Hybrid and enhanced PSO: Novel first order reliability method-based hybrid intelligent approaches

Computing the sensitivity vector in the traditional first order reliability method may provide inaccurate reliability outcomes for discrete performance functions and inefficient computation burden for high-dimensional problems. In this study, two improved particle swarm optimization algorithms are proposed to enhance the convergence rate with global optimal results during the structural reliability analysis. The abilities for convergence speed and global convergence of the particle swarm optimization algorithm are improved using a novel hybrid method called particle swarm optimization-based harmony search algorithm (PSO-HS), and enhanced particle swarm optimization (EPSO). The proposed methods use a dynamic self-adaptive term to execute the local adjusting process. Using twelve numerical-based engineering problems, the structural reliability frameworks developed based on modified versions of particle swarm optimization algorithms are compared to numerous FORM algorithms and the current metaheuristic methods. Results indicated that the novel proposed methods using the improved PSO algorithms are more robust and efficient than the analytical FORM methods for solving high-dimensional engineering problems. Furthermore, compared to the previous metaheuristic approaches, the suggested methods enabled faster convergence.