The large-scale pervasion of wind generators in electric power systems has raised the requirement for fast and reliable forecasting algorithms, aimed at allowing system operators to effectively manage the intrinsic uncertainties induced by their non-programmable generation profiles. To face with the dichotomy between forecasting accuracy and efficiency, this paper explores the potential role of Principal Component Analysis in training a neural-network forecaster, by extracting actionable intelligence from large data sets of historical climatic variables. The results obtained on a real case study are presented and discussed in order to assess the benefits from the application of the proposed method.
The role of principal component analysis in neural-based wind power forecasting
De Caro, F.
;Vaccaro, A.
;Villacci, D.
2018-01-01
Abstract
The large-scale pervasion of wind generators in electric power systems has raised the requirement for fast and reliable forecasting algorithms, aimed at allowing system operators to effectively manage the intrinsic uncertainties induced by their non-programmable generation profiles. To face with the dichotomy between forecasting accuracy and efficiency, this paper explores the potential role of Principal Component Analysis in training a neural-network forecaster, by extracting actionable intelligence from large data sets of historical climatic variables. The results obtained on a real case study are presented and discussed in order to assess the benefits from the application of the proposed method.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
