Abstract:

Binomial distribution is preferred in testability demonstration, but it is discrete. Because of calculation difficulty, the sampling plan is usually worked out with some adaptations. Through the research of binomial probability distribution function(PDF), it has been proved that the PDF is a monotone increasing function of probability p and allowable failure time r in test, and is a monotone decreasing function of test time n. With these conclusions, an accurate algorithm is presented to solve the simultaneous inequality of binomial PDF. An optimum solution can be worked out by adopting multivariate half\|folded searching algorithm, so that one can get the sampling plan of testability demonstration accurately and quickly. At last a software package of testability demonstration is developed, it better meets the requirements of testability demonstration, its calculation results agree with IEC1123, and its algorithm can also be applied to sampling plan of Poisson PDF.

Key words: testability demonstration, sampling plan, binomial distribution, software package