Deep Learning Based Energy Disaggregation and On/Off Detection of Household Appliances
share
The availability of large-scale household energy consumption datasets boosts the studies on energy disaggregation, a.k.a. non-intrusive load monitoring, that aims to separate the energy consumption of individual appliances from the readings of a mains power meter measuring the total consumption of multiple appliances for example in a house. various neural network models such as convolutional neural network and recurrent neural network have been investigated to solve the energy disaggregation problem. Neural network models can learn complex patterns from large amount of data and have outperformed traditional machine learning methods such as hidden Markov models. However, current neural network methods for energy disaggregation are either computational expensive or are not capable of handling long-term dependencies. In this paper, we investigate the application of the recently developed WaveNet model for the task of energy disaggregation. Based on a real-world energy dataset collected from 20 households over two years, we show that the WaveNet model outperforms the state-of-the-art deep learning methods proposed in the literature for energy disaggregation in terms of both error measures and computational cost. On the basis of energy disaggregation, we then investigate the performance of two deep-learning based frameworks for the task of on/off detection which aims at estimating whether an appliance is in operation or not. The first framework obtains the on/off states of an appliance by binarising the predictions made by a regression model trained for energy disaggregation, while the second framework obtains the on/off states of an appliance by directly training a binary classifier with binarised energy readings of the appliance serving as the target values. We show that for the task of on/off detection the second framework achieves better performance in terms of F1 score.
READ FULL TEXT