Convolutional Sequence to Sequence Non-intrusive Load Monitoring

TL;DR

This paper introduces a convolutional sequence-to-sequence model for non-intrusive load monitoring, utilizing gated linear units and residual blocks to improve appliance energy disaggregation accuracy.

Contribution

It presents a novel convolutional sequence-to-sequence architecture with residual connections for better energy disaggregation in NILM tasks.

Findings

Achieves satisfactory disaggregation performance on the REDD dataset.

Outperforms previous convolutional sequence-to-point models.

Effectively handles appliances with diverse characteristics.

Abstract

A convolutional sequence to sequence non-intrusive load monitoring model is proposed in this paper. Gated linear unit convolutional layers are used to extract information from the sequences of aggregate electricity consumption. Residual blocks are also introduced to refine the output of the neural network. The partially overlapped output sequences of the network are averaged to produce the final output of the model. We apply the proposed model to the REDD dataset and compare it with the convolutional sequence to point model in the literature. Results show that the proposed model is able to give satisfactory disaggregation performance for appliances with varied characteristics.