Multicriteria Evaluation and Optimization Nonlinear Trade-off Scheme

Abstract

Finding a multicriteria solution is inherently a compromise and is based on using subjective information.Given this information and a compromise scheme selected, it is possible to pass fromthe general vector expression to the scalar convolution of partial criteria, which provides a basis for a constructive apparatus to solve multicriteria problems. Solving the problem is based on the hypothesis that there exists a utility function [1] appearing in the DM’s brain during the solution of a specific multicriteria problem. We may state that virtually all the approaches to determining the scalar convolution of criteria are reduced to constructing one mathematical model or another of the DM’s utility function. The problem is to approximate correctly the utility function and to construct a substantial mathematical model as a scalar convolution, adequate to the given situation, to solve different multicriteria problems.