CAJ | 학술논문

风速时间序列的预测对风能的可持续发展至关重要，研究出准确的风速预测算法可以提高电网的安全性，改善环境效益。根据卡尔曼滤波和广义回归神经网络（GRNN）的特点，提出一种新的混合风速预测模型以及滤波误差阈值（FET）预测的方法，可以实现长期的风速预测，具有较高的精度和可靠性。
풍속시간서렬적예측대풍능적가지속발전지관중요，연구출준학적풍속예측산법가이제고전망적안전성，개선배경효익。근거잡이만려파화엄의회귀신경망락（GRNN）적특점，제출일충신적혼합풍속예측모형이급려파오차역치（FET）예측적방법，가이실현장기적풍속예측，구유교고적정도화가고성。
The forecast of wind speed time series is crucial for the sustainable development of wind energy, an accurate wind speed prediction algorithm is developed to improve the security of grid and environmental benefits.The paper proposes a new hybrid wind forecasting models and filtering error threshold (FET) predicted idea based on the Kalman filter and generalized regression neural network (GRNN). The model can predict the long-term wind speed with high precision and reliability.