Projectile drag coefficient identification based on extreme learning

Aerodynamic parameters play a decisive role in the ballistic characteristics of projectile. How to accurately obtain aerodynamic projectile is an important task development process In order further improve identification accuracy drag coefficient, this paper generates huge data through numerical simulation and uses extreme learning method identify coefficient under three kinds noise conditions. The avoids iterative updating weights thresholds by randomly generating input threshold values hidden layer neurons overcomes problem long time traditional back propagation (BP) neural network algorithm. Based on least squares principle, Moore–Penrose generalized inverse matrix output was solved determine optimal weight network, then, identified. Comparing with BP method, results show that proposed has higher faster convergence speed can effectively which meet practical needs engineering.