This content is not included in
your SAE MOBILUS subscription, or you are not logged in.
A New Hybrid Particle Swarm Optimization and Jaya Algorithm for Optimal Weight Design of a Gear Train
- Abdennour Hosna - Universite Ferhat Abbas Setif 1, Institut d’Optique et de Mecanique de Precision, Applied Mechanics, Algeria ,
- Ferhat Djeddou - University Ferhat Abbas Sétif 1, Institute of Optics and Precision Mechanics, Applied Precision Mechanics Laboratory, Algeria ,
- Abdelatif Hamouda - University Ferhat Abbas Sétif 1, Institute of Optics and Precision Mechanics, QU.E.R.E Laboratory, Algeria
Journal Article
05-16-02-0012
ISSN: 1946-3979, e-ISSN: 1946-3987
Sector:
Topic:
Citation:
Hosna, A., Djeddou, F., and Hamouda, A., "A New Hybrid Particle Swarm Optimization and Jaya Algorithm for Optimal Weight Design of a Gear Train," SAE Int. J. Mater. Manf. 16(2):141-155, 2023, https://doi.org/10.4271/05-16-02-0012.
Language:
English
Abstract:
Optimization is essential in real-life mechanical engineering problems that
mostly are nonlinear, depend on mixed decision variables, and are usually
subject to constraints. However, most of the studied problems are modelled
assuming continuous variables. A limited number of studies have been devoted to
cases with mixed variables. Moreover, there is a lack of algorithm treating
mixed variable problems properly. This article introduces a hybrid algorithm
that can handle constrained problems depending on continuous or mixed variables.
The proposed algorithm combines two meta-heuristics, Jaya and particle swarm
optimization (PSO). PSO is one of the most popular methods to solve nonlinear
problems, and Jaya is a novel parameter-free optimization algorithm. This new
hybrid optimization algorithm is proposed in order to improve the convergence
speed and to investigate what improvements it will bring to optimization problem
solutions. The developed algorithm was used to tackle two minimization problems
subject to nonlinear constraints. The first one relates to minimizing the weight
of a gear and the second one concerns the minimization of a two-stage planetary
gear train volume since they are widely used in power transmission systems. The
developed algorithm proposes an easy and efficient way to deal with discrete,
normalized, and integer variables. The obtained results are promising and show
that the proposed algorithm gives much better results compared to other
optimization methods that considered the examined problems, while satisfying the
constraints. The algorithm has also proven to be fast in arriving to the optimal
result in comparison with Jaya.