Sizing optimization of truss structures with discrete design variables using combined PSO algorithm with Special Particles Method

This paper proposes a modified particle swarm optimization (MPSO) algorithm for discrete sizing optimization of truss structures. The original particle swarm optimization (PSO) is a population-based metaheuristic that fluctuates the search agents about the best solution based on Eberhart functions. The efficiency of the PSO in solving standard optimization problems of well-known problems of truss structures has been demonstrated in literature. However, its performance in tackling the discrete optimization problems of truss structures is not competitive compared with the recent existing metaheuristic algorithms. In the framework of the proposed MPSO a number of worst solutions of the current population is replaced by some variants of the global best solution found so far. Moreover, an efficient mutation operator is added to the algorithm that reduces the probability of getting stuck in local optima. The efficiency of the proposed MPSO is illustrated through two benchmark optimization problems of truss structures.