Based on the classical bending rule that the plies composing the thinner region should be a subset of the ones of the thicker region for two adjacent laminates, a genetic algorithm–based dynamic boundary subset blending model is proposed to optimize the global stacking sequence of composite structures with ply-drops. Besides the stacking sequence chromosome of the guide laminate and ply number chromosome of each panel, a chromosome of a dynamic boundary subset factor is introduced for each panel to obtain a fully blended design. The lower and upper bounds of the dynamic boundary subset factor chromosome for each panel is determined by the ply number chromosomes of the panel and its adjacent panels. The stacking sequence of each panel can be determined by selection from combinations of various stacking sequences. The proposed blending model can solve the problem that laminates with identical thicknesses have the completely same layups even when they are not adjacent to each other. The optimal feasible designs outperform other published solutions for the 18-panel horseshoe configuration problem based on the classical bending rule.

基于经典弯曲规则，组成较薄区域的层应为两个相邻层压板的较厚区域的子集，提出了一种基于遗传算法的动态边界子集混合模型，以优化复合结构的整体堆叠顺序与ply-drops。除了引导叠层的堆叠序列染色体和每个面板的层数染色体之外，还为每个面板引入了动态边界子集因子的染色体，以获得完全融合的设计。每个面板的动态边界子集因子染色体的上下限由面板及其相邻面板的层数染色体确定。每个面板的堆叠顺序可以通过从各种堆叠顺序的组合中进行选择来确定。提出的混合模型可以解决厚度相同的层压板即使彼此不相邻也具有完全相同的铺层的问题。最佳可行设计优于基于经典弯曲规则的18面板马蹄形配置问题的其他已发布解决方案。