Validation of an algorithm for predicting the remaining useful life, for a model with linear degradation

Abstract

This paper aimed to validate a working tool, component of the Predictive Maintenance Toolbox ™, produced by Matlab (MathWorks), in the case of a procedure for monitoring the operation of mechanical systems, in order to diagnose a failure of the process and to estimate the remaining useful life (RUL). This toolbox provides toolsets, materialized in function files, for labeling data, designing condition indicators, and estimating a parameter named the remaining useful life of a machine. You can analyze and label machine data imported from local files, cloud storage, and distributed file systems. The algorithm suggested by Matlab (software owned by MathWorks) was used in detail to process part of the data set provided freely by NASA through The Prognostics Data Repository, The Prognostics Center of Excellence (PCoE) at Ames Research Center. Of the 4 data sets, only one was used for this paper. Each data set is composed of 3 working files, in text format, for training, test and algorithm validation, and solution statement, respectively. The results obtained confirm the validity of the computer-assisted training system, diagnostics, prognosis, and validation tools, on a statistical basis, in the case of consistent databases.