Quantum computing intelligence algorithm for structural topology optimization

Structural topology optimization methods that are driven by the geometric features of components, such as the Moving Morphable Component (MMC), are widely explored due to their convenient interaction with design software. However, these methods exhibit significant sensitivity to the initial positioning of components, limiting their suitability for applications involving complex geometric boundaries. This study addresses these challenges by introducing a novel dual-layer optimization framework that employs a quantum computing-based intelligent optimization algorithm, Quantum-behaved Particle Swarm Optimization (QPSO). The potential drawbacks of the MMC method in engineering applications, particularly its sensitivity to initial conditions, are critically examined, leading to the proposal of a global-gradient hybrid framework for geometry feature-driven topology optimization. This proposed method demonstrates superior capability in optimizing material arrangements compared to the original MMC framework and enhances engineering applicability, making it more suitable for real-world applications. Through three representative examples, the limitations of the original MMC method are illustrated, and the advantages of the proposed dual-layer framework are highlighted. The results indicate that this method not only overcomes sensitivity issues but also stably identifies superior configurations, particularly for structures with complex geometric boundaries, providing models that facilitate interaction with CAD systems. This method offers a robust and precise approach for optimizing designs in various engineering fields.