PROJECT TITLE :

Multistep Wind Power Forecast Using Mean Trend Detector and Mathematical Morphology-Based Local Predictor

ABSTRACT:

This paper proposes a novel forecasting model primarily based on a mean trend detector (MTD) and a mathematical morphologybased local predictor (MMLP) to undertake short-term forecast of wind power generation. Within the proposed MTD/MMLP model, the nonstationary time series describing wind power generation is 1st decomposed by the MTD, that employs some new notions and typical morphological operators. The decomposition yields 2 elements-the mean trend, which reveals the tendency of the time series, and therefore the stochastic part, that depicts the fluctuations caused by high frequency of the variability. Subsequently, the p-step forecast is conducted for these two components separately. The mean trend is forecasted on the basis of the least-square support vector machine (LS-SVM) model, whereas the p-step forecast for the stochastic part is applied by the MMLP, that involves performing morphological operations using a novel structuring element (SE) within the phase house. Finally, the forecast of wind power generation is achieved by combining the separate forecasts of 2 parts. In order to guage the accuracy and stability of the MTD/MMLP model, simulation studies are meted out using the info obtained from 3 widely used databases sampled in several periods. The results demonstrate that the MTD/MMLP model provides a more accurate and stable forecast compared to the traditional ways.