The welded beam design based on the Salp Swarm Algorithm

Introduction: The design problem of welded beams is a typical civil engineering design task. The goal is to minimize the manufacturing cost by optimizing parameters such as the thickness, height, length, and weld thickness of the reinforcing bars, while meeting multiple constraints including shear stress, bending stress, and buckling load requirements. Due to the nonlinearity of the objective function, the complexity of the parameter variables, and the numerous constraints, traditional methods are ineffective for this problem. However, swarm intelligence algorithms have the advantages of rapid convergence and strong optimization capabilities. Method: This study proposes a welded beam design method based on the Salp Swarm Algorithm. Firstly, by imitating the foraging and aggregation behaviors of the Salp Swarm in nature, a new population update model is developed, and the optimization objective function is established. Secondly, to improve the original SSA, a circle function is introduced for chaotic mapping, generating a more uniformly distributed initial population, expanding the search space, and enhancing the individual search ability. Finally, the improved SSA is applied to the optimization of welding beam design. Results: The experimental results show that the improved SSA has advantages such as flexibility, no gradient dependence mechanism, and strong local optimization ability. Its search performance is superior to the original SSA. Conclusion: The solution obtained by this method is relatively optimal in solving the welded beam design problem. Compared with other swarm intelligence algorithms (such as genetic algorithms and gravitational search algorithms), it performs better and verifies its importance in practical engineering.