Application of wavelet and neural network models for wind speed and power generation forecasting in a Brazilian experimental wind park

Wind power forecast using wavelet neural network trained by improved Clonal selection algorithm

“…In [33], it has been discussed that wavelets can effectively be used for both stationary and non-stationary time series analysis, and that is one of the reasons for the wide and diverse applications of wavelets. Wind speed and power prediction approaches based on wavelet transform, as a preprocessor to decompose wind speed/power time series, and ANFIS [34], Auto Regressive Moving Average (ARMA) [35], Artificial Neural Network (ANN) [36], and Support Vector Regression (SVR) [37], as forecast engines, have been presented. As for SVM-based models, they highly depend on appropriately tuning of parameters and involve complex optimization process [32].…”

“…Wavelet transform has been used in some recent research works for wind forecasting, as a preprocessor to decompose wind speed/power time series to a set of sub-series [34,37,35,36]. The future values of the sub-series are predicted by ANFIS [34], SVR [37], ARMA [35] and ANN [36] and then combined by the inverse WT to form the forecast value of wind power/speed.…”

“…The future values of the sub-series are predicted by ANFIS [34], SVR [37], ARMA [35] and ANN [36] and then combined by the inverse WT to form the forecast value of wind power/speed. Another approach to utilize wavelet in a forecast process is through constructing wavelet neural network in which a wavelet function is used as the activation function of the hidden neurons of an ANN.…”

“…1, as the activation function of the hidden neurons have been applied for another application, i.e. price forecast of electricity markets, in [40,36], respectively. Due to the local properties of wavelets and the ability of adapting the wavelet shape according to the training data set instead of adapting the parameters of the fixed shape activation function, WNNs offer higher generalization capability compared to the classical feed forward ANNs [40].…”

Combinação Linear Wavelet Sarima-Rna Com Estágios Multiplos Na Previsão De Séries Temporais

“…Da Análise Wavelet (ou Teoria Wavelet), por sua vez, advém importantes métodos auxiliares de pré-processamento que consistem, basicamente, em fazer a decomposição, filtragem ou alisamento dos dados temporais, antes de sua efetiva modelagem [Aquino et al, 2009]. Diversas abordagens utilizam, de forma integrada, métodos preditivos individuais e métodos de pré-processamento wavelet.…”