A robust maintenance management of transformers considering uncertainties in fault data

In this work it is performed a robust management of reliability centered maintenance in transformers, considering uncertainties associated with fault and repair data (transition rates of Markov models). The uncertainties are propagated to reliability indexes using probabilistic techniques such as Monte Carlo Simulation and Latin Hypercube Sampling. In addition, these techniques are used to perform a robust adjustment in transformers inspection interval through maximizing reliability indices. This optimization pattern makes reliability indexes immunized against variations in Markov models due to these uncertainties.