RT info:eu-repo/semantics/article
T1 A Granularity Invariant Method for the Classification of Energy Production Time Series in Photovoltaic Plants
A1 Paz Centeno, Iván de
A1 García Ordás, María Teresa
A1 García-Olalla Olivera, Óscar
A1 Girón Casares, Carlos
A1 Alaiz Moretón, Héctor
A2 Ingenieria de Sistemas y Automatica
K1 Energía
K1 Ingenierías
K1 PV production
K1 Classification
K1 Time series
K1 Invariance
K1 Granularity
K1 Deep-learning
K1 3306 Ingeniería y Tecnología Eléctricas
K1 3322.03 Generadores de Energía
AB [EN] The treatment of photovoltaic power production time series often faces the challenge ofunifying the granularity of the series when generating a predictive model. This can limit the generationof a dataset in terms of the time covered and the number of examples. In addition, models built with data of static granularities tend to show rigidity when facing granularity variations, invalidating them for scenarios different from those of the data they were trained on. To address this issue, this paper presents a novel method specifically indicated for Deep-Learning models that shows invariance to granularity called Synthesis. This operation can be added as a layer to an artificial neural network, allowing it to be applied to any power production time series and synthesizing the content of an arbitrarily long time series into a fixed-size vectorwhich can be used for classification or regression regardless of the initial time series length. The experiments with the NIST Campus Photovoltaic dataset demonstrate the ffectiveness of the method, showing an F1-Score of 1.0 for the classification of series with granularities between 2 minutes and 2 hours, and an F-Score of 1.0 for the classification of time series with variations of granularity throughout time when training with 5-minute granularity samples.
PB IEEE
LK https://hdl.handle.net/10612/21187
UL https://hdl.handle.net/10612/21187
NO De-Paz-Centeno, I., Garcia-Ordas, M. T., Garcia-Olalla, O., Giron-Casares, C., & Alaiz-Moreton, H. (2023). A Granularity Invariant Method for the Classification of Energy Production Time Series in Photovoltaic Plants. IEEE Access, 11, 16923-16933. https://doi.org/10.1109/ACCESS.2023.3246159
DS BULERIA. Repositorio Institucional de la Universidad de León
RD Jul 11, 2024