Evaluation of lubricant condition and engine health based on soft computing methods

Evaluating the condition of the lubricant plays an influential role in the maintenance engineering of mechanical systems. This study has several objectives and finally offers a method based on the neural network. That method uses more limited indicators to determine the condition of the lubricant and the engine's health. That will reduce testing costs and encourage equipment and machine owners to perform lubricant analyses. The 681 results of the engine lubricant spectral analysis had used. The statistical analysis results showed; out of twelve indexes, only seven indexes, including iron, chromium, lead, copper, aluminum, nickel, and TDPQ, could have been influential in determining the three conditions of normal, caution, and critical wear. Soft computing methods, including KNN and RBF-ANN, were used to diagnose engine health conditions in three classes of normal, caution, and critical based on seven indexes of engine lubricant in three sizes of the training data set includes 40, 60, and 80%. The results showed that the engine health diagnosis accuracy by KNN of the training set sizes of 80, 60, and 40% was equal to 99.71, 98.38, and 97.36%, while detection accuracy of the RBF-ANN for all three training set sizes was approximately 99.85%. Also, the sensitivity analysis results showed soft computing methods could have the high ability to diagnose engine health. Graphical abstract