Forecasting model for degradation path and parameter estimation based on neural network

Traditional life evaluation theory established on the basis of mass failure data, the phenomena of little or naught failure put forward challenges for existed life evaluation theory. The performance degradation data provide useful information for products' reliability and gives feasible way for products' life evaluation. The limitations of existing degradation models are analyzed, a new forecasting model and parameter estimation method based on neural network is brought forward. By using back propagation neural network(BPNN), the nonlinear degradation path of product performance can be got, and the parameters can be estimated by self-adaptive neural network. An example is given out to validate the effectiveness of the method and compared with existing model.

[1]  William Q. Meeker,et al.  Statistical tools for the rapid development and evaluation of high-reliability products , 1995 .

[2]  V. Crk,et al.  Reliability assessment from degradation data , 2000, Annual Reliability and Maintainability Symposium. 2000 Proceedings. International Symposium on Product Quality and Integrity (Cat. No.00CH37055).

[3]  Enrico Zio,et al.  A neuro-fuzzy technique for fault diagnosis and its application to rotating machinery , 2009, Reliab. Eng. Syst. Saf..

[4]  Martin T. Hagan,et al.  Neural network design , 1995 .

[5]  Sanjeev Arora,et al.  Artificial neural network models for predicting degradation trends in system components and sensors , 2003, Proceedings AUTOTESTCON 2003. IEEE Systems Readiness Technology Conference..

[6]  J.S.R. Jayaram,et al.  Reliability prediction through degradation data modeling using a quasi-likelihood approach , 2005, Annual Reliability and Maintainability Symposium, 2005. Proceedings..

[7]  C. Joseph Lu,et al.  Using Degradation Measures to Estimate a Time-to-Failure Distribution , 1993 .