Condition-based monitoring (CBM) has advanced to the stage where industry is now demanding machinery that possesses\r\nself-diagnosis ability. This need has spurred the CBM research to be applicable in more expanded areas over the past decades.\r\nThere are two critical issues in implementing CBM in harsh environments using embedded systems: computational efficiency\r\nand adaptability. In this paper, a computationally efficient and adaptive approach including simple principal component analysis\r\n(SPCA) for feature dimensionality reduction and K-means clustering for classification is proposed for online embedded machinery\r\ndiagnosis. Compared with the standard principal component analysis (PCA) and kernel principal component analysis (KPCA),\r\nSPCA is adaptive in nature and has lower algorithm complexity when dealing with a large amount of data. The effectiveness of\r\nthe proposed approach is firstly validated using a standard rolling element bearing test dataset on a personal computer. It is then\r\ndeployed on an embedded real-time controller and used to monitor a rotating shaft. It was found that the proposed approach\r\nscaled well, whereas the standard PCA-based approach broke down when data quantity increased to a certain level. Furthermore,\r\nthe proposed approach achieved 90% accuracy when diagnosing an induced fault compared to 59% accuracy obtained using the\r\nstandard PCA-based approach.
Loading....