Efficiency improvement of short-term forecast for wind power

Wind power is an increasingly used form of renewable energy. However, the inherent randomicity and intermittency of wind resource brings challenges to operators of power systems and wind farms. Therefore, preliminary forecasting of the wind power is necessary. We propose a statistical method for it based on linear regression model. Moreover, many of farms need to be worked on at the same time in some cases. If we predict every wind turbine using the same method such as linear regression, it must be time-consuming. In this paper, we partition the wind turbines into several groups according to the practical need, and choose one representative in each group. Then we predict all the turbines through transformation. Our method is applied to a real data set of one of the 3 largest wind farm operators in China.