Determination method of acceleration factor based on life distribution and Bayes

The acceleration factor is difficult to determine for small sample products, a determination method of acceleration factors, based on life distribution and Bayes, is proposed. First, according to the stored data information of similar products within the field under normal stress, obtain the value of the distribution parameters of the product life by maximum likelihood estimation (MLE), then use the Fisher information matrix inversion to obtain its variance-covariance matrix, and eventually get the distribution of life distribution parameters. Then, based on the testing information of small sample researching products under accelerated stress, combine some priori information, such as the accelerated stress test information of similar products and expert experience, and use the ML-II method for determining the weight of a mixed prior distribution. The posterior distribution of the distribution parameters of product life can be achieved by the Bayes method. Finally, by combining with distribution information distribution parameters of the product life under the normal stress and accelerated stress, acquire the acceleration factor. Taking an accelerometer as an example, the correctness and the validity of the proposed method are proved.