舰船全磁场数据融合的炸点精确选择

介绍了一种鱼雷磁引信炸点精确选择技术，即基于舰船全磁场的三个分量，利用神经网络将三分量数据进行序贯融合处理，从而进行了对舰船龙骨部位（鱼雷引信炸点）的精确选择，文中给出了处理方法及检验结果，分析了提高网络推广性能的措施，并与常归特征提取的炸点选择精度进行了比较。

Accurate Burst-point Selection Based on the Data Fusion of the Ship Magnetic field

This paper provides a method used to select a burst point accurately for a torpedo magnetic fuze. With the method, the fuze can accurately select the keel of a ship, which usually is the burst point for the torpedo fuze, through using a neural network to fuse the three components of a ship magnetic field. The paper gives a processing method and inspection results, analyzes the approaches to improve network generalization capability, and compares the accuracy of the burst point selection of the method used in this paper with the one of common characteristics extraction.