Development of an intelligent marine engines health assessment system based on digital twins and data driven models

The engine instantaneous torque is a critical measurement which can provide deep insight into the cylinders condition, can be cheaply obtained, and thus also allow for continuous engine monitoring, which is critical in identifying the onset of failures and degradation before incurring increased costs in fuel, spare parts, and vessel down time. This study presents the development of an intelligent engine health assessment system which relies on an engine digital twin, inclusive of both a thermodynamics and a crankshaft dynamics model to simulate the engine's instantaneous torque. Subsequently, the engine digital twin is utilised in tandem with the measured instantaneous torque to successfully identify frequently occurring malfunctioning conditions through an engine faults mapping approach. Furthermore, an even deeper insight to the cylinder's condition is gained, by reconstructing the in-cylinder pressure diagram for all cylinders using just the instantaneous torque.