A NEW METHOD OF NORMALIZATION BASED ON THE CONCEPT OF IDEAL REFERENCE IN FUZZY LOGIC.

Multicriteria, Fuzzy, Ideal reference, Normalization.

For many years, researchers have been devoting themselves to the study of tools that assist individuals or organizations in their decision-making processes. Thus, one of the most used decision-making methodologies in science, engineering and business is Multicriteria Decision Support (AMD), which is based on the purpose of improving the quality of decisions related to situations in which two alternatives, at least, are analyzed in a comparative way based on criteria and objectives that are often related in a conflicting way. Recently, in the context of AMD, the authors Cables, Lamata and Verdegay (2016) presented the ideal reference concept from the Ideal Reference Method (RIM). Considered by the authors as a variant method of the Technique for Order of Preference by Similarity with the Ideal Solution (TOPSIS) and Multicriteria Optimization and Compromise Solution (VIKOR). This method allows the ideal solutions to vary between the maximum value and the minimum value, which does not occur in the TOPSIS and VIKOR methods, since the ideal solutions for these methods are the extreme values. In 2018, the RIM was extended to the Fuzzy environment by the same authors who called it an Ideal Fuzzy Reference Method (FRIM). Despite the new possibility that the RIM and FRIM method propose, it is suggested in this work that in some situations, the normalization procedure applied in these methods may not adequately represent the decision maker preference structure. Thus, this study aims to propose a new method of normalization, based on the ideal reference conception and Fuzzy logic, allowing the differentiation of alternatives even when their performances in a given criterion belong to the ideal reference range. For this, the current method and the proposed method will be applied in a real case study, making possible a comparative analysis and verification of the merit of the new method.