Optimal Structural Design Using Modern Metaheuristic Methods

Abstract

In fact, structural design optimization represents one of the major challenges that most of the practical applications face due to an increase in the problems’ complexity. Traditional methods based on mathematical programming or gradient approaches are limited when there are a large number of design variables with nonlinear constraints. Inspired by natural phenomena, metaheuristic algorithms can present a robust and efficient alternative for solving complex problems.Herein, we introduce the Partial Reinforcement Optimizer (PRO), a novel algorithm inspired by the partial reinforcement extinction (PRE) psychological theory within an evolutionary learning framework. The PRO is applied to two engineering problems, namely the welded beam and the three-bar truss design optimization problems. The obtained results are compared with those of some reference metaheuristic algorithms. In fact, the obtained results ensured that the PRO is quite accurate, robust, and faster in convergence toward optimal solutions.