Year: 2013
Type: article
Abstract: This paper presents the results of an energy\npredictor that predicts the energy demand of\ncommercial buildings using Case Based Reasoning\n(CBR). The proposed approach is evaluated using\nmonitored data in a real office building located in\nVarennes, Quebec. The energy demand is predicted\nat every hour for the following three hours using\nweather forecasts. The results show that during\noccupancy, 7:00 to 17:00, the coefficient of\nvariance of the root-mean-square-error (CVRMSE)\nis below 12.3%, the normalized mean bias\nerror (NMBE) is below 1.3% and the root-meansquare-\nerror (RMSE) is below 16.6 kW. When the\nstatistical criteria are calculated for all hours of the\nday, the CV-RMSE is 13.9%, the NMBE is 2.7%\nand the RMSE is 17.9 kW. The case study\ndemonstrates that CBR can be used for energy\ndemand prediction and could be implemented in\nbuilding operation systems.
Language: English
Subfield: Building and Construction
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2271222151