ABSTRACT
Monitoring systems for rotating machines has been largely used in industry even with the high reliability achieved in turbines and compressors. These systems can reduce the number of non-scheduled shutdowns and time for the scheduled ones. In this way, it remains a challenge for researchers and industry to develop better monitoring, diagnosis and prognostic techniques. These systems generally use multiple sensors to analyze the machine behavior, so the Mahalanobis-Taguchi strategy (MTS), which is recognized for its suitability for multivariate data analysis, was applied in order to integrate these sensors. Also, in order to compare and validate MTS results, a more classic approach, based in vibration analysis is also considered. Therefore, the objective of this paper is to present some initial results in the unbalance estimation in rotating machines using this method. A centrifugal compressor used in an off-shore unit is selected for application. Actual data and computational simulations were used to validate the method and to provide the unbalance estimation of the selected machine.
