EUVIMEDEUVIMED entdecken
Crossref

SPGD: Steepest Perturbed Gradient Descent Optimization

Amir M. Vahedi, Horea T. Ilies

Journal of Mechanical Design · 01.04.2026 · 148 (4)

DOI: 10.1115/1.4070858 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Mechanical Design, online erschienen am 09.03.2026. Verfasst von Amir M. Vahedi und 1 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 50 Literaturangaben. Gefördert von: Defense Advanced Research Projects Agency, Division of Civil, Mechanical and Manufacturing Innovation.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Optimization algorithms are pivotal in advancing various scientific and industrial fields but often encounter obstacles such as trapping in local minima, saddle points, and plateaus (flat regions), which makes the convergence to reasonable or near-optimal solutions particularly challenging. This article presents the steepest perturbed gradient descent (SPGD), a novel algorithm that innovatively combines the principles of the gradient descent method with periodic uniform perturbation sampling to effectively circumvent these impediments and lead to better solutions whenever possible. SPGD is distinctively designed to generate a set of candidate solutions and select the one exhibiting the steepest loss difference relative to the current solution. It enhances the traditional gradient descent approach by integrating a strategic exploration mechanism that significantly increases the likelihood of escaping suboptimal local minima and navigating complex optimization landscapes effectively. Our approach not only retains the directed efficiency of gradient descent but also leverages the exploratory benefits of stochastic perturbations, thus enabling a more comprehensive search for global optima across diverse problem spaces. We demonstrate the efficacy of SPGD in solving the 3D component packing problem, an NP-hard challenge. Preliminary results show a substantial improvement over six established methods, particularly on response surfaces with complex topographies and in multidimensional nonconvex continuous optimization problems. Comparative analyses with established 2D benchmark functions over 30 randomized initial points highlight SPGD’s robustness and reliability in nonconvex optimization. These results emphasize SPGD’s potential as a versatile tool for a wide range of optimization problems.

Förderung

  • Defense Advanced Research Projects Agency
  • Division of Civil, Mechanical and Manufacturing Innovation

Zitieren

APA

Vahedi, A. M., & Ilies, H. T. (2026). SPGD: Steepest Perturbed Gradient Descent Optimization. Journal of Mechanical Design, 148(4). https://doi.org/10.1115/1.4070858

Vancouver

Vahedi AM, Ilies HT. SPGD: Steepest Perturbed Gradient Descent Optimization. Journal of Mechanical Design. 2026;148(4). doi:10.1115/1.4070858

BibTeX
@article{Vahedi2026,
  author = {Vahedi, Amir M. and Ilies, Horea T.},
  title = {SPGD: Steepest Perturbed Gradient Descent Optimization},
  journal = {Journal of Mechanical Design},
  year = {2026},
  volume = {148},
  number = {4},
  doi = {10.1115/1.4070858}
}
Kostenlos starten →