| Full text | |
| Author(s): |
Fernandes, R. A. S.
;
Deus, L. O.
;
Gomes, L.
;
Valel, Z.
;
Omatu, S
;
Rodriguez, S
;
Villarrubia, G
;
Faria, P
;
Sitek, P
;
Prieto, J
Total Authors: 10
|
| Document type: | Journal article |
| Source: | DISTRIBUTED COMPUTING AND ARTIFICIAL INTELLIGENCE; v. 620, p. 8-pg., 2018-01-01. |
| Abstract | |
This paper presents a statistical analysis of a database generated by voltage and current measurements acquired in a laboratorial environment, which simulates a residential kitchen. In this sense, data were acquired during one month in order to verify both the occurrence of errors as well as the possible identification of the loads. Thus, it is intended that the statistical analysis allows the database to be used to the purposes of Demand Response. However, at first, there was an analysis by histograms in order to verify the occurrence of errors on the measurements and then the feature extraction stage. In the sequence, these features were used to define decision rules that could perform the identification of loads. The results obtained demonstrated an average precision rate of more than 90%. (AU) | |
| FAPESP's process: | 16/00641-4 - Clustering algorithm embedded in smart meter for residential load identification |
| Grantee: | Letícia Oliveira de Deus |
| Support Opportunities: | Scholarships in Brazil - Scientific Initiation |