Characterization of electric consumers through an automated clustering pipeline

Publications - Paper

Characterization of electric consumers through an automated clustering pipeline

Description

Related tags

#Energy Consumption #Smart Grids

Publications - Paper

Characterization of electric consumers through an automated clustering pipeline

Cluster analysis of daily load profiles represents an effective technique for classifying and aggregating electrical users based on their actual consumption patterns.

Among other purposes, it can be leveraged as a preliminary phase for load forecasting, which is similarly applied to consumers within the same cluster. Numerous clustering algorithms have been proposed and developed in the literature, and selecting the most appropriate set of clustering parameters is crucial to ensure reliable results. This paper introduces an automated service suitable for repeated clustering analysis.

The pipeline can process a set of generic time series data and is easily adjustable to test other input clustering parameters; hence, it can be used to find the best parameter set for specific datasets. Furthermore, it facilitates repeated characterization of real-time load profiles aimed at detecting sudden changes in consumer behaviors and variable external conditions, which impact real power forecasting activity on a short time scale.

Download Publication

Related tags

#Energy Consumption #Smart Grids

Search in the site by keyword

Publications - Paper

Characterization of electric consumers through an automated clustering pipeline

Publications - Paper

Characterization of electric consumers through an automated clustering pipeline

Comments