Özet:

This study investigates the topology optimization problem using various optimization approaches, taking inspiration from the 99-line MATLAB code developed by Sigmund. The educational MATLAB code is based on the Solid Isotropic Material with Penalization (SIMP) model of the artificial material density method. The objective is to minimize the compliance function with a weight constraint, with the design variables being the densities of all elements. The aim is to identify a more efficient optimization technique as an alternative to the commonly used optimality criteria algorithm provided by other MATLAB built-in tools. Two types of optimization algorithms are examined: gradient-based methods such as Interior-Point, Sequential Quadratic Programming (SQP), and Active-Set, as well as metaheuristic methods including the Genetic Algorithm. The results are verified and validated by comparing them with existing literature, demonstrating good agreement. Performance assessments are conducted to compare the results obtained from these algorithms in terms of quality and computational efficiency. The numerical findings indicate that the interior-point method outperforms the other investigated methods, although the optimality criteria algorithm remains the most efficient for solving topology optimization problems.

Anahtar Kelimeler: