Принятие проектных решений на основе нечеткого отношения предпочтения

Abstract

Purpose: One of the most crucial stages in designing sophisticated technical objects is analyzing alternative design decisions and
choosing the best one. At early stages of the design, this problem has to be solved when the initial data are uncertain, relying on developers' knowledge and experience. The goal of this study is finding intellectual algorithms for analysis of design alternatives. Results: A method is chosen for the analysis of alternative decisions, using a fuzzy order relation on a set of linguistic vector evaluations from experts. It assumes the following operations: calculating the values of a μ< membership function for every paired comparison of the alternatives by every r-th criterion μ<(Kr(xi), Kr(xj)); calculating the respective values of μ≥ membership functions; calculating the preference degree for every alternative by convolution of the membership functions values according to the criteria based on the given structure of preferences; comparison of the alternatives by their ranking. An example is given for calculating the values of preference degrees for the case of two alternatives estimated by four parameters. Conclusion: The proposed approach can be used for a well-reasoned design decision making when the initial data are uncertain.