Research On Short-term Wind Power Forecast Of Dong Tuan Bao Wind Farm

Shortage of fossil resources and serious environmental pollution all over the world,makes the development of global renewable energy faster and faster.The installed capacity of global wind turbines is also rising very fast,and annual installed capacity of global wind turbines in our country occupies the first place all over the world.However,extremely strong random volatility of wind speed results in the instability of wind power.Large-scale grid connection of wind turbine will disturb the stable operation of power grid,which will stop the development of wind power industry.Short-term prediction of wind power can realize the control of wind power output.This paper proposes prediction model of short-term wind power based on PSO-BP neural network algorithm.Firstly,this paper introduces the construction situation and fan type in Dong Tuan Bao wind farm,and analyzes the distribution situation of wind resource in the place of wind farm—Lai Yuan city.The distribution situation of monthly average wind speed shows the wind speed in summer is smaller than the other.And then,combining the BP neural network algorithm and Particle Swarm Optimization(PSO),prediction model of short-term wind power in wind farm is proposed.Weights and thresholds are optimized based on PSO,which confirms initial weight and threshold in BP Neural network algorithm.And the basic process prediction model of short-term wind power based on PSO-BP neural network algorithm is proposed.Finally,this paper predicts the wind power in Dong Tuan Bao wind farm based on PSO-BP Neural network algorithm.The probability distribution of wind speed in spring,summer,autumn and winter is analyzed based on two-parameter weibull distribution,which shows the probability distribution of wind speed is different in different season.Therefore,the wind power of four seasons in this wind farm is predicted.The wind power prediction accuracy is analyzed by the root-mean-square error(RMSE),which shows the accuracy of the prediction of wind power is high and the proposed PSO-BP neural network algorithm can achieve good effect.