Abstract:
Hypersonic flow characteristics such as high temperature real gas effect, viscous interference effect, Mach number effect and scale effect, do not follow the similarity simulation criterion of experimental gas dynamics, which makes the hypersonic flow phenomenon beyond the range that can be accurately predicted by the classical gas dynamics theory. How to use the ground experimental data to predict the flight state, that is, the problem of ground-to-flight (G2F) correlation, is the key scientific problem that restricts the development of new aerospace vehicles. This paper summarizes the latest research progress of G2F correlation, and focuses on the multi-space correlation theory and correlation method of intelligent functional optimization. According to the theory, from a high-dimensional point of view, the experimental results of different wind tunnels are intrinsically related, and the flight test can be regarded as an ideal wind tunnel experiment. Based on the experimental data of wind tunnel groups (different types of wind tunnels which can simulate different parameter sections), using a specialized intelligent learning algorithm in functional space, the correlation method is performed by starting from the high-dimensional full parameter space. And then by carrying out a series of dimension reduction and adaptive space transformation, the invariant law is automatically deduced that different wind tunnels abide by together, so as to get a formula for G2F correlation. The results of verification examples and preliminary applications show that the multi-space correlation theory and functional intelligent optimization correlation method are effective, which would be a new trend in the research of G2F correlation for hypersonic aerodynamic force and heating.