Search

Due to the non-linear and non-stationary characteristics of the wind speed time series, it is generally difficult to model and predict such series by single forecasting models. In this paper, two novel hybrid models, which combine empirical mode decomposition (EMD), feature selection with artificial neural network (ANN) and support vector machine (SVM), are proposed for short-term wind speed prediction. First, the original wind speed time series is decomposed into a set of sub-series by EMD. Second, the initial features (input variables) and targets are constructed from all the sub-series and the original series. Then, a feature selection process is introduced to constitute the relevant and informative features. Finally, a predictive model (ANN or SVM) is established using these selected features. The effectiveness of the proposed models has been assessed on the real datasets recorded from three wind farms in China. Compared with the single ANN, SVM, traditional EMD-based ANN, and traditional EMD-based SVM, the experimental results show that the proposed models have satisfactory performance, which are suitable for the wind speed prediction. [ABSTRACT FROM AUTHOR]

Wind energy is rapidly emerging as a potential and viable replacement for fossil fuels owing to its clean way of power production. However, integration of this abundantly available renewable energy into the power system is constrained by its intermittent nature and unpredictability. Efforts to improve the prediction accuracy of wind speed is therefore imperative for its successful integration into the grid. The uncertainty associated with the prediction is also an important information needed by the system operators for reliable and economic operations. This paper presents the implementation of a multi-objective differential evolution (MODE) algorithm for generation of prediction intervals (PIs) for capturing the uncertainty related to forecasts. Support vector machine (SVM) is used as the machine learning technique and its parameters are tuned such that multiple contradictory objectives are satisfied to generate Pareto-optimal solutions. Several case studies are performed for data from wind farms located in the eastern region of United States. The obtained results prove the successful implementation of the methodology and generation of high quality PIs. [ABSTRACT FROM AUTHOR]

In this paper, a hybrid methodology based on Support Vector Regression for wind speed forecasting is proposed. Using the autoregressive model called Time Delay Coordinates, feature selection is performed by the Phase Space Reconstruction procedure. Then, a Support Vector Regression model is trained using univariate wind speed time series. Parameters of Support Vector Regression are tuned by a genetic algorithm. The proposed method is compared against the persistence model, and autoregressive models (AR, ARMA, and ARIMA) tuned by Akaike's Information Criterion and Ordinary Least Squares method. The stationary transformation of time series is also evaluated for the proposed method. Using historical wind speed data from the Mexican Wind Energy Technology Center (CERTE) located at La Ventosa, Oaxaca, México, the accuracy of the proposed forecasting method is evaluated for a whole range of short termforecasting horizons (from 1 to 24 h ahead). Results show that, forecasts made with our method are more accurate for medium (5–23 h ahead) short term WSF and WPF than those made with persistence and autoregressive models. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, a hybrid forecasting approach, which combines the Ensemble Empirical Mode Decomposition (EEMD) and the Support Vector Machine (SVM), is proposed to improve the quality of wind speed forecasting. The essence of the methodology incorporates three phases. First, the original data of wind speed are decomposed into a number of independent Intrinsic Mode Functions (IMFs) and one residual series by EEMD using the principle of decomposition. In order to forecast these IMFs, excepting the highest frequency acquired by EEMD, the respective estimates are yielded using the SVM algorithm. Finally, these respective estimates are combined into the final wind speed forecasts using the principle of ensemble. The proposed hybrid method is examined by forecasting the mean monthly wind speed of three wind farms located in northwest China. The obtained results confirm an observable improvement for the forecasting validity of the proposed hybrid approach. This tool shows great promise for the forecasting of intricate time series which are intrinsically highly volatile and irregular. [Copyright &y& Elsevier]

Abstract: This paper proposes a least-square (LS) support vector machine (SVM)-based model for short-term solar power prediction (SPP). The input of the model includes historical data of atmospheric transmissivity in a novel two-dimensional (2D) form and other meteorological variables, including sky cover, relative humidity, and wind speed. The output of the model is the predicted atmospheric transmissivity, which then is converted to solar power according to the latitude of the site and the time of the day. Computer simulations are carried out to validate the proposed model by using the data obtained from the National Solar Radiation Database (NSRDB). Results show that the proposed model not only significantly outperforms a reference autoregressive (AR) model but also achieves better results than a radial basis function neural network (RBFNN)-based model in terms of prediction accuracy. The superiority of using transmissivity over sigmoid functions for data normalization is testified. Simulation studies also show that the use of additional meteorological variables, especially sky cover, improves the accuracy of SPP. [Copyright &y& Elsevier]

This paper introduces support vector machines (SVM), the latest neural network algorithm, to wind speed prediction and compares their performance with the multilayer perceptron (MLP) neural networks. Mean daily wind speed data from Madina city, Saudi Arabia, is used for building and testing both models. Results indicate that SVM compare favorably with the MLP model based on the root mean square errors between the actual and the predicted data. These results are confirmed for a system with order 1 to system with order 11. [Copyright &y& Elsevier]