Multi-objective optimization of web-reinforced basalt fiber sandwich panels using improved particle swarm algorithm

Abstract

The present paper proposes a multi-objective optimization design method with web-reinforced basalt fiber sandwich panels as the object of study. The objective of optimization is to minimize the mid-span deflection and cost, given the constraints on ultimate load-bearing capacity. The optimization model for the combination of three key structural parameters—panel thickness, core height, and web spacing—was constructed. In order to enhance the convergence speed and global optimization capability of the particle swarm optimization, the asymmetric adjustment learning factors and the nonlinearly decreasing inertial weight are adopted. Improvement of the multi-objective particle swarm optimization based on MATLAB, resulting in an 11.94% increase in convergence speed. The resulting Pareto solution set exhibits enhanced distribution and convergence properties. In addition, the proposed approach demonstrates superior performance in comparison to the multi-objective genetic algorithm. According to the optimization results, the best length-to-width ratio for balancing the economy and applicability of the sandwich structure is proposed to be 3:1-2:1. A total of 25 sets of solutions are selected on the Pareto frontier curve, uniformly, for the purpose of finite element validation. The mean discrepancy between the optimized and simulated values is merely 1.51%, thereby substantiating the model’s reliability. The Technique for Order Preference by Similarity to Ideal Solution method is employed to identify the optimal trade-off solution, which is the minimum deflection solution at this cost. This analysis elucidates the efficacy of the optimization design method proposed in this paper. It can serve as a reference point for the optimal design of multidimensional complex engineering problems.