Some issues of multi-criteria optimization of parameters of complex systems

Abstract

Objective. The purpose of the study is to substantiate the design criteria for optimality, taking into account the multi-criteria purpose and functioning of the object.Method. The study was carried out on the basis of multicriteria optimization methods, the hierarchy analysis method, the convolution method, and heuristic methods.Result. The paper describes general problems of multicriteria optimization of mechanical systems. The importance of the problem of substantiating the design criteria for optimality, taking into account the multi-criteria purpose and functioning of the object, and introducing quantitatively measurable goals from a variety of alternative options is indicated. It is noted that the choice and formation of a generalized optimality criterion is the most responsible in solving optimization problems. The need to ensure the correctness of the application of one or another convolution method is indicated. Various approaches to the formation of a generalized optimality criterion are analyzed. Their features, advantages and disadvantages are given. The possibility of correcting parametric constraints in problems of improving optimal solutions is pointed out.Conclusion. When solving multi-criteria problems, it is necessary to pay great attention to both the formulation of the problem and the choice of a system of criteria and the implemented method for solving the problem.

Keywords

• multi-criteria optimization
• optimality criteria
• pareto-optimal solutions
• criteria convolution
• multi-vector optimization