Reliability-based Design Optimization for Structures Using Particle Swarm Optimization

Studies on structure design optimization have been conducted extensively over the past decades because it can increase structural efficiency and maximize the engineers' profit. One of the significant current discussions is the reliability aspect of structure design in addition to optimal design. This problem becomes more important especially in sizing and shaping optimization of truss structures. To address the problem, this paper applies Reliability-based Design Optimization (RBDO) by combining Latin Hypercube Sampling (LHS) and metaheuristic algorithms which are Particle Swarm Optimization (PSO) and Barebone PSO (BBPSO). The presence of uncertainty is modeled using LHS and the reliability constraint is added to measure the probability of structural failure. This study aims to investigate the performance of these metaheuristic algorithms in optimizing the structure design while still satisfying the reliability constraints. Two case studies, a welded beam design problem and 15-bar planar truss problem, are used in this study. The performance of these two algorithms will be compared. The obtained results indicate that BBPSO has a better performance than PSO.