چكيده به لاتين
In this thesis, a new multi-community algorithm namely the Shuffled Shepherd Optimization Algorithm (SSOA) is proposed. SSOA is inspired by the herding behavior of shepherds in nature. In this optimization method, to calculate the step size, there are both movements toward better and worse members simultaneously. In the SSOA, the population of candidate solutions is divided into small communities. Next, the optimization process inspired by shepherd's behavior performs in each community, and the new position of candidate solution obtain. If the obtained position in terms of objective function value is better than the previous one, it will be updated. To evaluate the performance of the SSOA, 17 mathematical benchmark functions, and 2 classical engineering problems are examined. Moreover, the sizing optimization of 4 truss structures and a double-layer grid are investigated. Finally, the performance of the SSOA is verified in simultaneous sizing and layout optimization of truss structures. All obtained results indicate that SSOA has better performance than other considered methods in both aspects of accuracy and convergence speed.
