Utilize este identificador para referenciar este registo: http://hdl.handle.net/10400.22/5971
Título: Definition of MV Load Diagrams via Weighted Evidence Accumulation Clustering using Subsampling
Autor: Duarte, Jorge
Fred, Ana
Rodrigues, Fátima
Duarte, João
Ramos, Sérgio
Vale, Zita
Palavras-chave: Load Diagrams
Evidence accumulation
Electricity markets
Load profiles
Clustering
Weighted evidence accumulation clustering
Data: 2007
Editora: WSEAS - World Scientific and Engineering Academy and Society
Citação: Duarte, Jorge; Fred, Ana; Rodrigues, Fátima; Duarte, João; Ramos, Sérgio; Vale, Zita. Definition of MV Load Diagrams via Weighted Evidence Accumulation Clustering using Subsampling, Trabalho apresentado em Proceedings of the 6th WSEAS International Conference on Signal Processing, Robotics and Automation (ISPRA’ 07), In Proceedings of the 6th WSEAS International Conference on Signal Processing, Robotics and Automation (ISPRA’ 07), Corfu Island, Greece, 2007.
Relatório da Série N.º: ISPRA;2007
Resumo: A definition of medium voltage (MV) load diagrams was made, based on the data base knowledge discovery process. Clustering techniques were used as support for the agents of the electric power retail markets to obtain specific knowledge of their customers’ consumption habits. Each customer class resulting from the clustering operation is represented by its load diagram. The Two-step clustering algorithm and the WEACS approach based on evidence accumulation (EAC) were applied to an electricity consumption data from a utility client’s database in order to form the customer’s classes and to find a set of representative consumption patterns. The WEACS approach is a clustering ensemble combination approach that uses subsampling and that weights differently the partitions in the co-association matrix. As a complementary step to the WEACS approach, all the final data partitions produced by the different variations of the method are combined and the Ward Link algorithm is used to obtain the final data partition. Experiment results showed that WEACS approach led to better accuracy than many other clustering approaches. In this paper the WEACS approach separates better the customer’s population than Two-step clustering algorithm.
URI: http://hdl.handle.net/10400.22/5971
Versão do Editor: http://www.wseas.us/e-library/conferences/2007corfu/papers/540-335.pdf
Aparece nas colecções:ISEP – GECAD – Comunicações em eventos científicos

Ficheiros deste registo:
Ficheiro Descrição TamanhoFormato 
COM_FJD_2007_GECAD.pdf558,15 kBAdobe PDFVer/Abrir


FacebookTwitterDeliciousLinkedInDiggGoogle BookmarksMySpace
Formato BibTex MendeleyEndnote Degois 

Todos os registos no repositório estão protegidos por leis de copyright, com todos os direitos reservados.