The determination method on products sample size under the condition of Bayesian sequential testing

In reliability engineering, especially in the context of reliability test design, it is important to determine how many product samples should be put into field test to evaluate product reliability level or verify whether the product reliability satisfy the predefined requirement in the development contract. In this paper, the determination method on sample size under the condition of Bayesian sequential testing is proposed. According to the sequential posterior odd testing (SPOT) method, the calculation on overall probability of decisions is firstly given out. Then the determination program of sample size used to test is presented by controlling probability level of correct decision-making. Finally, we make the determination of sample size on the mean of normal distribution under the condition of Bayesian sequential testing as an example to demonstrate the proposed determination process of sample size. The comparisons with classical methods are made to prove the effectiveness of the proposed method.