Improved Electricity Load Forecasting via Kernel Spectral Clustering of Smart Meters

This paper explores kernel spectral clustering methods to improve forecasts of aggregated electricity smart meter data. The objective is to cluster the data in such a way that building a forecasting models separately for each cluster and taking the sum of forecasts leads to a better accuracy than building one forecasting model for the total aggregate of all meters. To measure the similarity between time series, we consider wavelet feature extraction and several positive-definite kernels. To forecast the aggregated meter data, we use a periodic autoregressive model with calendar and temperature information as exogenous variable. The data used in the experiments are smart meter recordings from 6,000 residential customers and small-to-medium enterprises collected by the Irish Commission for Energy Regulation (CER). The results show a 20% improvement in forecasting accuracy, where the highest gain is obtained using a kernel with the Spearman's distance. The resulting clusters show distinctive patterns particularly during hours of peak demand.

[1]  B. De Moor,et al.  Short-term load forecasting, profile identification, and customer segmentation: a methodology based on periodic time series , 2005, IEEE Transactions on Power Systems.

[2]  Arnaud Doucet,et al.  Autoregressive Kernels For Time Series , 2011, 1101.0673.

[3]  T. Warren Liao,et al.  Clustering of time series data - a survey , 2005, Pattern Recognit..

[4]  A. Atiya,et al.  Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond , 2005, IEEE Transactions on Neural Networks.

[5]  J. W. Taylor,et al.  Short-Term Load Forecasting With Exponentially Weighted Methods , 2012, IEEE Transactions on Power Systems.

[6]  Jean-Michel Poggi,et al.  Clustering Signals Using Wavelets , 2007, IWANN.

[7]  Anestis Antoniadis,et al.  Clustering Functional Data using Wavelets , 2010, Int. J. Wavelets Multiresolution Inf. Process..

[8]  Stamatis Karnouskos,et al.  Impact assessment of smart meter grouping on the accuracy of forecasting algorithms , 2013, SAC '13.

[9]  Philip Hans Franses,et al.  Periodic Time Series Models , 1996 .

[10]  Jean-Michel Poggi,et al.  OPTIMIZED CLUSTERS FOR DISAGGREGATED ELECTRICITY LOAD FORECASTING , 2010 .

[11]  R. Shah,et al.  Least Squares Support Vector Machines , 2022 .

[12]  Marco Cuturi,et al.  Fast Global Alignment Kernels , 2011, ICML.

[13]  Johan A. K. Suykens,et al.  Multiway Spectral Clustering with Out-of-Sample Extensions through Weighted Kernel PCA , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[14]  Thomas Schreiber,et al.  Classification of Time Series Data with Nonlinear Similarity Measures , 1997 .

[15]  Christoph Flath,et al.  Cluster Analysis of Smart Metering Data , 2012, Business & Information Systems Engineering.

[16]  U. Feige,et al.  Spectral Graph Theory , 2015 .

[17]  Ulrike von Luxburg,et al.  A tutorial on spectral clustering , 2007, Stat. Comput..