Forecasting models for wind speed using wavelet, wavelet packet, time series and Artificial Neural Networks

Article ID	Journal	Published Year	Pages	File Type
6692981	Applied Energy	2013	18 Pages	PDF

Abstract

âº Three new methods are proposed to predict wind speed for the wind power system. âº The Wavelet Packet-BFGS is better than the Wavelet Packet-ARIMA-BFGS. âº The Wavelet Packet-BFGS is better than the Wavelet-BFGS. âº They are compared to the Neuro-Fuzzy, ANFIS, RBF neural network and PM.

Keywords

WPD MSE NWP ANN Fnn SBM MAPE BFGS BPNN DWT RBFNN RNN Vector autoregression model Rprop SCMs Resilient propagation autoregressive integrated moving average model Wind speed predictions RBF CWT ARIMA EMD MLP ANFIS DRA back propagation Evolutionary programming MAE PSO Particle swarm optimization Discrete wavelet packet transform Continuous wavelet transform Discrete wavelet transform Wavelet packet decomposition Empirical mode decomposition signal decomposition Wavelet decomposition adaptive neuro-fuzzy inference systems Feed-forward neural network Backpropagation neural network Radial basis function neural network Recurrent neural network artificial neural networks Radial basis function Kalman filter Support vector machines SVM Hybrid model Autoregressive model Mean Absolute Error mean absolute percentage error Mean Square Error Bayesian theory Multilayer perceptron Numerical weather prediction Wind speed forecasting VAR GARCH