Search

Purpose This paper aims to present a new hybrid algorithm of Particle Swarm Optimization and the Genetic Algorithm (PSOGA) to optimize the space trusses with continuous design variables. The PSOGA is an efficient hybridized algorithm to solve optimization problems. Design/methodology/approach These algorithms have shown outstanding performance in solving optimization problems with continuous variables. The PSO conceptually models the social behavior of birds, in which individual birds exchange information about their position, velocity and fitness. The behavior of a flock is influencing the probability of migration to other regions with high fitness. The GAs procedure is based on the mechanism of natural selection. The present study uses mutation, random selection and reproduction to reach the best genetic algorithm by the operators of natural genetics. Thus, only identical chromosomes or particles can be converged. Findings In this research, using the idea of hybridization PSO and GA algorithms are hybridized and a new meta-heuristic algorithm is developed to minimize the space trusses with continuous design variables. To showing the efficiency and robustness of the new algorithm, several benchmark problems are solved and compared with other researchers. Originality/value The results indicate that the hybrid PSO algorithm improved in both exploration and exploitation. The PSO algorithm can be used to minimize the weight of structural problems under stress and displacement constraints.

Seismic safety of super tall buildings is of paramount importance, since if they get damaged, their repair is very costly on the one hand, and their service interruption requires a substitute space for the occupants, which is very difficult, if not impossible, on the other. Regarding these facts, in seismic design of super tall buildings the desired performance level (PL) is immediate occupancy (IO), and it is obvious that achieving this PL is not possible with conventional seismic design procedures. On this basis, in this paper, the efficiency of using specific energy absorbing stories (SEASs) in improvement of seismic behavior of a set of super tall steel buildings with 50, 60, 70 and 80 stories, with tubular frames combined with X-bracing, as their lateral load bearing system, has been investigated. To create SEASs in each building, in every other few stories the bracing elements have been substituted by dampers, as the additional energy absorption sources. The optimal number and the locations of SEASs in each case have been obtained by an optimization technique based on combination of crossover and mutation. Numerical results show that by considering SEASs in almost one tenth of the buildings’ stories, an optimal design is achieved in each case, and that the optimal use of the SEASs reduces the maximum drifts more than 50% in all considered buildings, resulting in a remarkably higher PL, comparing to that of the buildings designed based on conventional methods.

A new meta-heuristic algorithm is presented for optimum design of engineering problems. This algorithm is inspired by the natural behavior of the forests against the rapidly changing environment. This phenomenon is combined with natural regeneration of forests, and a simple but efficient optimization technique is developed which is called natural forest regeneration (NFR). Some well-studied benchmark optimization problems are investigated, and the results of the NFR are compared with those of some previously published algorithms.

In this paper the recently developed nature inspired metaheuristic algorithm is utilized for optimum design of truss structures with continuous and discrete variables. This algorithm is inspired by the natural process happening in the forests with the rapidly change of environment and their natural regeneration. Based on this process a simple powerful optimization technique is introduced so-called Natural Forest Regeneration (NFR). Some well-studied benchmark structural problems are investigated with both continuous and discrete sizing variables and the results of the NRF are compared to those of some previously developed algorithms.