Propulsion life prediction based on support vector machine

Abstract

The advanced technology of the propulsion device is a national key research project, and PHM prediction and health management are among the core technologies. Our prediction of the remaining effective life (RUL) of the propulsion unit is a key part of the PHM technology. The data-driven method can effectively predict RUL. For example, the artificial intelligence method is widely used because it has certain advantages. Therefore, this paper uses machine learning algorithms to carry out RUL prediction research of propulsion devices. In addition, on the basis of support vector machine (SVM) regression theory, we combined the method of SBI (similarity-based interpolation), using Matlab program to build a prediction model. Both the practice samples and test samples of the model come from the engine simulation data, CMAPSS, provided by NASA. Finally, we use test samples to test the prediction model. The results show that the constructed model has good accuracy and robustness, and it can better predict the RUL of the propulsion unit.