A Data Clustering Technique for Fault Detection and Diagnostics in Field-Assembled Air Conditioners

Published

May 14, 2018

Author(s)

William V. Payne, Piotr A. Domanski, Jaehyeok Heo

Abstract

Fault Detection and Diagnostics (FDD) can be used to commission and maintain the performance of residential style, split-system, air conditioners and heat pumps. Previous investigations by various researchers have provided techniques for detecting and diagnosing all common faults, but these techniques are ideally applied to systems that have been thoroughly tested in a carefully instrumented laboratory environment. Although such an approach can be applied to factory-assembled systems, installation variations of field-assembled systems require in-situ adaptation of FDD methods. Providing a workable solution to this problem has been the impetus for this work which describes a method for adapting laboratory style FDD techniques to field installed systems by automatically customizing the FDD fault-free performance models for random installation differences.

Citation

International Journal of Air-Conditioning and Refrigeration

Volume

26

Pub Type

Journals

Download Paper

DOI Link

Keywords

Air-Conditioner Fault, ANN Model, Fault Detection and Diagnosis, Heat Pump Fault, MPR Model, Split-System Fault

Citation

Payne, W. , Domanski, P. and Heo, J. (2018), A Data Clustering Technique for Fault Detection and Diagnostics in Field-Assembled Air Conditioners, International Journal of Air-Conditioning and Refrigeration, [online], https://doi.org/10.1142/S2010132518500153 (Accessed October 9, 2024)