Chaotic multi-swarm particle swarm optimization for an engineering problem: Welded beam design

Design optimization is an important engineering design activity. In general, design optimization determines the values required for design variables by satisfying constraints to optimize the objective function. Particle Swarm Optimization algorithm suffers unbalanced local search and global search. Meeting Room approach was introduced as a multi-swarm model to improve the Particle Swarm Optimization algorithm. However, in this approach called Multiple Swarm Particle Swarm Optimization, the algorithm can start from a bad position. Therefore, the algorithm may have a slow convergence. This problem can be overcome by using a position created with a chaotic logistics map. Welded Beam Design, which is an engineering problem, mainly aims to minimize the beam cost due to constraints on loading load, shear stress, bending stress and final deflection. The purpose of this thesis is to evaluate the performance of the Chaotic Multiple Swarm Particle Swarm Optimization algorithm in solving this problem. Experimental studies were carried out with different swarm sizes and iteration numbers. According to the results obtained, the Chaotic Multiple Swarm Particle Swarm Optimization algorithm showed a fast convergence process and performed better than other well-known algorithms. Key Words: Engineering Design Problem, Welded Beam Design, Optimization, Chaotic Multi-Swarm Particle Swarm Optimization.