Abstract
The same as for planning in different categorize, maintenance planning deals with probabilistic parameters, due to uncertainty in all cases. One of the probabilities which maintenance planning is facing is related to the machines and components failure. These kinds of failure are completely accidental. Hence, involving a reliability prediction model for these uncertain cases would magnetize predictive values. In this study, based on the Adaptive Neuro-Fuzzy Inference System, a model for prediction of different systems reliability is developed. Based on the outputs of this prediction model for a case study (failures of electromotor in south pars gas Complex), it is demonstrated that the errors of predicted values are less than those for non-Fuzzy models.
Keywords
Reliability, Prediction, Nuero-Fuzzy networks