Causal-relationship-assisted shape design optimization for blended-wing-body underwater gliders

Blended-wing-body underwater gliders (BWBUGs) are a new generation of underwater vehicles that have been successfully applied to long-range missions. It is generally recognized that shapes with a high aspect ratio have a higher lift–drag ratio (LDR) and superior fluid performance. However, the pursuit of higher LDR cannot achieve the optimal overall performance, and there are complex design factors. It is necessary but difficult to conduct a thorough examination of shape layout, energy-carrying and other issues. In this article, the layout and energy models are established, and the causal graph is used as a multidisciplinary analysis method to show the interaction between system variables. A shape design method based on surrogate models and causal graphs is proposed that incorporates knowledge into optimization by combing the information flow and finding the relationship between variables. The results show that causal-assisted optimization can mine the characteristics of variables and discover more valuable shape designs.