摘要：AbstractPredictive maintenance (PdM) is an open issue in industrial and manufacturing systems. Remaining Useful Life (RUL) estimation is a significant concern in PdM, which corresponds to the engine’s remaining time to perform its intended function. This work proposes Fuzzy Cognitive Maps (FCMs) as a Health Indicator (HI) prognostics method for engines’ RUL prediction. For the FCM model design, we apply, test, and compare four different training algorithms to determine the model parameters’ values that lead to the optimal prediction accuracy. In order to test the results obtained from the FCM-based approach, we compare them with those derived from an Artificial Neural Network (ANN), and we verify that the use of FCMs results in a decent engine’s health status predictor comparable to a machine learning method.
