Household Segmentation by Load Shape and Daily Consumption

Œe widespread global deployment of smart meters o‚ers a unique opportunity for utilities to understand the energy use lifestyles and needs of their consumers. It also presents a new challenge: mining these massive datasets and translating €ne-grained consumption data into meaningful insights for purposes of eciency and interpretability. In this paper, we focus on the segmentation of consumer lifestyles based on energy consumption data and propose a structured methodology that considers both the shape of the daily consumer load pro€le as well as its total energy consumption. We show that this distinction is critical to target recommendations of solar photo-voltaics (PV) and baŠeries for each consumer segment.