This paper presents a data-driven methodology for monitoring and quantifying the degradation of plastic chain conveyors, a relevant asset in industrial automation. The proposed approach leverages system vibrations and employs the synchrosqueezing Short-Time Fourier Transform and convolutional autoencoders to generate a compact representation space for the data. This space enables the construction of control charts to monitor the extracted metrics. The method is intended to provide a comprehensive assessment of system degradation. Applied to a plastic conveyor chain on a dataset collected over a four-month period, the methodology seeks to identify and quantify the two most significant degradation mechanisms: the chain’s elongation due to joint wear and the wear and tear of the slide rail. This research intends to address a significant gap in the literature, offering a practical and automatic solution for condition monitoring based on vibration for industrial equipment. Two aspects will be considered for the evaluation of the method: the capability to “reduce” the storage usage (dimensionality reduction) and the anomaly detection capabilities of the system.

Application of synchrosqueezing transform and autoencoders for monitoring of production systems. A case study on plastic chain conveyor systems

Radicioni L.;Bono F. M.;Cinquemani S.
2024-01-01

Abstract

This paper presents a data-driven methodology for monitoring and quantifying the degradation of plastic chain conveyors, a relevant asset in industrial automation. The proposed approach leverages system vibrations and employs the synchrosqueezing Short-Time Fourier Transform and convolutional autoencoders to generate a compact representation space for the data. This space enables the construction of control charts to monitor the extracted metrics. The method is intended to provide a comprehensive assessment of system degradation. Applied to a plastic conveyor chain on a dataset collected over a four-month period, the methodology seeks to identify and quantify the two most significant degradation mechanisms: the chain’s elongation due to joint wear and the wear and tear of the slide rail. This research intends to address a significant gap in the literature, offering a practical and automatic solution for condition monitoring based on vibration for industrial equipment. Two aspects will be considered for the evaluation of the method: the capability to “reduce” the storage usage (dimensionality reduction) and the anomaly detection capabilities of the system.
2024
MESA 2024 - 20th International Conference on Mechatronic, Embedded Systems and Applications, Proceedings
autoencoders
condition monitoring
conveyors
plastic chains
synchrosqueezing transform
vibrations
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/1282086
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? ND
social impact