中文摘要：

提出了一种基于聚类算法和实测数据的风电场动态等值建模方法。根据某风电场的实测数据，通过随机抽样比较的方式证明了风电场内机组间的空间效应，并且利用风速曲线在不同机组间的显著差别，说明风电机组间的空间效应在建立风电场动态等值模型时是不可忽略的。利用K-means聚类分析方法并以实测的数据作为分群指标，将某风电场的33台UP77-1．5MW风电机组聚成4个机群，每个机群对应建立一个等值模型，消除了机组间的空间效应。最后，通过将各个模型与实测的数据的等值比较与误差分析，验证了模型的合理性。与传统模型进行比较，实际验证结果表明该方法建立的模型比传统模型精确度高。

英文摘要：

A dynamic equivalent model suitable for wind farms based on clustering algorithms and experimental data is presented in this paper. According to the experimental data of one wind farm, Existence of dispersion between wind farm units is proved by the way of comparison of random sampling. Besides, the study making use of the significant difference between wind speed curves and power curves in different units indicates that the dispersion among wind farm units can＇t be ignored. In the new model, the wind farms ＇33 UP77-1 .SMW wind turbine units are clustered into four categories by making use of the K-means cluster analysis of SPSS platform and using the measured data as clustering index. At last, the macro comparison and error analysis between various models and measured data verifies the rationality of the model. In addition, the result of comparing with the traditional model indicates that the model in the new method of this paper has a higher accuracy.