Shape Optimization for Blended-Wing-Body Underwater Gliders with Structure Constraint

To improve the hydrodynamic efficiency and performance in shallower glide path angle, blended-wing-body (BWB) configuration is applied in underwater gliders. Plentiful works have been focused on shape optimization of this configuration to improve the lift to drag ratio. However the arrangement of structure components, such as the pressure hull, tends to be neglected in geometric model. In this paper, the pressure hull is set as constraint during optimization. Both planar and sectional shape parameters are taken into account for building parametric model and 20 design variables are used for optimization. Computational fluid dynamics (CFD) method is used for hydrodynamic calculation and the results are used for training neural network as surrogate model. Based on multiple island genetic algorithm, the optimization is conducted to search for the shape with higher lift to drag ratio. The hydrodynamic performance of the obtained optimized model is verified to be better at wide states and the lift to drag ratio is improved by 12.4% at the designed attack of angle.