An efficient operation of marine diesel engines, onboard ships, requires advanced monitoring and diagnostic techniques for early detection of faults and degradation in the propulsion or power generation system. This complex problem has been recently approached by digital-twin-based fault detection models. In this paper, we report on two methods for fault analysis on marine diesel engines exploiting (i) an Artificial Neural Network (ANN) combined with machine learning tools and (ii) a digital twin simulation model combined with a parameter estimator tool. In both cases, a digital twin model of the engine has been used for the generation of synthetic data, but in different simulation environments. These methodologies are applied to two distinct case studies, and their outcomes are discussed, focusing on the pros and cons. A proposal for a method combining the benefits of both is presented.
A Comparison Between Digital-Twin Based Methodologies for Predictive Maintenance of Marine Diesel Engine
Campora, Ugo
2024-01-01
Abstract
An efficient operation of marine diesel engines, onboard ships, requires advanced monitoring and diagnostic techniques for early detection of faults and degradation in the propulsion or power generation system. This complex problem has been recently approached by digital-twin-based fault detection models. In this paper, we report on two methods for fault analysis on marine diesel engines exploiting (i) an Artificial Neural Network (ANN) combined with machine learning tools and (ii) a digital twin simulation model combined with a parameter estimator tool. In both cases, a digital twin model of the engine has been used for the generation of synthetic data, but in different simulation environments. These methodologies are applied to two distinct case studies, and their outcomes are discussed, focusing on the pros and cons. A proposal for a method combining the benefits of both is presented.File | Dimensione | Formato | |
---|---|---|---|
1 SPEEDAM 2024-MPS3617[343]-A Comparison between Digital-Twin Predictive Maintenance Diesel Engine-Final.pdf
accesso aperto
Tipologia:
Documento in versione editoriale
Dimensione
855.04 kB
Formato
Adobe PDF
|
855.04 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.