Determination of Approximate Main Engine Power for Chemical Cargo Ships Using Radial Basis Function Neural Network

Burcu Kapanoglu

Electronics and Communications Engineering Department,
Faculty of Electric and Electronics, Yildiz Technical Universty, Istanbul, TURKEY
e-mail: bkapan@yildiz.edu.tr

Ugur Bugra Çelebi

Naval Architecture and Marine Engineering Department,
Faculty of Mechanical Engineering,Yildiz Technical Universty, Istanbul, TURKEY
e-mail: ucelebi@yildiz.edu.tr

Serkan Ekinci

Naval Architecture and Marine Engineering Department,
Faculty of Mechanical Engineering, Yildiz Technical Universty, Istanbul, TURKEY
e-mail: ekinci@yildiz.edu.tr

Tülay Yildirim

Electronics and Communications Engineering Department,
Faculty of Electric and Electronics, Yildiz Technical Universty, Istanbul, TURKEY
e-mail: tulay@yildiz.edu.tr

Abstract

This paper focuses on the use of genetic algorithms in developing an efficient optimization method for fixed geometry journal bearings, considering system stability along with other design criteria, namely power loss, film thickness, film temperature, and film pressure. The results obtained and presented in this study are compared to the results of numerical optimum design methods such as gradient-based method, and show the potential of genetic algorithms in design optimization of three-lobe journal bearings.

Keywords: stability, genetic algorithm, optimization, journal bearing, rotating machines