Optimal condition-based maintenance decisions for systems considering dependent stochastic degradations of components

This paper describes a probabilistic model to determine the optimal inspection interval for an engineered system containing multiple components that are subjected to dependent stochastic deteriorations. To this end, we employ the homogeneous gamma process to model the deterioration of the structural components and the Gaussian copula function to characterize the dependency of the deteriorating processes associated with different components. Following that, the expectation of the cost rate is formulated based on a particular maintenance policy and evaluated using the Monte Carlo simulation technique. The minimum expected cost rule is then used to determine the optimal inspection interval. An example dealing with the heat transport piping system of nuclear power plant is employed to illustrate the developed probabilistic methodology. The impact of the dependency of the degradation processes on the selection of the optimal inspection interval is investigated through the parametric analysis and also included in the example.