Comparitive study on wind forecasting models for day ahead power markets

摘要

With the increasing wind power penetration in the emerging power system, a precise wind power forecasting method is necessary to help the power system operators, to include wind generation into economic scheduling, unit commitment, power markets, and reserve allocation problems. The main objective of this paper is about the prediction of wind velocity. The uncertainties in wind velocity forecast are analysed using nonlinear regression method and compared with Kalman Filtering technique. The initial data is collected from the global data obtained from ARW (Advanced Research core of the Weather Research and Forecasting) for the region around Vijayawada city and using an algorithm developed by K L University the localized velocity in Vijayawada city is obtained from the global data. This localized data is used for the predictive methods discussed here. The methodology using regression method developed in this paper reduces the error in the initial predicted values of the wind velocity by adding to it, the predicted error based on the available error data. Velocity for the 8th day is calculated in advance using the available data over the past one week. The proposed method helps the producers of wind power to maximize their benefits by bidding in the day ahead power markets with more confidence. The accuracy of the Regression method is compared with the Kalman filtering method and the results show the superiority of the proposed method.