Improving reservoir based wind power forecasting with ensembles

Wind energy - generated from wind power - is plentiful, renewable, clean and available in many places in the world. This energy is generated by wind turbines, in which the wind captured by propellers is connected to a turbine that drives an electrical generator. The use of this source to generate electricity on a commercial scale began in the 1970s, when the international oil crisis escalated. The U.S. and some European countries became interested in the development of alternatives for the production of electricity sources, seeking to reduce dependence on oil and coal. The use of wind power to generate electricity has some drawbacks, however, such as uncertainties in generation and some difficulty in planning and operation of the power system. Several models for wind power forecasting using artificial neural networks have been presented with promising results. This paper presents the use of an ensemble approach to improve the results obtained by models using artificial neural networks, specifically reservoir computing. Reservoir computing is a new paradigm that offers an intuitive methodology for using the temporal processing power of RNNs without the inconvenience of training them. The main issue of using ensemble approach is the consideration of accuracy and diversity of individual predictors which constitute an ensemble.

[1]  Benjamin Schrauwen,et al.  An experimental unification of reservoir computing methods , 2007, Neural Networks.

[2]  Henry Markram,et al.  Real-Time Computing Without Stable States: A New Framework for Neural Computation Based on Perturbations , 2002, Neural Computation.

[3]  A. Timmermann.  Chapter 4 Forecast Combinations , 2006 .

[4]  Milde M. S. Lira,et al.  Application of wavelet and neural network models for wind speed and power generation forecasting in a Brazilian experimental wind park , 2009, 2009 International Joint Conference on Neural Networks.

[5]  Alan F. Murray,et al.  International Joint Conference on Neural Networks , 1993 .

[6]  J. Scott Armstrong,et al.  Principles of forecasting , 2001 .

[7]  Bogdan Gabrys,et al.  Classifier selection for majority voting , 2005, Inf. Fusion.

[8]  T. van der Zant,et al.  Identification of motion with echo state network , 2004, Oceans '04 MTS/IEEE Techno-Ocean '04 (IEEE Cat. No.04CH37600).

[9]  Teresa Bernarda Ludermir,et al.  Forecasting models of wind power in Northeastern of Brazil , 2013, The 2013 International Joint Conference on Neural Networks (IJCNN).

[10]  Thomas G. Dietterich.  Multiple Classifier Systems , 2000, Lecture Notes in Computer Science.

[11]  A. Timmermann.  Forecast Combinations , 2005 .

[12]  Lisa Werner,et al.  Principles of forecasting: A handbook for researchers and practitioners , 2002 .

[13]  Georges Kariniotakis,et al.  Next Generation Short-Term Forecasting of Wind Power – Overview of the ANEMOS Project. , 2006 .

[14]  Herbert Jaeger,et al.  The''echo state''approach to analysing and training recurrent neural networks , 2001 .

[15]  Vladimiro Miranda,et al.  Wind power forecasting : state-of-the-art 2009. , 2009 .

[16]  Benjamin Schrauwen,et al.  Event detection and localization for small mobile robots using reservoir computing , 2008, Neural Networks.

[17]  Lars Kai Hansen,et al.  Neural Network Ensembles , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  Xin Yao,et al.  Diversity creation methods: a survey and categorisation , 2004, Inf. Fusion.

[19]  Milde M. S. Lira,et al.  Recurrent neural networks solving a real large scale mid-term scheduling for power plants , 2010, The 2010 International Joint Conference on Neural Networks (IJCNN).

[20]  Gregor Giebel,et al.  The State-Of-The-Art in Short-Term Prediction of Wind Power. A Literature Overview , 2003 .

[21]  José Carlos Príncipe,et al.  Special issue on echo state networks and liquid state machines , 2007, Neural Networks.