The BP neural network algorithm and the model of the reliability growth predication

The reliability growth prediction is the most important part of the reliability engineering. This paper considers the backpropagation (BP) algorithm application in reliability growth prediction. The Gompertz model is a good reliability growth model. The paper examines the BP algorithm and compares it with the Gompertz model. The results show that they were both basically consistent. It indicates that the BP approach is feasible, effective, simple and adaptive. This paper also discusses the prediction of the reliability storage in the weapon system using the BP algorithm