前馈多层神经网络BP算法与可靠性增长模型

BP Algorithms of Feed-Forward Multi Layer Neural Network and the Model of Reliability Growth

本文叙述了可靠性增长和 BP算法的基本概念 ,着重研究神经网络 BP算法用于可靠性增长的预测方法。Gompertz模型是可靠性增长的一个很好的预测模型。本文引用文献中的多个实例 ,将BP算法预测结果与 Gompertz模型预测结果相比较 ,结论基本上一致。这说明该方法不但可行 ,而且有简便。

Reliability growth prediction is the most important part of reliability engineering.The artificial neural network is one of the recently hot subjects.This paper discusses reliability growth and BP algorithms,and then considers emphatically its application in reliability growth prediction.The Gompertz model is a better prediction model of reliability growth.The paper examines the algorithm by some instances and compares with the results of the Gompertz model.The results are almost the same.It shows the approach is feasible and effective,and has special features:simple and adaptive.[WT5HZ]