Abstract:

The condition based maintenance decision modeling of deteriorating systems is focused on. Considering the economic factors and system availability, two general maintenance optimization models are given. The gamma distribution is used to model the system degradation, and monitoring data are used to estimate the parameters. Then based on the deteriorating model, historical maintenance information and monitoring data are used to make certain variables in decision models. For the maintenance effects and maintenance time of the main impacts in maintenance actions, geometric distribution models are given to model maintenance effects with and without human factors respectively, which is considered as the random variable of system condition and maintenance numbers, and an interpolation method is used to construct the distribution function of maintenance time. Finally, the decision model is applied to a control system, and the comparison results of the decision model between ideal and random variable situation demonstrate the efficiency and practicability of the proposed models.