计及风机不同控制方式的新能源场站等值建模研究

Study on Equivalent Modeling of New Energy Station Considering Different Controlling Strategy

大规模新能源场站的各台风机在实际运行中会受到尾流效应的影响,单独使用风速作为分群指标不能准确地表示新能源场站的运行特性。为此,提出一种适用于双馈风机和直驱风机构成的新能源场站的动态等值建模方法。选取能够表示双馈风机和直驱风机不同运行状态的风速、桨距角、有功功率和转速作为综合分群指标,提出采用PSO粒子群改进算法与DBSCAN相结合的优化聚类算法。利用PSCAD/EMTDC仿真软件,对波动风速和电网侧故障2种工况进行仿真,与单机等值模型和详细模型对比,随后与传统的K均值聚类算法对比,结果表明综合分群指标和优化聚类算法能更好地提升等值建模精度,可以用来表示新能源场站的实际运行特性。

In the actual operation,the wind turbine of large-scale new energy station will be affected by wake effect,so using wind speed in isolation as the cluster index cannot accurately represent the operation characteristics of the new energy station.This paper presents a dynamic equivalent modeling method for new energy station composed of doubly-fed fans and direct-driven fans.By selecting wind speed,pitch angle,active power and speed,which can represent the different running state of doubly-fed wind generator and direct-driven wind generator,as comprehensive clustering indexes,an optimized clustering algorithm combining PSO particle swarm optimization algorithm and DBSCAN is proposed.PSCAD/EMTDC simulation software is used to simulate two conditions of fluctuating wind speed and grid side fault,which are compared with first the single equivalent model and the detailed model and then the conventional K-means clustering model.The results show that the integrated clustering index and the optimized clustering algorithm can better improve the accuracy of equivalent modeling,and can be used to represent the actual operating characteristics of the new energy station.