Wind Farm's Integral Wind Power Model And Wind Speed Forecast

Intensive fluctuation of output power will be caused as a result of wind's randomicity,fluctuation and uncontrollability. The power fluctuation from large-scale wind farm will impactthe voltage and frequency stability of power grid when the percentage of wind farm's capacity inthe total power grid's capacity exceeds a proportion value. So, lots of hybrid energy systems hadbeen considered to solve this problem. According to the practical conditions in Xinjiang region, anovel hybrid power system involves a wind farm and several complementary small gas turbinepower plants (Wind-gas turbine hybrid power system) that supply a constant power has been putout.In order to serve for the wind-gas turbine hybrid power system, Firstly, Auto-Regressiveand Moving-Average (ARMA) model is used to forecast a short-term wind speed on the platformof MATLAB/System Identification Toolbox.Then, the dissertation uses Artificial Nerve Network (ANN) model to forecast the windspeed on the platform of MATLAB/Neural Network Toolbox and makes some comparisonsbetween ARMA method and ANN method.At last, the dissertation builds "the integral wind power model of wind farm" and uses thewind speed predicted in the ARMA model and ANN model to calculate the general output powerof the whole wind farm, which can be used in the wind-gas turbine hybrid power system.