Building a graph signal processing model using dynamic time warping for load disaggregation

He, Kanghang and Stankovic, Vladimir and Stankovic, Lina (2020) Building a graph signal processing model using dynamic time warping for load disaggregation. Sensors, 20 (22). 6628. ISSN 1424-8220 (https://doi.org/10.3390/s20226628)

[thumbnail of He-etal-Sensors-2020-Building-a-graph-signal-processing-model-using-dynamic-time]
Preview
Text. Filename: He_etal_Sensors_2020_Building_a_graph_signal_processing_model_using_dynamic_time.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo

Download (343kB)| Preview

Abstract

Building on recent unsupervised Non-intrusive load monitoring (NILM) algorithms that use graph Laplacian regularization (GLR) and achieve state-of-the-art performance, in this paper, we propose a novel unsupervised approach to design an underlying graph to model the correlation within time-series smart meter measurements. We propose a variable-length data segmentation approach to extract potential events, assign all measurements associated with an identified event to each graph node, employ dynamic time warping to define the adjacency matrix of the graph, and propose a robust cluster labeling approach. Our simulation results on four different datasets show up to 10% improvement in classification performance over competing approaches.