RADIS: A real-time anomaly detection intelligent system for fault diagnosis of marine machinery

作者：

Highlights：

• Data-driven models can empower strategies in relation to O&M activities.

• The need of developing a real-time intelligent system to promote smart maintenance.

• RADIS, a DLSTM-based VAE NN in tandem with image thresholding, is presented.

• A diesel generator of a tanker ship is considered as a case study.

• RADIS can detect an average of 92.5% of anomalous instances.

摘要

•Data-driven models can empower strategies in relation to O&M activities.•The need of developing a real-time intelligent system to promote smart maintenance.•RADIS, a DLSTM-based VAE NN in tandem with image thresholding, is presented.•A diesel generator of a tanker ship is considered as a case study.•RADIS can detect an average of 92.5% of anomalous instances.

论文关键词：Anomaly detection,Smart maintenance,Ship systems,Marine machinery,Deep learning,Intelligent real-time systems

论文评审过程：Received 19 August 2021, Revised 16 May 2022, Accepted 19 May 2022, Available online 26 May 2022, Version of Record 28 May 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117634