Optimal design of truss structures using a hybrid method based on particle swarm optimizer and cultural algorithm

In the present study, an efficient hybrid algorithm is proposed based on the particle swarm optimizer (PSO) and the cultural algorithm (CA) for the optimal design of truss structures. In this method, the cultural space defined by the CA has been used to improve the PSO method. The cultural space models the cultural information about individuals in the population space. In the so-called particle swarm optimizer cultural (PSOC) algorithm, three modifications are made on the standard PSO. First, the components related to the best memory of each particle are modified according to the range of cultural space. Second, cultural space is used in order to modify the variables violating the allowable range. And third, with the introduction of Personal Current Bound Strategy (PCBS), any unnecessary structural analyzes, that consequently impose a lot of additional cost in the optimization process, are avoided. Therefore, if the particles move in an inappropriate direction, and in case a weaker situation is achieved, the relevant situation will initially be rectified using cultural space, and after ensuring that it has been placed in the upper limits of society, analysis will then be carried out on it. To show the computational advantages of the PSOC, several benchmark examples are presented. The results show that the PSOC algorithm can converge to a better solution and effectively accelerate the convergence rate compared to other variants of PSO and some other well-known metaheuristic methods.

Automated summary

In this study, an efficient hybrid algorithm combining particle swarm optimizer (PSO) and cultural algorithm (CA) is proposed for optimal truss structure design. The algorithm utilizes cultural space to improve PSO and includes modifications to components, variable adjustments based on cultural space, and the introduction of Personal Current Bound Strategy (PCBS) to avoid unnecessary structural analyzes. Computational advantages of the PSOC algorithm are demonstrated through benchmark examples, showing faster convergence and better solutions compared to other PSO variants and metaheuristic methods.