摘要
基于机组同调性的影响因素,提出一种新的鼠笼型风电场动态等值建模方法。该方法通过采集不同工况和短路故障类型组合下风力发电机的转速向量,利用two-step法对风电机组聚合分类,并根据Fisher判别分析进行聚类结果的显著性检验。在综合考虑风电场全年风资源统计信息和系统侧不同类型故障发生比例的基础上,以概率最大的机群划分结果,建立风电场概率等值模型。使用DIgSILENT Power Factory平台进行风电场机电暂态仿真,并与传统等值模型和详细模型对比。仿真结果表明,该文提出以转速向量作为分群判据是合理的,所建立的风电场概率等值模型能较全面表征风电场全年的运行外特性,具有重要的工程应用价值。
Based on the coherence influencing factors, a new equivalencing modeling method for squirrel-cage induction generator （SCIG）-based wind farm was proposed. The rotor speed vectors of the wind turbines were sampled in different combinations of operating conditon and short-circuit faults. The two-step cluster method was utilised to divide the wind turbines into groups, and the significance between different groups was tested with the Fisher discriminant analysis. Considering annual wind resource statistics of wind farms and the rates of different system fault types, the probabilistic equivalent model for wind farm was established depending on the largest probability group result. The electromechanical transient model of wind farm was simulated on DIgSILENT PowerFactory platform and the results were compared to that of the traditional equivalent model and detailed model. The simulation results reveal that it is reasonable that the speed vector was considered as cluster-dependent index. The probabilistic equivalent model is able to reflect annual external characteristics of wind farm, and has a great value in engineering applications.
出处
《中国电机工程学报》
EI
CSCD
北大核心
2014年第28期4770-4780,共11页
Proceedings of the CSEE
基金
国家自然科学基金项目(51207039)
国家电网公司科技项目(NY17201200073)
国家能源应用技术研究工程示范项目(NY20110406-1)~~
