GJO-MLP: A NOVEL METHOD FOR HYBRID METAHEURISTICS MULTI-LAYER PERCEPTRON AND A NEW APPROACH FOR PREDICTION OF WEAR LOSS OF AZ91D MAGNESIUM ALLOY WORN AT DRY, OIL, AND h-BN NANOADDITIVE OIL

Altay O., Gurgenc T.

Surface Review and Letters, cilt.31, sa.6, 2024 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 31 Sayı: 6
  • Basım Tarihi: 2024
  • Doi Numarası: 10.1142/s0218625x24500483
  • Dergi Adı: Surface Review and Letters
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Aerospace Database, Applied Science & Technology Source, Chemical Abstracts Core, Chimica, Communication Abstracts, Compendex, Metadex, Civil Engineering Abstracts
  • Anahtar Kelimeler: Magnesium alloy, wear condition, golden jackal optimization-multi-layer perceptron, metaheuristic optimization, hybrid artificial neural network
  • Manisa Celal Bayar Üniversitesi Adresli: Evet

Özet

In this study, the AZ91D magnesium alloy was worn at different wear conditions (dry, oil, and h-BN nanoadditive oil), loads (10–60 N), sliding speeds (50–150 mm/s) and sliding distances (100–1000 m). Wear losses increased with the increase of applied load, sliding speed, and sliding distance. Wear losses were decreased in the h-BN nanoadditive oil conditions. For the first time, the wear losses were predicted using the hybrid golden jackal optimizer-multi-layer perceptron (GJO-MLP) method proposed in this study, using the experimentally obtained data. In addition, the performance of the proposed method was compared with the whale optimization-MLP (WOA-MLP), genetic algorithm-MLP (GA-MLP) and ant lion optimization-MLP (ALO-MLP) methods, which are widely used in the literature. The results showed that GJO-MLP outperformed other methods with a performance of 0.9784 in R2 value.