https://doi.org/10.1109/tte.2026.3676834
Vancouver
Thapliyal S, Gupta S, Song B, Oh S, Hwang K. Neural-Based Ensemble AI Meta-Model for Enhanced Motor Design Optimization. IEEE Transactions on Transportation Electrification. 2026;12(4):5708-5719. doi:10.1109/tte.2026.3676834
BibTeX
@article{Thapliyal2026,
author = {Thapliyal, Shivanshu and Gupta, Shubham and Song, Baik-Kee and Oh, Seon-Yeol and Hwang, Kyu-Yun},
title = {Neural-Based Ensemble AI Meta-Model for Enhanced Motor Design Optimization},
journal = {IEEE Transactions on Transportation Electrification},
year = {2026},
volume = {12},
number = {4},
pages = {5708--5719},
doi = {10.1109/tte.2026.3676834}
}
EUVIMED