Development and Validation of a Data-Driven Fault Detection and Diagnosis System for Chillers Using Machine Learning Algorithms

Fault detection and diagnosis (FDD) systems enable high cost savings energy that could have economic environmental impact. This study aims to develop validate a data-driven FDD system for chiller. The uses historical operation data capture quantitative correlations among variables. evaluated the effectiveness robustness of eight classification methods based on experimental chiller (the ASHRAE 1043-RP project). training used is classified into four cases. Moreover, true false positive rates are characterize performance methods. results show local fault not significantly sensitive data, shows accuracy all has significant effect amount severity levels accuracy.