人工神经网络 BP 网的应用

ANN BP Networks and Their Applications

目的研究ＢＰ网在人工神经网络领域中的应用及其特点．方法采用误差反向传播学习算法，对不同的声纳回波进行分类处理．结果建立了３层ＢＰ网络结构的声纳信号处理系统．结论隐蔽层单元增加有益于正确分类率的提高，但连接权计算复杂度增加．

Aim To study the features of BP networks and its applications in ANN Methods Adopting the error back propagation learning algorithm,different back waves in sonar system are processed and classified Results A three layer BP structure system is made up to process the sonar signals Conclusion Increase in the number of hidden layers,heightens the rate of classification,but the algorithm of the link weighting will become more complex