Multi-Step-Ahead Forecasting of Wind Speed Based on EMD-RBF Model

Abstract

Wind speed forecasting is critical for wind energy conversion systems. Adaptive and reliable methods and techniques of wind speed forecasting are urgently needed in view of the stochastic nature of wind resource, which is varying from time to time and from site to site. Multi-step-ahead speed forecasting is built with empirical mode decomposition (EMD) method and RBF neural network, which makes use of non-linear and non-stationary signal characteristics. Time series of original wind speed data is decomposed by EMD method. And RBF neural network is used to predict the decomposition of the various components. Experimental results show that the method effectively improves the accuracy and the reliability of wind speed forecasting.

Multi‐step wind speed and wind power forecasting using variational momentum factor and deep learning based intelligent neural network models

“…The best way to produce higher power generation from wind resources is to locate the wind mills in the places where there is higher wind speed. Hence, early researchers have worked to develop conventional and heuristic models 1–12 for predicting wind speed based on wind parameters, so that it facilitates the power engineers in constructing wind mills at the predicted locations and thereby more power is been generated. Henceforth, this research article is intended to develop intelligent neural computing models for forecasting wind speed and thereby wind power so that it helps the wind people to locate wind mills at desired locations.…”

“…The ELMAN network model was used to perform the direct multi‐step forecasting 1 . In a work, multi‐step‐ahead speed forecasting was built with empirical mode decomposition (EMD) method and RBF neural network, which makes use of non‐linear and non‐stationary signal characteristics 2 . A modified EMD‐FNN model (EMD based feed‐forward neural network (FNN) ensemble learning paradigm) and also unscented Kalman filter (UKF) was proposed for wind speed forecasting 3,4 .…”

“…1 In a work, multi-step-ahead speed forecasting was built with empirical mode decomposition (EMD) method and RBF neural network, which makes use of non-linear and non-stationary signal characteristics. 2 A modified EMD-FNN model (EMD based feed-forward neural network (FNN) ensemble learning paradigm) and also unscented Kalman filter (UKF) was proposed for wind speed forecasting. 3,4 A work proposed to extract the characteristic components of original data using the ensemble EMD and sample entropy techniques.…”