Wind Power Forecasting Technology in Wind Farm Based on Artificial Neural Network

The intermittent and randomness of wind power will bring serious challenges to the power grid for safety, high quality and economic operation, but timely and accurate wind power prediction is the effective method to solving the above problems. In order to forecast the wind power of wind farms in Poyang lake more accurate, through the special wind meteorological forecast station, the numerical forecast model and wind power statistical forecast model are combined in this project to improve the precision of the wind power short-term forecast: based on the numerical weather prediction data, realize the wind speed short-term forecasting. The increasing of the wind speed prediction time scale is expected to get the short-term forecast in wind speed curve in the future 1h-72h. to forecast the short-term wind power by the neural network model and meet the operation of the power grid dispatching, it can get at least the next 24 h of wind power prediction data. The wind power generation power prediction error will be controlled within 20%, making the grid dispatching more effective use of wind power resources, arranging power generation plan more reasonable, improving the system safety of power grid and the economic benefits of wind farm.