Abstract: A new prediction method based on radial basis function (RBF) neural network and an unscented Kalman filter technology is proposed for the precise and quick prediction of impact-point without exact flight data. Firstly, RBF neural network approximated external ballistics equation is used to predict the projectile impact-point, and the improved quantum-behaved particle swarm optimization algorithm is used to optimize the training method. On this basis, the tentative prediction data is processed with unscented Kalman filter. At last, the prediction simulation is carried out. The results show that a high prediction precision can be reached under the condition of input data with noise. The method proposed in this paper is efficient and available for impact-point prediction.

Key words: ordnance science and technology, radial basis function neural network, particle swarm optimization, unscented Kalman filter, impact-point prediction