CAJ | 학술논문

风力发电是当今新能源领域最具有前景的发展方向，而准确有效地预测风电场的输出功率对于风电场的顺利并网运行具有重要的意义。从实际的风力发电场中获得了有关风速、风向以及实际输出功率等历史数据，建立了基于 GMDH 神经网络的风电短期功率预测模型并将其运用于实际的风电场功率短期预测当中。最后，通过将预测结果和实际输出功率比较，表明 GMDH 方法在风电功率预测中具有较高的预测精度。
풍력발전시당금신능원영역최구유전경적발전방향，이준학유효지예측풍전장적수출공솔대우풍전장적순리병망운행구유중요적의의。종실제적풍력발전장중획득료유관풍속、풍향이급실제수출공솔등역사수거，건립료기우 GMDH 신경망락적풍전단기공솔예측모형병장기운용우실제적풍전장공솔단기예측당중。최후，통과장예측결과화실제수출공솔비교，표명 GMDH 방법재풍전공솔예측중구유교고적예측정도。
Wind power generation is the most promising direction of the new energy fieId,and forecasting the wind power accurateIy is significant for connetting wind farms with the power grid smoothIy.This paper we get the speed and direction of wind,and the actuaI output power of historicaI data from actuaI wind farms,buiIt short-term wind power prediction modeI based on GMDH neuraI network and appIicated this method for short-term wind power prediction.FinaI y,this paper prepares the predic-tion resuIts with the actuaI output,shows that GMDH method has higher prediction accuracy in wind power prediction.