基于WRF模式与相似Kalman滤波法订正后的风速预报效果分析

Analysis of Bias Correction of Wind Speed Forecasting Based on WRF Model and Similar Kalman Filter Method

为提高张家山风电场风速预报质量,比较了张家山测风塔处的实测风速WRF模式预报的风速的误差,再利用相似Kalman滤波方法订正预报的风速,以减小WRF模式预报风场的系统误差和随机误差。结果表明,WRF模式对敏感区的风速预报能力最好,对小风区的预报效果较差,对冬季的预报效果略好于春秋季;再使用相似Kalman滤波方法对模拟风速进行误差订正后,所有时刻的预报能力得到提高,平均偏差、均方根误差变小,从而提高了风速预报的准确率。

In order to improve the quality of wind speed forecasting in Zhangjiashan wind-farm site, the bias of the forecast wind speed was analyzed by using the observation of wind speed and the numerical prediction model WRF. For reducing systematic errors and random errors of the WRF model, similar Kalman filter was used to correct the wind speed forecasting. The results show that the prediction of middle wind speed was the closest to the observed speed by using WRF model, and the wind speed forecasting in winter was better than the other seasons; the forecasting ability improved in all periods after using similar Kalman filter method to correct the simulated wind speed; mean deviation and root mean square error were decreased, and the forecasting accuracy was enhanced.