Fault detection and diagnosis of marine diesel engines: A systematic review of fault taxonomies, data acquisition and generation pathways, and diagnostic approaches

This article provides a comprehensive synthesis of the current state of research on fault detection and diagnosis of marine diesel engines, organized along three main threads: fault taxonomy, data acquisition and generation pathways, and diagnostic approaches. Firstly, guided by the combustion functional chain and key components, a hierarchical modeling of fault modes is developed to construct a “system–component–fault” fault taxonomy. Secondly, fault-injection and parameter-equivalent simulation strategies are systematically consolidated, revealing a structural landscape in which bench-test data with controllable fault injection constitutes the primary evidence base, simulation-generated data serves as important complements, while in-service data remains underrepresented and public datasets are scarce. Subsequently, the methodological landscape of marine diesel engine diagnosis and its evolutionary trends are summarized, showing that data-driven approaches dominate, model-based and hybrid approaches provide critical support, and knowledge-based approaches play a supplementary role. Finally, future research directions for this field are discussed.