Assessment of modern metaheuristic algorithms -- HS, ABC and FA -- in shape and size optimisation of structures with different types of constraints (Q2256924)

Summary: Shape and size optimisation of structures combining different types of constraints normally causes difficult sensitivity analysis. Thus, choosing the appropriate method to solve this kind of problem is of paramount importance. There are many metaheuristic methods and new ones continue being developed. An assessment of these algorithms may help to choose one. In this paper, three recent metaheuristic algorithms, harmony search (HS), artificial bee colony (ABC) and firefly algorithm (FA), are evaluated. The effectiveness of these methods are demonstrated by solving seven benchmark structural optimisation problems. Additionally, the optimisation of a realistic transmission tower is performed, which may serve as benchmark to validate new optimisation methods. A parametric study was also carried out and the statistical results were presented. The results showed that the three algorithms reached, in a relatively low computational time, better results than most literature, emphasising their excellent ability. Comparisons among the algorithms could be established.