Subspace Projection Method Based Clustering Analysis in Load Profiling

From demand response to integrated demand response: review and prospect of research and application

“…Its core technique is clustering algorithm, including direct clustering and indirect clustering [22]. The direct clustering is performed on the data of power consumption itself, and various types of clustering techniques have been studied, including K-means [23], fuzzy K-means [24], agglomerative hierarchical clustering [25], self-organizing map (SOM) [26], support vector clustering [27] and subspace clustering [28]. Various types of indexes have been used to quantitatively evaluate the effect of clustering, including Clustering Dispersion Indicator (CDI), Scatter Index (SI), Davies-Bouldin Index (DBI) and Mean Index Adequacy (MIA) [29].…”

Evaluation of the performance of clustering algorithms for a high voltage industrial consumer

“…Algorithms that do not belong to the above categories, like the Expectation Maximization (EM), Support Vector Clustering (SVC), Renyi entropy (Nizar et al, 2006;Chicco and Akilimali, 2009;Chicco and Ilie, 2009), Ant Colony Clustering and a set of Subspace and Projection Clustering Methods (Piao et al, 2014).…”

“…Low voltage consumers (Räsänen et al, 2010;Figueiredo et al, 2005;López et al, 2011;Rodrigues et al, 2003;Nizar et al, 2006, Anuar andZakaria, 2012;Hino et al, 2013;Cao et al, 2013;Iglesias and Kastner, 2013;Piao et al, 2014).…”

A Load Classification Framework Based on VMD and Singular Value Energy Difference Spectrum

“…However, with the increasing demand for load data analysis, the proliferation of load data collection types and quantities has made data analysis more difficult. At the same time, the high load data dimension has a serious impact on the accuracy of the analysis results and brings great difficulty to data analysis [1][2][3]. Therefore, it is urgent to conduct an in-depth study on power system load data analysis.…”