Analysis of flight performance using neural networks

The analysis of flight performance is a necessary process in system concept design for aircrafts. A key problem of the analysis is to determine the relationship between the design parameters and the performance parameters, which generally relies on a lot of time-consuming simulations and even expensive experiments. In order to improve analysis efficiency and save money, the neural network modeling based on uniform design is employed to determine the relationship models. A relative fewer input-output data are required to training the models. By the models the influence characteristics of the design parameters on the performance can be analyzed rapidly without extensive simulations and experiments. The approach is tested taking the low performance Common Aero Vehicle (CAV-L) as an example. The results show that neural networks are effective in the analysis of flight performance for aircrafts.