A flexible and efficient optimization design framework for the shape of blend-wing-body underwater glider

Abstract

Shape optimization design plays an important role in improving the gliding performance of blended-wing-body underwater glider (BWBUG), but it faces the challenges of how to handle the geometric constraints and improve the optimization efficiency when building the surrogate-based optimization (SBO) framework. To address these issues, a flexible and efficient optimization design framework for the shape of BWBUG is proposed to perform the shape optimization design of BWBUG. In the framework, an axis-driven geometric free-form deformation method is presented to reduce cost of the BWBUG shape deformation by reasonably controlling the scale of optimization variables, and the rapid calculation method of geometric constraints are derived for the proposed shape deformation method. Based on these, a constrained SBO optimization strategy is developed to directly employ the geometric constraints into the training dataset generation and infill criteria. Finally, the shape of one BWBUG example is optimized, and the results show the flexibility and efficiency of the proposed optimization framework.