Effect of normalization techniques in multi-criteria decision making methods for the design of ship internal layout from

  • PDF / 1,083,815 Bytes
  • 15 Pages / 595.276 x 790.866 pts Page_size
  • 35 Downloads / 168 Views

DOWNLOAD

REPORT


RESEARCH PAPER

Effect of normalization techniques in multi-criteria decision making methods for the design of ship internal layout from a Pareto optimal set H. Jafaryeganeh 1 & M. Ventura 1 & C. Guedes Soares 1 Received: 22 October 2019 / Revised: 9 February 2020 / Accepted: 16 March 2020 # Springer-Verlag GmbH Germany, part of Springer Nature 2020

Abstract This work aims to assess the suitability of different formulations of normalization techniques within multi-criteria decision-making (MCDM) for applying to final design selection from the obtained Pareto frontier. The MCDM methods are applied as the postoptimization process for decision-making about the selection of final design among the non-dominated solutions for a case study. The Pareto front of a hull layout design problem for a ship is considered as the available set of alternatives that are obtained by multiobjective optimization. The selection of the final design is defined as an MCDM problem, and widely used methods of decisionmaking are adopted to select the final design. Different normalization formulations are applied to transfer the solution values of the Pareto frontier to the required data range for the input of MCDM methods. The effect of the normalization methods is discussed in a comparative study in the decision-making of the internal ship design problem. The results are assessed by comparison of top-ranked alternatives and by evaluating the correlation coefficients of obtained rank for all the alternatives. The normalization methods that are able to keep the dominancy order of the alternatives mostly resulted in similar final design selection. Keywords Internal layout of tankers . Pareto solutions . Multi-objective optimization . MCDM . Normalization of the attribute values

1 Introduction Achieving the best possible design is a changeling process for the designers at the stage of conceptual design because the process needs to integrate several interactive stages for approaching the ideal design, considering the limitations. Those stages include identification of the effective merit functions and constraints (Nowacki 2010), development of a parametric model to be used in an automatic procedure (Koelman 2012), associating the tools for the evaluation functions (Harries et al. 2011), implementing a suitable strategy to obtain the optimum designs (Hart and Vlahopoulos 2010), and finally making decision for selecting the final design among

Responsible Editor: Fred van Keulen * C. Guedes Soares [email protected] 1

Centre for Marine Technology and Ocean Engineering (CENTEC), Instituto Superior Técnico, Universidade de Lisboa, Lisbon, Portugal

the obtained solutions, especially for global optimization processes (Zavala et al. 2014). In general, two strategies of local and global optimization were implemented in the process of improvement of engineering designs. Contrary to the local optimization that results in a unique design with convergence to an optimum solution (Yu et al. 2015), the global optimization usually provi