为研究风电场尾流特征及与气象条件的关系，选取内蒙古中部地区某风电场33台风电机组，统计分析了2021—2023年平均风速、风向、风频分布等风资源评估参数。基于Jensen尾流模型，计算不同风向及精细化主导风向尾流区风速，探讨考虑尾流效应后的风速与其他气象要素的相关性。结果表明：（1） 2021—2023年内蒙古中部地区风电场以西南（SW）风为主，高频风向年内变化由偏西向偏南转变，月内风向集中且风速差较小。主导风向下平均风速最大，风速频率曲线呈现正偏态分布。（2） 各风向平均风速下，受尾流影响最大的风电机组风速损失率超过10%，其中西北（NW）、东南（SE）风向超过50%风电机组受尾流影响，风速损失集中分布在风电场东北（NE）向偏后位置，偏西风向风速减小更明显。（3） 气压、气温和湿度对不同风向风速日变化的影响程度不同，上述气象因子对风速的影响下，SW风向在4~5 m·s^-1风速区间内尾流模型计算效果相对好于其他风速段，风速平均绝对百分比误差与相对湿度呈负相关。NW风向在9~10 m·s^-1风速区间内尾流模型计算风速与实测更接近，误差与气压和气温都呈正相关。SE、NE风向分别在9~10 m·s^-1、7~8 m·s^-1风速区间尾流模型计算效果较好。研究结果可为风电机组尾流效应分析及风电场风速预测提供一定参考。

To investigate the characteristics of wind farm wake effects and their relationship with meteorological conditions, 33 wind turbines from a wind farm in central Inner Mongolia, China were selected for analysis. Wind resource assessment parameters, including average wind speed, wind direction, and wind frequency distribution, were statistically analyzed from 2021 to 2023. Using the Jensen wake model, wind speeds in the wake area were calculated for different wind directions, with a focus on the refined dominant wind direction. The correlation between wind speeds and meteorological factors, accounting for wake effects, was also explored. The findings are as follows: (1) From 2021 to 2023, the wind farm in central Inner Mongolia was predominantly influenced by southwest winds. High-frequency wind directions shifted from west to south throughout the year. Monthly wind directions were relatively stable, with concentrated wind directions and small wind speed variations. The average wind speed was highest under the dominant wind direction, and the wind speed frequency curve exhibited a positively skewed distribution. (2) Under average wind speeds for each direction, turbines most affected by the wake experienced wind speed losses exceeding 10%. More than half of the turbines were affected by wake effects under northwest and southeast winds, with the most significant losses occurring in the northeasterly downstream positions of the wind farm. Wind speed reductions were particularly pronounced under westerly winds. (3) The impact of barometric pressure, air temperature, and humidity on daily wind speed variation differed across wind directions. For southwest winds, the wake model performed best in the 4-5 m·s^-1 wind speed range, with the average absolute percentage error of wind speed negatively correlated with relative humidity. For northwest winds in the 9-10 m·s^-1 range, the wake model calculations closely matched measured wind speeds, with errors positively correlated with barometric pressure and temperature. In addition, the wake model performed well in the 9-10 m·s^-1 and 7-8 m·s^-1 ranges for southeast and northeast winds, respectively. These results provide valuable insights into the analysis of wind turbine wake effects and wind speed predictions for wind farms.