Predicting uncertainty of a chiller plant power consumption using quantile random forest: A commercial building case study

Abstract

Selecting an appropriate confidence level for chillers models could lead to better prediction of the peak electricity load, to more efficiently participate in demand response programs, and to manage the operation and performance of chiller plants. Therefore, this study aims to predict the power consumption of a chiller plant in a high-rise commercial building based on Quantile Random Forest (QRF). This study proposed a simplified data-driven approach to compute the lower and upper bounds of the prediction intervals (PIs) of power consumption for a chiller plant, consisting of four chillers. An evaluation of the proposed data-driven approach using measured datasets and computed PIs metrics demonstrates that the PI coverage probability for the samples in test datasets for all chiller models with the confidence level of 80% and 90%, are above 80% and 90%, respectively. Also, the results show that by increasing the confidence level from 80% to 90%, the upper and lower bounds of the demand charge differ from the actual value by a factor of 3.3 and 1.7 times greater, respectively. This approach has implications for developing sequences of operations for chiller plants in building automation systems for evaluating past performance and for achieving future optimization of these plants.