Seminars and Events at automatic control
All seminars are held at the Department of Automatic Control, in the seminar room M 3170-73 on the third floor in the M-building, unless stated otherwise.
Master Thesis presentation by Viktor Fredriksson: Residential Load Disaggregation Using Deep Learning Methods
Disputation
From:
2025-06-09 13:15
to
14:00
Place: Large Conference Room 2485-88 in the M-building, LTH
Contact: richard [dot] pates [at] control [dot] lth [dot] se
Date & Time: June 9th, 13:15-14:00
Location: Large Conference Room 2485-88 in the M-building, LTH
Author: Viktor Fredriksson
Title: Residential Load Disaggregation Using Deep Learning Methods
Supervisor: Stephane Velut (Emulate Energy), Anders Rantzer
Examiner: Richard Pates
Abstract: Increasing availability of smart meter data and growing interest in energy efficiency create new opportunities for understanding household energy consumption patterns. Load disaggregation is the task of finding individual loads from aggregate electrical power data. This thesis evaluates three neural network architectures for disaggregating load data, focusing on energy-intensive appliances. The architectures evaluated are recurrent neural networks, state-space models, and attention mechanisms. The results show which approaches are most effective for load disaggregation tasks and demonstrate the benefits of external conditioning through temperature and time-of-day data and different fine-tuning approaches.