Fig. 14From: Deep learning for inverse design of low-boom supersonic configurationsConvergence process of the optimized objective functionBack to article page