Based on the parameterization study of the axisymmetric basic flowfield

a variable center body with properly small radius was selected

then the sensitivity analysis of other design parameters was conducted to obtain the influence law of design parameters on the overall performance of the basic flowfield. This showed that the coefficient c had the most obvious effect

and the coupling effect between various design parameters was also significant. The sample database was used to construct the corresponding neural network approximate model

and the three-objective optimization of the axisymmetric basic flowfield at Mach number of 6 was carried out in combination with the neighborhood cultivation multi-objective genetic algorithm. After the optimization

the internal compression ratio in the basic flowfield reduced by 17.7%

the total pressure recovery coefficient increased by 2.3%

and the static pressure ratio increased by 7.1%. Based on the optimization results

the inward turning inlet profile was designed. The results from its viscous numerical simulation at Mach number of 3-6 show that the optimized inward turning inlet can start normally at Mach number of 3

and has higher compression

better flow capture ability and total pressure recovery performance at Mach number of 4-6.