Credibility Evaluation of Missile Flight Simulation Model

In this paper, credibility evaluation issue of missile flight simulation model is studying by applying neural network technique. Aiming at the subsistent insufficiency of model validation method in application, we present a credibility evaluation method based on neural network. Which uses the powerful ability of nonlinearity mapping of neural network by utilizing missile flight state data of simulation model as training set to train neural network firstly, ascertain the accurate degree of models by the value of the output(reliability series)of trained neural network when flight test data of missile act on the trained neural network then, thereby giving the credibility evaluation to the models. Validation simulation results of a certain type of missile flight simulation model lastly further show the feasibility and validity of this method.