신경회로망 기반 유도탄의 종말속도 추정기법

초록

In this study, a method of estimating the terminal speed of a missile based on the neural network technique when a guidance law for the missile is specified. The data set of the terminal speeds corresponding to randomly selected initial conditions for learning neural network is produced via 3-DOF simulations. Once the learning process of the neural network is done, we obtain weighting matrices and biases of the neural network. Through simple matrix calculations, we can produce the terminal speed of the missile is dependent on the initial condition, we can control the terminal speed by changing the initial condition. The proposed method can be widely applied to make a algorithm for terminal speed estimation under a terminal guidance law is specified.