Meta-sezgisel Algoritmaların Çok Boyutlu Problem Uzaylarındaki Arama Performanslarının Araştırılması

Depending on technological and scientific developments, industrial systems are becoming more powerful, effective and efficient. Parallel to these developments, the number of items that make up the systems increases and they have a much more complex structure. Increasing the number of design variables, which are the basic elements of systems, and modeling the complex relationships between these variables make it difficult to optimize the problems. In the past, heavily heuristic search algorithms have been used to optimize complex problems. However, in the past, constricted optimization problems with sizes from 10 to 30 have been characterized as multidimensional and complex, and today, problems with design variables ranging from 50 to 10,000 are evaluated in this category. However, there is currently insufficient information about the performance of heuristic search algorithms in multi-dimensional (over 50) search spaces. On the other hand, in the optimization of problems with high complexity and high number of design variables, the performance of algorithms in multi-dimensional search space is determinant. In this study, search performances are investigated in the optimization problems which have the most frequently used modern and traditional algorithms 50 and over design variables in the literature. The results obtained are unique to researchers studying in multi-dimensional search space.