Abstract:

An arbitrary space shape free-form deformation (FFD) technique is first established in this paper based on the non-uniform rational B-splines basis function, and any complex configuration can be parameterized through choosing an FFD shape and lattice reasonably. First an airliner wingtip is parameterized using the FFD technique. Then the multi-block structure grid deformation technique is established by the Delaunay graph mapping method. An aerodynamic optimization design system is established by combining the FFD technique, the grouping particle swarm optimization arithmetic with the back propagation (BP) neural network approximation model. Finally, it processes the robust aerodynamic optimization design of the winglet by taking the swept angle, deflection angle and height of the airliner as design variables. The surface pressure contour, pressure distribution of the wing section and load distribution of the initial and optimized winglet are analyzed. The results show that the optimized winglet has significantly better aerodynamic characteristics.

Key words: winglet, FFD technique, robust design, BP neural network, Delaunay graph mapping, particle swarm optimization arithmetic