Investigation on intelligent prediction technology of aerodynamic characteristics of multi-body separation
The next-generation supersonic or hypersonic fighter aircraft may use the concept of aft ejection separation(AES)for the internal weapon system.The AES belongs to a type or form of multi-body separation,not only has the flow and aerodynamic interference between the multi-body aircraft,but also has the relative motion between the multi-body aircraft,which seriously affects the separation safety of the aircraft.In this paper,an internal wind tunnel captive trajectory system is developed and integrated in the wind tunnel to obtain the aerodynamic and motion variation laws under typical conditions,and the artificial intelligence algorithm based on comprehensive random decision tree is used to preliminarily predict the aerodynamics of the aft ejection separation of the internal weapon.The results show that the optimization algorithm of nonlinear model regression technology based on comprehensive random decision tree has a certain predictability in the sample space of Mach numberMa=2.0,angle-of-attack(-5°≤αA≤ 5°)and sideslip angle(-3°≤β≤ 8°)for tail separation test.
internal weaponaft ejection separationaerodynamic interferenceartificial intelli-gencecomprehensive random decision tree
NETL
NSTL
万方数据
