Reliability growth estimation for unmanned aerial vechicle during flight-testing phases

It is necessary for airplanes to be flight-tested during the development process, and they should pass the testing/failurefinding/improvement/re-testing reliability growth process during the flight-testing phases to ensure its reliability. However, due to airplane complexity and the high costs of flight-testing, the reliability growth testing is usually done with small samples. It is thus difficult to estimate the reliability growth during the flight-testing phases. In this paper, Bayesian method for binomial reliability growth based on the Dirichlet prior distribution is applied to reliability growth estimation, and the parameters of the posterior distribution are calculated by using the simulation method of Markov-Chain Monte Carlo. The method is applied to the Unmanned Aerial Vehicle test flight phases, and the example shows that the method based on the Dirichlet prior distribution can save the flight-testing time. It is easy to confirm the parameters of the prior distribution by using the prior information. The proposed method is suitable for reliability growth testing estimation during flight-testing stages.