A statistical approach for condition monitoring for maintenance management in railway

This paper presents an approach for detecting and identifying faults in railway infrastructure components. The method is based on pattern recognition and data analysis algorithms. Principal Component Analysis (PCA) is employed to reduce the complexity of the data to two and three dimension. PCA involves a mathematical procedure that transforms a number of variables, which may be correlated, into a smaller set of uncorrelated variables called "principal components". In order to improve the results obtained, the signal has been filtered. The filtered has been carried out employing a State Space system model, estimated by Maximum Likelihood with the help of the well-known recursive algorithms Kalman Filter and Fixed Interval Smoothing. The models explored in this paper to analyse system data fits within the so called Unobserved Components class models.