The Application Of NWP-Based Wind Power Load Prediction Method In Inner Mongolia Power Grid

Wind power, as a kind of clean energy, has spacious room for development. But due to the randomicity and the uncertainty of the wind, it is very difficult for large-scale wind power plant to integrate into the power grid. A prediction system of the wind will reduce the influence of wind power on power grid, increase the controllability and predictability of wind power, and thereby increase the power grid's capacity of the wind power.Numerical Weather Prediction (NWP) is a method to forecast the future weather, which uses numerical calculation solving the evolution equations of fluid mechanics and thermodynamics. The NWP is quantitative and objective forecast. Using the NWP's wind speed and direction data, we could predict the future load of the wind power plants.In this paper, we firstly analyzed the operational data of the wind farm in the Inner Mongolia power grid and established the reasonable time interval of prediction. We found that most forecasting methods, used in numerical statistics now, are not suitable in term of time interval. Secondly, we analyzed the historical data of the turbine operation and revised the original speed-power curve according to the meteorological characteristics of Inner Mongolia autonomous region. Thirdly, we discussed the conversion method of numerical weather prediction data and electric load of the turbine. At the last part, we discussed how to convert high-altitude wind velocity to a single machine wind speed. To predict the future load of a whole wind power plant, we use the ANN method to correct the wake effect and yaw effect of the turbines.Finally, we analyzed the prediction error of NWP-based wind power load prediction system and discussed how to use the result from NWP-based wind power load prediction system to make power generation scheduling.