Electrical load pattern clustering provides useful information on how to partition the customers on the basis of the shape of their actual consumption. The results of the clustering procedure are then used to group the customers into classes by identifying a suitable set of attributes describing each class in a non-overlapping way. Starting from an initial set of centroids, it is possible to construct a clustering algorithm to upgrade the clustering results by refining the initial centroids. The Electrical Pattern Ant Colony Clustering (EPACC) method recently introduced by the authors, based on ant colony clustering principles, has been shown to outperform the classical k-means for electrical load pattern clustering based on an initial centroid model. This paper illustrates the genesis of the EPACC algorithm by providing additional information on how the method parameters have been chosen and set up. Examples are taken from a case study application with actual load patterns.

Formation of Load Pattern Clusters Exploiting Ant Colony Clustering Principles / Chicco, Gianfranco; Ionel, OCTAVIAN MARCEL; Radu, Porumb. - STAMPA. - (2013), pp. 1460-1467. (Intervento presentato al convegno IEEE Eurocon 2013 tenutosi a Zagreb, Croatia nel 1-4 July 2013) [10.1109/EUROCON.2013.6625170].

Formation of Load Pattern Clusters Exploiting Ant Colony Clustering Principles

CHICCO, GIANFRANCO;IONEL, OCTAVIAN MARCEL;
2013

Abstract

Electrical load pattern clustering provides useful information on how to partition the customers on the basis of the shape of their actual consumption. The results of the clustering procedure are then used to group the customers into classes by identifying a suitable set of attributes describing each class in a non-overlapping way. Starting from an initial set of centroids, it is possible to construct a clustering algorithm to upgrade the clustering results by refining the initial centroids. The Electrical Pattern Ant Colony Clustering (EPACC) method recently introduced by the authors, based on ant colony clustering principles, has been shown to outperform the classical k-means for electrical load pattern clustering based on an initial centroid model. This paper illustrates the genesis of the EPACC algorithm by providing additional information on how the method parameters have been chosen and set up. Examples are taken from a case study application with actual load patterns.
2013
9781467322300
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11583/2583377
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo