鄞州风电场基于STMAS-WRF模式的风力预报性能评估

Wind Forecasting Performance Evaluation of Yinzhou Wind Power Plant Based on STMAS-WRF Model

利用STMAS-WRF模式,对宁波鄞州风电场2021年12月23-26日的一次强冷空气大风过程进行数值模拟试验,并采用风机的观测资料进行误差检验,对此次大风过程的模式模拟能力进行评估。结果表明,STMAS-WRF模式对此次强冷空气大风过程具有较好的模拟能力,基本模拟出了风速随时间的变化,相关系数均达到0.8以上,均方根误差为2.83~3.86 m·s^(-1),平均绝对百分比误差为21.8%~29.8%。模拟结果表现出了冷空气影响过程中风速减弱增强再减弱的过程,但是对于风速减弱时段大幅度波动的风速,模式模拟普遍还存在偏大的现象。从不同风速段来看,模式对于小风速段和额定风速段的模拟效果较好,对变化风速段的模拟误差较大。通过对预报场和实况场环流的对比分析,发现模式对高层风速模拟效果较好,低层风速预报较实况偏大,可能与模式对于地面摩擦的影响考虑不足有关。

In this paper,the STMAS-WRF model is used to simulate a strong cold air gale process in Yinzhou Ningbo wind power plant from December 23 to 26,2021 and the observation data of the wind turbine is used for error checking to evaluate the model simulation capability for this gale process.The results show that the STMAS-WRF model has a good simulation capability for this strong cold air gale case.The variation of wind speed with time is basically simulated,and their correlation coefficient exceeds 0.8,with the root mean square error between 2.83 m·s^(-1) and 3.86 m·s^(-1) and the average absolute percentage error between 21.8% and 29.8%.The simulation results illustrate the pattern of wind speed weakening and strengthening and then weakening during the cold air influence process,but the model simulations generally have bigger biases for the large fluctuation in wind speed during the wind weakening period.From the view of different wind speed sections,the model has better simulation effect for the small wind speed section and the rated wind speed section,but bigger simulation error for the varying wind speed section.Through the comparative analysis of the circulations between the model forecast field and the observation field,we have found that the simulation effect of the model is better for the upper-air wind speed,but for the low-level wind speed,the model forecast has larger value than the observation,which may be because the model does not sufficiently consider the influence of the surface friction.