Date and time: 03.12.2019 14:15-15:00

Place: D111

Slides can be found here (TBA)

Speaker: Luca di Narzo

Title: Residential Load Disaggregation with Temporal Feature Extraction

In 2017, buildings and appliances were responsible for around 30% of global Energy consumption. The two fastest-growing end-uses in buildings are space cooling and appliances. Energy consumption for space cooling has nearly doubled since 2000, driven largely by increased penetration of the use of cooling devices. In order to reduce the impact of the space cooling based consumption on the electrical grids Demand Side Management (DSM) based approaches can be employed. DSM allows customers to make informed decisions regarding their energy consumption, and helps the energy providers reduce the peak load and reshape the load profile. This results in an improved balancing in the smart grid, as well as reduction in overall operational cost and carbon emission levels. Energy Disaggregation (ED), which estimates the consumption of individual appliances from a single meter, can facilitate the implementation of DSM. For the above-mentioned case, the data on the consumption of the air-conditioner instead of the overall consumption can notably improve the implementation of DSM methodologies that act on modifying the air-conditioners’ schedules. The conventional method employed for energy disaggregation is Nonintrusive load monitoring (NILM), which needs dedicated sensors for each appliance (e.g. air conditioner). The objective of this project is a disaggregation technique using a machine learning model, based on Extra Trees Regression, with hourly data from smart meters without requiring additional sensors for supporting sustainable demand-side energy management. A state of the art feature extraction methodology has been strengthen and implemented, followed by a feature selection step to reduce the computational load and improve the results.


I am an M.Sc. student of Energy Engineering – Renewables and Environmental Sustainability Program at Politecnico di Milano and I am currently conducting my thesis in Data Analytics and Optimization for Energy Applications laboratory in the context of a collaboration of the lab with the Department of Electrical Engineering of HVL Norway