Outlier Detection in Buildings’ Power Consumption Data Using Forecast Error

Buildings play a central role in energy transition, as they were responsible for 67.8% of the total consumption electricity France 2017. Because that, detecting anomalies (outliers) is crucial order to identify both potential opportunities reduce and malfunctioning metering system. This work aims compare performance several outlier detection methods, such classical statistical methods (as boxplots) applied actual measurements difference between their predictions, task outliers power data tertiary building located France. The results show that combination regression method, random forest, adjusted boxplot method have promising this type quality problem consumption.