Performance evaluation of metaheuristic search techniques in the optimum design of real size pin jointed structures

In recent years a number of metaheuristic search techniques have been widely used in developing structural optimization algorithms. Amongst these techniques are genetic algorithms, simulated annealing, evolution strategies, particle swarm optimizer, tabu search, ant colony optimization and harmony search. The primary goal of this paper is to objectively evaluate the performance of abovementioned seven techniques in optimum design of pin jointed structures. First, a verification of the algorithms used to implement the techniques is carried out using a benchmark problem from the literature. Next, the techniques compiled in an unbiased coding platform are evaluated and compared in terms of their solution accuracies as well as convergence rates and reliabilities using four real size design examples formulated according to the design limitations imposed by ASD-AISC (Allowable Stress Design Code of American Institute of Steel Institution). The results reveal that simulated annealing and evolution strategies are the most powerful techniques, and harmony search and simple genetic algorithm methods can be characterized by slow convergence rates and unreliable search performance in large-scale problems.