Missile Health State Prediction Based on CA-RBF Neural Network

Abstract

In order to solve the problem of selecting missile health state evaluation indexes and the mapping relationship between the selected indexes and missile health state, a missile health state prediction method based on CA-RBF neural network is proposed. Firstly, the influencing factors of missile health state are analyzed through missile life profile, and the quantitative method is given. Then, the corresponding analysis (CA) method is used to screen the influencing factors of missile health state. Taking the selected factors and the evaluation results of missile health state as the input and output of neural network, the training samples of RBF neural network are established to predict the missile health state. Finally, an example is given to illustrate the practicability and effectiveness of the proposed method. This method can provide a new idea for the selection of missile health state indexes and missile health state prediction, and provide a basis for missile preventive maintenance decision-making.

Keywords

• |missile|rbf neural network|state prediction|correspondence analysis|index quantization