Optimal test-times for intermittent faults

Su et al. (1978) considered continuous and repetitive tests for a continuous-parameter Markov model with intermittent faults. Periodic tests for intermittent faults are scheduled at times k·T (k=1, 2,...). This paper presents a simple algorithm to compute the optimal time to minimize the mean cost until detection when the test model is imperfect. First an upper bound is found for the optimal time. Then a bisection-algorithm is used to minimize the cost of detecting faults for a system in which faults are intermittent and unpredictable. Using this algorithm, the solution of example 1 is better than that of Nakagawa and Yasui (1989) by at least 10%. This algorithm can be more useful than the Newton-Raphson method to locate an optimum because Newton-Raphson involves the first derivative whereas the bisection method does not