Model of monthly electricity consumption of healthcare buildings based on climatological variables using PCA and linear regression

Abstract

At this time, due to the global pandemic that has occurred, public administrations want to optimize resources and reduce greenhouse gases with more interest than before. It is the case of the Energy Regional Entity of the Junta de Castilla y León (Spain) that pursues the optimization of the energy consumption in particular of healthcare sector buildings. For this purpose, this work focuses on estimating electricity consumption for each month, for which different scenarios will be generated and the corresponding model is obtained for each scenario. This model has been developed considering the historical monthly data of consumption and climatic variables for the last 3 years. Electricity consumption in public sanitary buildings is related to their climatology, due to the use of air conditioning to adjust the indoor temperature. Subsequently, from the models obtained, the results will be analyzed. Significant differences have been observed in the estimation of electricity consumption with respect to the real data provided by the Junta de Castilla y León. The results obtained show how the availability of climatic variables increases the accuracy of the model obtained by about 30%.

Keywords

• PCA
• Data mining
• Regression models
• Building energy index
• Smart metering