Multiple-criteria Analysis of Plasterboard Systems

Abstract

Multiple-criteria analysis methods are a useful tool for decision making support in everyday engineering practice. They turn out to be especially helpful when one considers numerous variants described by many parameters, such as commercially available drywall systems - plasterboard partition walls, claddings and suspended ceilings. In this study four methods of multiple-criteria analysis are investigated (weighted sum, weighted product, ideal point and PROMETHEE II) and 7 normalization algorithms (vector, Manhattan, maximum, Weitendorf linear, Jüttler-Korth linear, Peldschus nonlinear, Zavadskas and Turskis logarithmic) and their influence on the rankings obtained for 5 groups of products: plasterboard partition walls without fire resistance, EI30 partition walls, hard partition walls, suspended ceilings and ceiling claddings. The study shows high sensitivity of the commonly used methods to input data normalization algorithms and indicates the need for a cautious approach to the results obtained by them.