摘要
模块化多电平换流器(MMC)发生直流故障后,换流站会出现明显的过电流,对MMC的可靠性产生严重影响。而故障暂态电流的精确计算是实现对暂态过电流有效评估的最直接的手段。为了精确计算MMC故障过电流,提出一种同时考虑交流系统与电容放电影响的故障电流精确计算方法,对桥臂等效模型进行了建模与分析,指出了桥臂电压对时间的微分应由电容项与子模块投入函数项组成。基于桥臂等效模型,建立了MMC直流故障下的状态方程,计算得出了MMC直流故障电流的数值解,对可能出现的最大过电流与最小过电流进行评估,并将此结果与电磁暂态仿真模型(EMT)中的仿真结果进行对比验证,证明了该方法的准确性。最后,将该计算方法与两种传统的故障电流计算分析方法进行了比较,指出了传统故障电流计算方法的不足之处,并进一步验证了该方法的有效性。该文提出的计算和分析方法可以为MMC关键部件的选型提供一定的参考,并为直流故障保护方案的设计提供依据。
The modular multilevel converter(MMC)is vulnerable to DC fault.Significant overcurrent will occur on the converter after the fault,which has a serious impact on the reliability of MMC.The accurate calculation of fault transient current is crucial to evaluate the overcurrent.This paper presents an precise method to calculate MMC fault overcurrent.The proposed calculation method considers both the influence of grid voltage and DC capacitor discharging.The equivalent model of MMC is established and analyzed,it can be concluded that the differential of arm voltage to time should be composed of capacitance term and submodule switch function term.Based on the method,the numerical solution of the fault current under the DC fault was calculated.Maximum and minimum possible overcurrent was derived.The precision of calculation method was verified by the electromagnetic transient(EMT)simulation results.Finally,the proposed approach was further compared with the two conventional calculation methods.It is verified that the proposed approach has significantly higher accuracy than the conventional methods. The proposed approaches provide anumerical evaluation tool for the design of MMC and DC fault protection.
作者
蔡洋
郭文勇
赵闯
桑文举
Cai Yang;Guo Wenyong;Zhao Chuang;Sang Wenjun(Institute of Electrical Engineering Chinese Academy of Sciences,Beijing,100190,China;Key Laboratory of Applied Superconductivity Chinese Academy of Sciences,Beijing,100190,China;University of Chinese Academy of Sciences,Beijing,100049,China)
出处
《电工技术学报》
EI
CSCD
北大核心
2021年第7期1526-1536,共11页
Transactions of China Electrotechnical Society
基金
国家重点研发计划(2018YFB0905800)
国家自然科学基金(51877206,51721005)资助项目。