Title: Genetic algorithm for chance constrained reliability stochastic optimisation problems

Authors: Vincent Charles; A. Udhayakumar

Addresses: CENTRUM Católica, Graduate School of Business, Pontificia Universidad Católica del Perú, Jr. Daniel Alomía Robles 125–129, Los Álamos de Monterrico, Santiago de Surco, Lima 33, Peru ' Department of Computer Applications, Hindustan University, Chennai 603 103, Tamil Nadu, India

Abstract: This paper addresses the chance constrained reliability stochastic optimisation problem, in which the objective is to maximise system reliability for the given chance constraints. A problem specific stochastic simulation-based genetic algorithm (GA) is developed for finding optimal redundancy to an n-stage series system with m-chance constraints of the redundancy allocation problem. As GA is a proven robust evolutionary optimisation search technique for solving various reliability optimisation problems and the Monte Carlo (MC) simulation, which is a flexible tool for checking feasibility of chance constraints, we have effectively combined GA and MC simulation in the proposed algorithm. The effectiveness of the proposed algorithm is illustrated for a four-stage series system with two chance constraints.

Keywords: redundancy optimisation; system reliability; stochastic simulation; GAs; genetic algorithms; Monte Carlo simulation; redundancy allocation; reliability optimisation; chance constraints.

DOI: 10.1504/IJOR.2012.047513

International Journal of Operational Research, 2012 Vol.14 No.4, pp.417 - 432

Published online: 11 Jan 2015 *

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