Development of a Fuzzy Matlab inference algorithm for decision-making under uncertainty conditions for designed objects

Abstract

The decision -making method in the face of uncertainty for designed objects is proposed. Uncertainty arises in open decision -making tasks in which the dice does not know the total set of factors. The uncertainty situation is characterized by the fact that the choice of a specific action plan can lead to any result from a certain set of options, but the likelihood of influencing random factors is unknown. The purpose of the article is to create a fuzzy algorithm for making decisions in the face of uncertainty for  designed objects in the Matlab environment. The study is based on the use of decision -making method by means of a linguistic variable, which helps to increase the effective decision in a new difficult situation for designed objects. Two types of Mamdani and Sugeno algorithms in the Fuzzy Logic Matlab package were used to realize this goal. Previous studies were used to build a fuzzy knowledge base, namely indexes and quantitative indicators of the impact of the projected object on the components of the environment (air, surface water, soils). Expert rules "If - then" for both algorithms (Mamdani and Sugeno) were built. As a result, numerical values are obtained - indices of evaluation of influences on the environment of the projected objects. The used decision -making method based on fuzzy logical output models is effective in evaluating the influences of industrial objects on the components of the environment, but are more difficult to formalize than numerical methods that do not require the designer of skills in a specific software environment.

  Keywords: fuzzy output, uncertainty, fuzzy logic matlab, fuzzy knowledge base.