Skip to Main content Skip to Navigation

Diagnostic vibratoire autonome des roulements

Abstract : The industrial and transportation sectors require more and more efficient and complex machines and installations increasing the risk of failure and disruption. This can lead to the immediate shutdown of a machine and disrupts the proper functioning of the entire production system. The diagnosis of industrial machines is essentially based on the monitoring of symptoms related to different degradation conditions. These symptoms can be derived from various sources of information, including vibration and acoustic signals. Nowadays, many effective techniques are well established, based on powerful tools offered by the theory of cyclostationary processes. The complexity of these tools requires an expert to use them and to interpret the results based on his/her experience. The continuous presence of the expert is expensive and difficult to achieve in practice. Condition indicators for rotating machines exist in the literature but they are conceived under the assumption of perfect operating conditions. They are limited, dispersed and generally not supported by theoretical frameworks. The main objective of this thesis is to reduce the use of human intervention by proposing strategies to design two optimal indicators that summarize diagnostic information into a scalar value. A distinction is made between two families in diagnosis: the case where prior information on the faults is known and the case where it is unknown. These indicators are designed to be used in an autonomous process without requiring human intervention, using statistical hypothesis tests. The capacity of these indicators is validated on real data and compared with other indicators from the literature in terms of detection performance.
Complete list of metadatas

Cited literature [117 references]  Display  Hide  Download
Contributor : Abes Star :  Contact
Submitted on : Friday, July 17, 2020 - 6:26:16 PM
Last modification on : Saturday, July 18, 2020 - 3:38:00 AM


Version validated by the jury (STAR)


  • HAL Id : tel-02902122, version 1


Souhayb Kass. Diagnostic vibratoire autonome des roulements. Vibrations [physics.class-ph]. Université de Lyon; Université libanaise, 2019. Français. ⟨NNT : 2019LYSEI103⟩. ⟨tel-02902122⟩



Record views


Files downloads