A multiobjective optimization problem is considered, in which binary choice relations are used instead of optimized functions. To solve this problem, it is proposed to use an evolutionary random search algorithm, in which the function of choice in the form of a lock is used instead of the choice function in the form of a preference,. The convergence of the proposed evolutionary algorithms is analyzed, and sufficient conditions for convergence are formulated. The results of the proposed evolutionary search are compared with the results of well-known evolutionary algorithms for one test problem.

No abstract

Problems formulation. In this article, the term "solar architecture" refers to the architectural approach to solve complex problems related to renewable solar energy. In this case, such results that relate to thermodynamics, heat transfer, construction, electronics, materials science, energy saving should be used. These results should lead to the common decision making that goes beyond these scientific disciplines. The problem is to determine the common decisions in the problems of solar energy. Recently research analysis. The scientific results of recent years in this area were limited to solve analysis problems, and synthesis problems were not considered, since no general mechanisms for making decisions were proposed. A significant role in solving the problems of solar architecture, of course, is given to decision making. Algorithms for evolutionary search for solutions can serve as a method for solving systemic problems in the field of solar architecture. Purpose and tasks. The purpose of this study is to develop a general approach to solve the problems of synthesis of solar architecture. The developed approach should provide the calculation of tasks of complex nature in the field of solar architecture. The task is to develop evolutionary search algorithm schemes that are used to solve solar architecture decision-making problems. Conclusions. Specific schemes of evolutionary algorithms that can be applied to solve the tasks are proposed. At the final stage of decision-making with different characteristics, it is proposed to apply the algorithm of evolutionary search for decisions. Two schemes of algorithm of search of the decision are presented: scheme of algorithm of evolutionary search with function of choice in the form of advantage; diagram of an evolutionary search algorithm with a block selection function.