【作者】 史彩成； 赵保军； 韩月秋；

【Author】 Shi Caicheng,Zhao Baojun and Han Yueqiou (Department of Electronic Engineering,Beijing Institute of Technology)

【机构】 北京理工大学电子工程系；

【摘要】 本文利用自适应神经网络对电子系统可靠性增长和失效密度分布进行拟合。对电子系统失效时间序列进行分析,可以发现时间序列存在混沌吸引子。根据时间序列的嵌入维,确定神经网络的输入数,建立三层BP神经网络。分析和仿真结果表明:此方法比其它方法有更好的拟合和预测结果。更多还原

【Abstract】 An adaptive neural network model of reliability growth and invalidation density of electronic system is presented in this paper.Invalidation sequences of electronic system are analyzed,it can be discovered that the sequences have chaotic attractor.Input number of the neural network is established with inset-dimensions of the sequences,the adaptive neural network is built in the form of three layers.It is shown through analysis and simulation that the adaptive neural network model is better than others.更多还原