In this paper, we propose a computationally efficient approach for solving complex multicriteria lexicographic optimization problems, which can be complicated by the multiextremal nature of the efficiency criteria and extensive volume of computations required to calculate the criteria values. The formulation of problems is assumed to be the subject to change as well, which, in turn, may require solving dynamically defined sets of multicriteria optimization problems. The proposed approach is based on reducing multidimensional problems to one-dimensional global optimization problems, utilizing efficient global search algorithms, and reusing the search information obtained in the process of calculations. The results of numerical experiments confirm that the proposed approach demonstrates high computational efficiency of solving multicriteria global optimization problems.

在本文中，我们提出了一种用于解决复杂的多准则词典优化问题的计算有效方法，该方法可能由于效率准则的多极端性质和计算准则值所需的大量计算而变得复杂。假设问题的提出方式也将发生变化，这又可能需要解决动态定义的多准则优化问题集。所提出的方法是基于将多维问题简化为一维全局优化问题，利用有效的全局搜索算法，并重用在计算过程中获得的搜索信息。