Simple Wind Power Prediction System Using Self-Tuning Fuzzy Reasoning and Error Persistent Model

In this paper, we propose a simple wind power prediction system using a self-tuning fuzzy reasoning and an error persistent model. The self-tuning fuzzy of which fuzzy rules are automatically adjusted so that the tuning error between a measured and a predicted wind power decreases, and converges to a small constant value by using predicted wind speed and direction data of the mesoscale spectral model (MSM)-GPV providing from the Meteorological Agency. After tuning process is completed, adjusted fuzzy rules imply wind power prediction rules which is considered geographical features, distance between the nearest wind forecast point and the site of the wind power generator system. Moreover, to decrease a prediction error, we apply an error persistent model to prediction results under the fuzzy reasoning. Some simulation results demonstrate the usefulness of the proposed simple wind power prediction system.