论文部分内容阅读
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。
An exploratory research on intelligent information processing of fuze based on neural network is introduced. The image recognition system based on unsupervised feature extraction is established by BP algorithm, including three-layer network automatic associative machine and two-layer network information extractor, which can realize the training, memory and classification of input image. The results show that this method is feasible and partly reflects the intelligence of information processing.