A dispatching method of active power in wind farm clusters considering probability distribution of forecasting errors

To smooth out the unbalance in power grid active power dispatch caused by integration of large-scale wind power, it is an inevitable tendency to merge wind power forecasting information into power dispatch. Existing dispatching methods of wind farm clusters are normally based on deterministic wind power forecasting data. Because of the limitation of wind power characteristics and forecasting level with existing technology, it is difficult to meet the requirements of power system dispatching precision. A new dispatching method considering probability distribution of forecasting errors was proposed in this paper. First, the probability distribution of wind power forecasting errors of every wind farm was obtained by analysis of historical data. Then active power requests received by wind farm clusters were allocated based on optimization method whose optimal target was the minimum of the total mathematic expectation of each wind farm's power shortage. As a result, the power shortage of each wind farm caused by forecasting errors was reduced. Finally, the validity and advancement of proposed method were verified with the practical data. © 2013 Chinese Society for Electrical Engineering.