Fault Prognosis of Key Components in HVAC Air-Handling Systems at Component and System Levels
Ying Yan, Peter B. Luh, Krishna R. Pattipati
Abstract
Fault prognosis of the air-handling systems, which are the key subsystems of heating, ventilation, and air conditioning systems, allows system operators to know the remaining useful life (RUL), thus preventing unexpected breakdowns and reducing the operational and maintenance costs. In this article, a new hidden semi-Markov model-based method is developed. In the method, only relevant state-transition points are selected and estimated, leading to computational efficiency. Physics-based models are used in a novel way to provide “mapping matrices” relating component capacities to fault severities, capturing impacts of multiple failure modes. Experimental results show that our method can effectively estimate the RUL of the components and the systems.