Title :
A statistical method for obtaining the factors in electronic-component reliability-prediction models
Author :
Yang, Zhongsen ; Raafat, Hazem
Author_Institution :
Dept. of Comput. Sci., Regina Univ., Saskatchewan, Sask., Canada
fDate :
12/1/1992 12:00:00 AM
Abstract :
According to US Mil-Hdbk-217, the failure rate of most electronic components can be predicted as λp=λb×π1 ×π2···πn. A statistical method of obtaining these parameters is presented. The method is based on large quantities of field data, and the more current the data are, the more accurate the values of each level of λb and πi. Using these values, the models in US Mil-Hdbk-217 can be applied to predict the life of electronic products. The method can also be used to verify the present values of each level of λb and πi, and to provide the basis of further amendment
Keywords :
failure analysis; reliability theory; statistical analysis; US Mil-Hdbk-217; electronic components; failure rate; life prediction; reliability prediction models; statistical method; Accuracy; Costs; Electronic components; Failure analysis; Laboratories; Least squares methods; Predictive models; Raw materials; Statistical analysis; Testing;
Journal_Title :
Reliability, IEEE Transactions on
