摘要
高准确度仿真模型是进行大规模风电并网暂态稳定分析的基础,然而双馈风电机组(DFIG)控制策略与参数属于技术秘密难以获取,模型仿真准确性难以保证。针对DFIG故障暂态精确建模难题,提出了基于实测数据的DFIG建模及参数辨识方法。首先,基于电力系统综合稳定程序(PSASP)中DFIG模型及控制结构,建立低电压穿越(LVRT)控制数学模型并分析故障暂态过程,明确LVRT暂态控制核心参数。其次,基于DFIG的LVRT部分现场实测工况数据,利用混沌粒子群算法实现了DFIG故障暂态控制参数辨识。最后,基于剩余实测工况数据进行辨识参数准确性分析与校验,仿真验证了所提参数辨识方法的有效性及准确性。所提方法辨识结果泛化能力强、准确度高,具有较高的工程应用价值。
The high-accuracy simulation model is the basis for transient stability analysis of large-scale wind power integration.However,the control strategies and parameters of doubly-fed wind turbines are technical secrets that are difficult to obtain,and the accuracy of model simulation is difficult to guarantee.In order to address the fault transient modeling problems of doubly-fed wind turbines,a measured data-based modeling and parameter identification method of doubly-fed wind turbines is proposed.Firstly,based on the DFIG model and control structure of the Power System Integrated Stability Program(PSASP),a low voltage ride through(LVRT)control mathematical model is established to analyze the fault transient process,and the LVRT transient control core parameters are clarified.Secondly,based on part of the field measured LVRT data of doubly-fed wind turbines,the fault transient parameters are identified with the chaotic particle swarm optimization algorithm.Finally,the accuracy of the identification parameters are analyzed and verified based on the remaining measured data.The simulation results have verified the effectiveness and accuracy of the proposed parameter identification method.The proposed method has strong generalization ability and high accuracy of identification results,and is of great engineering application value.
作者
李丹
秦世耀
李少林
贺敬
LI Dan;QIN Shiyao;LI Shaolin;HE Jing(National Key Laboratory of Renewable Energy Grid-Intergration(China Electric Power Research Institute),Beijing 100192,China)
出处
《中国电力》
CSCD
北大核心
2024年第8期75-84,共10页
Electric Power
基金
国家电网有限公司科技项目(新能源电站的实测建模与模型参数优化技术研究,5100-202155481A-0-5-ZN)。
关键词
双馈风电机组
低压穿越
参数辨识
实测数据
混沌粒子群
double-fed induction generator
low voltage ride through
parameter identification
measured data
chaotic particle swarm
