Abstract—Fault Detection and Diagnosis (FDD) are vital for maintaining the energy efficiency and reliability of centrifugal chillers. This study proposes a hybrid data-driven approach that combines Radial Basis Function (RBF) regression and Gaussian Process Regression (GPR) to develop an accurate residual-based reference model. Deviations in thermodynamic parameters are continuously tracked using an Exponentially Weighted Moving Average (EWMA) control chart to detect condenser fouling and refrigerant leakage at multiple severity levels. The proposed RBF_GPR_EWMA framework was validated using the ASHRAE RP-1043 dataset and real chiller data obtained from a hospital in Ho Chi Minh City. The results demonstrate high prediction accuracy (R2 > 0.99) and reliable detection of early-stage performance degradation. The proposed framework does not require fault labels or complex feature design, offering robustness and interpretability. The simplicity and adaptability of the framework make it suitable for integration into building management systems to support condition-based maintenance and energy-efficient operation of heating, ventilation, air conditioning, and refrigeration equipment.

Keywords—Fault Detection and Diagnosis (FDD), Exponentially Weighted Moving Average (EWMA) control chart, radial basis function, Gaussian process regression, ASHRAE RP-1043 dataset

Cite: Nguyen Hoang Son, Dinh Anh Tuan Tran, and Huong Nguyen Thi Cam, "A Hybrid RBF_GPR_EWMA Model for Data-driven Fault Detection in Centrifugal Chiller Systems," International Journal of Mechanical Engineering and Robotics Research, Vol. 15, No. 4, pp. 391-403, 2026. doi: 10.18178/ijmerr.15.4.391-403

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).