基于NExT-DMD的电力系统机电振荡参数提取  

Extraction of Power System Electromechanical Oscillation Parameters Based on NExT-DMD

作　　者：高晗 蔡国伟[1] 杨德友[1] 王丽馨[1] GAO Han;CAI Guowei;YANG Deyou;WANG Lixin(School of Electrical Engineering,Northeast Electric Power University,Jilin 132012,Jilin Province,China)

机构地区：[1]东北电力大学电气工程学院,吉林省吉林市132012

出　　处：《电网技术》2022年第1期284-291,共8页Power System Technology

基　　金：国家自然科学基金资助项目(51877032)。

摘　　要：电力系统随机响应中蕴含丰富的系统动态特征信息,基于系统随机响应所提取的机电特征参数能够准确反映互联电网机电小干扰稳定性。文章在深入分析环境激励下随机响应信号特征基础上,提出了基于自然激励技术的动态模式分解算法(nature excitation technology based dynamic mode decomposition,NExT-DMD)以提取电力系统机电小干扰特征参数。NEx T-DMD采用自然激励技术预处理随机响应信号,获得具有明显振荡特征的振荡分量,有效改善动态模式分解算法(dynamic mode decomposition,DMD)存在的噪声鲁棒性差的问题。进而采用DMD从获得的自由振荡信号中有效提取机电特征参数(振荡频率、阻尼比和模态振型)。IEEE 16机系统和某实际电网量测数据验证了所提方法的可行性和有效性;与随机子空间辨识算法和基于自然激励技术的特征系统实现算法相比,该方法在模式参数提取的精度和跟踪性能方面具有较好的表现,为电力系统机电小干扰稳定评估提供了一种切实可行的新途径。Ambient data from power systems contains abundant information of the system dynamic characteristics.Hence, the extraction of the oscillation mode parameters based on the ambient data can effectively reflect the small signal stability of the interconnected power systems. Based on the in-depth analysis of the characteristics of ambient data under the condition of ambient excitation, this paper proposes a method for extracting the electromechanical oscillation parameters using the Nature Excitation Technology based Dynamic Mode Decomposition(NEx T-DMD) in power systems. The NExT-DMD algorithm, using the Nature Excitation Technology(NExT) to preprocess the ambient data,obtains the oscillation components with the obvious oscillation characteristics, effectively enhancing the noise immunity of the Dynamic Mode Decomposition algorithm(DMD). And then the DMD algorithm is applied to effectively extract the electromechanical parameters(the oscillation frequency, the damping ratio and the mode shape) from the obtained oscillation components. The IEEE 16-generator simulation system and a real power system are used to validate the effectiveness of the proposed method. Compared with the Stochastic Subspace Identification(SSI) and the Nature Excitation Technology based Eigensystem Realization Algorithm(NEx T-ERA), the proposed method shows a better performance in the aspect of extraction accuracy and tracking ability, providing a feasible approach for the small signal stability evaluation in the power system.

关 键 词：小干扰稳定 随机响应 动态模式分解 自然激励技术 模式跟踪 

分 类 号：TM721[电气工程—电力系统及自动化]