A data-driven prognostics approach for RUL based on principle component and instance learning

Abstract

The research of Remaining Useful Life (RUL) estimation is one of the most common tasks of Prognostics and Health Management (PHM). This paper presents a data-driven approach for estimating RUL using principle component and instance learning. The approach is especially suitable for situations in which abundant run-to-failure (RtF) data are available. Firstly, the principal component analysis (PCA) is used to find the low-dimensional principal components (PCs) from the statistical features of the measured signals. Then, the health indicators (HI) can be obtained by using weighted Euclid distance (WED), and regressed by the data-driven methods or model-based methods. Finally, the method based on instance learning is employed to estimate the RUL of the machine under operation. The performance of the prognostics approach introduced in this paper is demonstrated by using turbofan engine degradation simulation data set, which is supplied by NASA Ames.