Multi-Objective Optimization of Supersonic Projectiles using Evolutionary Algorithms

Abstract

An optimization of supersonic projectiles was performed using aerodynamic data obtained from the CART3D CFD solver and the NEAR MISL3 semi-empirical aeropredictive code. Multi-objective optimization using the NSGA-II algorithm was performed on body-tail and body-canards-tail design configurations. The results show that the multi-objective optimization approach provides valuable information about the design space, and identified that the body-tail configuration was a trade-off between all of the objectives, each of which individually lead to different optimal configurations. Two general types of projectile category were identified, equating to highly maneuverable but less stable and slower, and less maneuverable but more stable and faster designs. The body-canards-tail design case showed significant differences to the body-tail case. In particular, it was determined from the Pareto fronts that an improvement in the lateral acceleration could be made in the baseline design without any negative consequences on the other objectives.