Combining Reservoir Computing and Over-Sampling for Ordinal Wind Power Ramp Prediction

Áreas de investigación:

• Methodology - Machine learning

Año:

2017

Tipo de publicación:

Artículo en conferencia

Autores:

• Dorado-Moreno, Manuel
• Cornejo-Bueno, Laura
• Gutiérrez, Pedro Antonio
• Prieto, Luis
• Salcedo-Sanz, Sancho
• Hervás-Martínez, César

Volumen:

10305

Título del libro:

14th International Work-Conference on Artificial and Natural Neural Networks (IWANN2017)

Serie:

Lecture Notes in Computer Science (LNCS)

Páginas:

708-719

Organización:

Cádiz, Spain

Mes:

14th-16th June

ISBN:

978-3-319-59152-0

BibTex:

@conference{dorado2017bd, author = "Manuel Dorado-Moreno and Laura Cornejo-Bueno and Pedro Antonio Guti{\'e}rrez and Luis Prieto and Sancho Salcedo-Sanz and C{\'e}sar Herv{\'a}s-Mart{\'i}nez", abstract = "Wind power ramp events (WPREs) are strong increases or decreases of wind speed in a short period of time. Predicting WPREs is of vital importance given that they can damage the turbines in a wind farm. In contrast to previous binary approaches (ramp versus non-ramp), a three-class prediction is proposed in this paper by considering: negative ramp, non-ramp and positive ramp, where the natural order of the events is clear. The independent variables used for prediction include past ramp function values and meteorological data obtained from physical models (reanalysis data). The proposed methodology is based on reservoir computing and an over-sampling process for alleviating the high degree of unbalance of the dataset (non-ramp events are much more frequent than ramps). The reservoir computing model is a modified echo state network composed by: a recurrent neural network layer, a nonlinear kernel mapping and an ordinal logistic regression, in such a way that the order of the classes can be exploited. The standard synthetic minority oversampling technique (SMOTE) is applied to the reservoir activations, given that the direct application over the input variables would damage its temporal structure. The performance of this proposal is compared to the original dataset (without over-sampling) and to nominal logistic regression, and the results obtained with the oversampled dataset and ordinal logistic regression are found to be more robust.", booktitle = "14th International Work-Conference on Artificial and Natural Neural Networks (IWANN2017)", doi = "10.1007/978-3-319-59153-7_61", isbn = "978-3-319-59152-0", month = "14th-16th June", organization = "C{\'a}diz, Spain", pages = "708-719", publisher = "Springer International Publishing", series = "Lecture Notes in Computer Science (LNCS)", title = "{C}ombining {R}eservoir {C}omputing and {O}ver-{S}ampling for {O}rdinal {W}ind {P}ower {R}amp {P}rediction", url = "dx.doi.org/10.1007/978-3-319-59153-7_61", volume = "10305", year = "2017", }

Abstract:

Wind power ramp events (WPREs) are strong increases or decreases of wind speed in a short period of time. Predicting WPREs is of vital importance given that they can damage the turbines in a wind farm. In contrast to previous binary approaches (ramp versus non-ramp), a three-class prediction is proposed in this paper by considering: negative ramp, non-ramp and positive ramp, where the natural order of the events is clear. The independent variables used for prediction include past ramp function values and meteorological data obtained from physical models (reanalysis data). The proposed methodology is based on reservoir computing and an over-sampling process for alleviating the high degree of unbalance of the dataset (non-ramp events are much more frequent than ramps). The reservoir computing model is a modified echo state network composed by: a recurrent neural network layer, a nonlinear kernel mapping and an ordinal logistic regression, in such a way that the order of the classes can be exploited. The standard synthetic minority oversampling technique (SMOTE) is applied to the reservoir activations, given that the direct application over the input variables would damage its temporal structure. The performance of this proposal is compared to the original dataset (without over-sampling) and to nominal logistic regression, and the results obtained with the oversampled dataset and ordinal logistic regression are found to be more robust.