Flexible positioning process design of aircraft composite panels based on surrogate model optimization

Aircraft composite panels are sensitive to positioner layouts during flexible positioning assembly, and inappropriate layouts may cause non-uniform deformation and local stress concentration. To reduce these effects, this study proposes a positioning-process optimization method based on a simulated component that approximately represents the positioning process of a real panel. First, a parametric positioner model is established using CATIA Automation, and sample data are obtained through batch finite element analyses. Gaussian process regression is then used to construct a surrogate model between the positioner coordinates and the combined objective function, and particle swarm optimization is used to search for an optimized positioner layout. Finally, the optimized layout is reintroduced into the finite element model for verification. The results show that the combined process index, consisting of deformation and stress terms, is reduced by 10.389%-28.896% compared with the pre-optimization cases.