神经网络在引信智能信息处理中应用的研究

Application of Neural Network in the Intelligent Information Processing for Use in Fuzes

介绍了应用神经网络进行引信智能信息处理的探索性研究。采用BP算法建立了无监督特征提取的图象识别系统,包括三层网络自动联想机与两层网络信息提取器,实现了对输入图象的训练、记忆与分类。研究结果表明该方法技术可行,在一定程度上体现了信息处理的智能性。

Recognition and processing of graphic patterns using artificial neural network (ANN) technique is introduced. A pattern recognition system of unsupervised feature extraction based on BP algorithm is established, including an automatic associative processor with 3 layers of network and an information extractor with 2 layers. Training, recognition and classification of input patterns are realized in the system.