Local preserving projections-based feature selection and Gaussian mixture model for machine health assessment

Abstract

The sensitivity of various features that are characteristics of machine health may vary significantly under different working conditions. Thus, it is critical to devise a systematic feature selection (FS) approach that provides a useful and automatic guidance on choosing the most effective features for machine health assessment. This article proposes a locality preserving projections (LPP)-based FS approach. Different from principal component analysis (PCA) that aims to discover the global structure of the Euclidean space, LPP can find a good linear embedding that preserves local structure information. This may enable LPP to find more meaningful low-dimensional information hidden in the high-dimensional observations compared with PCA. The LPP-based FS approach is based on unsupervised learning technique, which does not need too much prior knowledge to improve its utility in real-world applications. The effectiveness of the proposed approach was evaluated experimentally on bearing test-beds. A novel machine health assessment indication, Gaussian mixture model-based Mahalanobis distance is proposed to provide a comprehensible indication for quantifying machine health state. The proposed approach has shown to provide the better performance with reduced feature inputs than using all original candidate features. The experimental results indicate its potential applications as an effective tool for machine health assessment.