Title: A fuzzy multi-objective genetic algorithm for system reliability optimisation
Authors: Michael Mutingi
Addresses: Department of Mechanical Engineering, University of Botswana, P/Bag 0061, Gaborone, Botswana
Abstract: The problem of optimising system reliability is often confronted with imprecise goals concerned with reduction of system costs and improvement of system reliability. Due to the presence of imprecise parameters, the impact of the decision is fuzzy and multi-objective. The present paper models the problem as a fuzzy multi-objective nonlinear program. To effectively handle the fuzzy goals and constraints of the multi-objective decision problem, a fuzzy multi-objective genetic algorithm approach (FMGA) is proposed. The proposed approach is flexible; it allows for generation of intermediate solutions, which eventually lead to high quality solutions. By using fuzzy membership functions, FMGA incorporates the decision maker's preferences and choices that influence system costs and reliability goals. Computations based on benchmark problems demonstrate the utility of the approach.
Keywords: system reliability; multi-objective optimisation; genetic algorithms; fuzzy optimisation; fuzzy set theory; fuzzy evaluation; nonlinear programming; fuzzy logic.
DOI: 10.1504/IJISE.2016.073257
International Journal of Industrial and Systems Engineering, 2016 Vol.22 No.1, pp.1 - 16
Received: 06 Jul 2013
Accepted: 05 Apr 2014
Published online: 30 Nov 2015 *