Multi-objective optimization of stochastic failure-prone manufacturing system with consideration of energy consumption and job sequences

S. S. Amelian, S. M. Sajadi*, M. Navabakhsh, M. Esmaelian

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

In this paper, multi-objective optimization of energy-aware multi-product failure-prone manufacturing system is explored. The purpose is to determine the best sequence of jobs, optimal production rate and optimum preventive maintenance time for simultaneous optimization of three criterions of total weighted quadratic earliness and tardiness, system reliability and energy-consumption cost. Considering the uncertainties of the problem such as stochastically machine breakdown and maintenance, stochastic processing times as well as NP-hard nature of the problem, it is not possible to propose an analytical solution to this problem. Therefore, two novel algorithms by combining (1) simulation and NSGA-II/PSO and (2) simulation and NSGA-II/GA are proposed for solving this problem. A set of Pareto optimal solutions was obtained via this algorithm. Results show that the both methods converge to a same optimal solution, but the rate of convergence with NSGA-II/PSO is faster than NSGA-II/GA. The algorithms are evaluated by solving small-, medium- and large-scale problems. To the best of our knowledge, multi-product failure-prone manufacturing systems by considering sequence of jobs have not been explored in any paper and for the first time a new hedging point policy is presented for the mentioned problem.

Original languageEnglish
Pages (from-to)3389-3402
JournalInternational Journal of Environmental Science and Technology
Volume16
Issue number7
Early online date28 May 2018
DOIs
Publication statusPublished - 1 Jul 2019

Keywords

  • Energy consumption
  • Failure-prone manufacturing system
  • Modified hedging point policy
  • Multi-objective optimization
  • NSGA-II

Fingerprint

Dive into the research topics of 'Multi-objective optimization of stochastic failure-prone manufacturing system with consideration of energy consumption and job sequences'. Together they form a unique fingerprint.

Cite this