Abstract Topology optimization (TO) with genetic algorithm (GA) is a bio-inspired heuristic optimization technique. In practical applications, it suffers from unstable results, a large number of redundant computations, and low convergence rates. To address this issue, an improved GA, specifically tailored to TO by introducing a strong shape constraint and enriched information obtained from the embedded finite element analyses, is proposed in this study. The strong shape constraint adds a filter to all the individuals at the beginning of each iteration to prevent the analysis of topologies that would not lead to feasible structures. Meanwhile, multiple-input genetic operators leverage additional information from the finite element analyses to guide the mutation and reproduction process, accelerating the convergence rate to the optimal shape. Three case studies present the contributions of the proposed algorithm in terms of robustness, efficiency, and refinement compared to conventional GAs and the Solid Isotropic Material with Penalization (SIMP) method. The results show that the proposed algorithm achieves high robustness and the probability of convergence to impractical shapes is negligible, the computational cost is reduced to one half, the convergence is expedited (compared to conventional GAs), and a refined shape can be obtained and manufactured.
Wang et al. (2026) studied this question.