Testability verification based on sequential probability ratio test method

Testability plays an important role in the readiness of equipment as a good design for testability (DFT) can greatly decrease the fault detection and isolation time, which will accelerate the maintenance actions. Testability verification is a procedure to check that whether the testability indexes such as fault detection rate (FDR) and fault isolation rate (FIR) meet the requirement in the contract. Currently, standards and statistical methods used in testability verification have the problems such as large sample, long period and so on. Sequential probability ratio test (SPRT) method can decrease the test sample size with almost a same operation characteristic as classical method based on binomial distribution. SPRT method and its truncated rules are introduced and the spectrum of expected test number is proposed. Then, the sample size allocation method and failure mode selection method based on failure rate used in sequential testability verification are illustrated. Testability verification of a control system is implemented with the given method and steps. Software named testability demonstration and evaluation system (TDES) which can calculate the decision criteria, plot decision chart, select failure mode and make judgment is used in the test as assistance. The result shows that the test sample size is remarkably decreased while comparing with the classical method.