Open Access

ARTICLE

A New Method for Maintenance Management Employing Principal Component Analysis

Fausto Pedro García Márquez¹

1 Ingenium Research Group, Universidad Castilla-La Mancha, Campus Universitario s/n, 13071 Ciudad Real, Spain. FaustoPedro.Garcia@uclm.es

Structural Durability & Health Monitoring 2010, 6(2), 89-100. https://doi.org/10.3970/sdhm.2010.006.089

Abstract

This paper presents a simple graphic method for detecting and classifying faults in point mechanisms based on the study of some statistical parameters of the force and current signals of the point machine. Principal Components Analysis (PCA) employed in order to reduce the number of these parameters. PCA is utilised in this paper for modifying the parameter dataset, and reducing the coordinate system by linear transformation. It is then possible to plot the new coordinate system in 2 or 3 dimensions, where the faults can be detected and identified. In this work most of the faults could be detected, but only a few experiments could be identified.

---

Keywords

---