In this paper, we propose methods of handling,analyzing, and profiling monitoring data of energy systemsusing their thermal coefficient of performance seen in unevensegmentations in their time series databases. Aside fromassessing the performance of chillers using this parameter,we dealt with pinpointing different trends that this para-meter undergoes through while the systems operate. Fromthese results, we identified and cross-validated with domainexperts outlier behavior which were ultimately identified asfaulty operation of the chiller. Finally, we establish correla-tions of the parameter with the other independent variablesacross the different circuits of the machine with or withoutthe observed faulty behavior.


Malinao, J., Judex, F., Selke, T., Zucker, G., Caro, J., & Kropatsch, W. (2015). Pattern mining and fault detection via textitCOP_textittherm-based profiling with correlation analysis of circuit variables in chiller systems. Computer Science - Research and Development, 31(1–2), 79–87. https://doi.org/10.1007/s00450-014-0277-5