Title: Optimal design of planetary gear train for automotive transmissions using advanced meta-heuristics

Authors: Hammoudi Abderazek; Sadiq M. Sait; Ali Riza Yildiz

Addresses: Applied Precision Mechanics Laboratory, Institute of Optics and Precision Mechanics, Setif-1-University, 19000, Setif, Algeria ' Computer Engineering Department, King Fahd University of Petroleum and Minerals, Dhahran, 31261, Saudi Arabia ' Department of Automotive Engineering, Bursa Uludağ University, Görükle, Bursa, 16059, Turkey

Abstract: In this paper, nine recent meta-heuristics have been employed to search for optimal design of an automatic planetary gear train. The function of the designed system is to automatically transmit power and motion in automobiles. Nine mixed decision parameters are considered in the optimisation procedure. The geometric conditions such as the undercutting, the maximum overall diameter of the transmission, as well as the spacing of multiple planets are taken into account to ensure an optimum design. All the above algorithms are tested both quantitatively and qualitatively for solution quality, robustness, and their time complexity is determined. Results obtained illustrate that the utilised approaches can effectively solve the planetary gearbox problem. Besides this, the comparative study indicates that roulette wheel selection-elitist differential evolution (ReDE) outperforms the other algorithms in terms of the statistical results, and FA has the best convergence behaviour. Meanwhile, multi-verse optimisation (MVO) and butterfly optimisation algorithm (BOA) performed better than the other used algorithms when computation time was considered.

Keywords: planetary gearbox; automotive transmissions; discrete optimisation; optimal design; meta-heuristics; engineering optimisation; differential evolution; multi verse optimiser; neural network.

DOI: 10.1504/IJVD.2019.109862

International Journal of Vehicle Design, 2019 Vol.80 No.2/3/4, pp.121 - 136

Received: 03 Aug 2019
Accepted: 09 Mar 2020
Published online: 28 Sep 2020 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article