Design of Large-Scale Structures by an enhanced Metaheuristic utilizing Opposition-based Learning

Structural sizing problems are treated as rewarding optimization tasks characterized by non-differentiable constraints and objective functions. In this category, a number of new large-scale space trusses are offered here as structural benchmarks to declare effectiveness and efficiency of zero-order optimization methods. Furthermore, an optimization method is offered for such a task utilizing opposition-based learning that switches between certain vectors directed to the best individual. The proposed algorithm is developed with simple yet efficient operators as a specific-parameter-Iess method. Although it does not bring about high tuning effort for control parameters, numerical simulations exhibit competitive performance of the proposed method in structural sizing problems compared with recent meta-heuristic algorithms.