Combining multi-attribute decision-making methods with multi-objective optimization in the design of biomass supply chains

WHEELER, JONATHAN; PÁEZ, M.A.; GUILLÉN-GOSÁLBEZ, GONZALO; MELE, FERNANDO DANIEL

COMPUTERS AND CHEMICAL ENGINEERING

PERGAMON-ELSEVIER SCIENCE LTD

Año: 2018 vol. 113 p. 11 - 31

0098-1354

Multi-objective optimization (MOO) is widely appliedin sustainability problems where several objectives must be accounted for inthe analysis. Unfortunately, its complexity grows with the number of objectives,which hampers its practical use. In this paper, we simplify MOO problems viatheir combination with multi-attribute decision-making (MADM) methods. Theapproach identifies a unique Pareto solution of the MOO problem, which bestreflects the decision-makers? preferences, by using weighting factors generatedvia four well-known MADM methods: SWING, SMART, AHP and TRADE OFF. Thecapabilities of this approach are illustrated through its application to thedesign and planning of a sugar/ethanol supply chain using questionnaires filledin by academic experts in the problem. We find that the weights obtained usingMADM algorithms may well differ from the ones given by standard life-cycleassessment methods employed in systems engineering problems. Overall, ourapproach simplifies the MOO problem by identifying solutions consistent with thedecision-makers? preferences and by providing valuable insight on how thesepreferences are articulated in practice.