초록 일부
Fault detection and diagnostics (FDD) can be used to monitor the performance of air conditioners (ACs) and heat pumps (HPs), signal any departure from their optimal performance, and provide diagnostic...
더보기
초록 전체
Fault detection and diagnostics (FDD) can be used to monitor the performance of air conditioners (ACs) and heat pumps (HPs), signal any departure from their optimal performance, and provide diagnostic information indicating a possible fault if degradation of performance occurs. For packaged systems fully assembled in a factory, an FDD module can be fully developed for all units of a given model based on laboratory tests of a single unit. For field-assembled systems, laboratory tests of a representative AC or HP installation can lead to the development of a “back-bone” preliminary FDD algorithm; however, in situ adaptation of these algorithms is required because of installation variations in the field. This paper describes a method for adapting a laboratory-based FDD module to field-assembled systems by automatically customizing the in situ FDD fault-free performance correlations. We validated the developed data-clustering technique with a set of nearly 6000 data points to generate fault-free correlations for an HP operating in the cooling mode in our laboratory. The study evaluated several fault-free feature models and indicated that the use of different order correlations during stages of data collection produced better fits to the data.
더보기 닫기