利用有限时间扰动后的响应辨识电力系统的主导特征值

ONLINE IDENTIFICATION OF POWER SYSTEM DOMINANT EIGENVALUES BASED ON ZERO-INPUT RESPONSES FOLLOWING LIMITED-TIME DISTURBANCES

研究电网在运行点处线性化系统的特征值分布是分析小干扰稳定的重要手段。文章分析了零输入响应下线性系统特征值辨识的特点,给出了可辨识性条件,提出了一种新的获得电力系统主导特征值分布的方法,即运用改进Prony算法,在线辨识电力系统在有限时间扰动后不同测点多路零输入响应的信号模型参数,从而得出电力系统主导特征值的分布,辨识过程与系统激励无关。该方法克服了传统离线分析方法中由于运行方式、模型参数不确定而导致结果可信度低等诸多缺点。文章给出了基于WAMS实时数据平台实现电网特征值在线监视的流程和框架。36节点多机仿真系统和实际电网现场测量的分析结果验证了方法的有效性。

Studying the distribution of dominanteigenvalues of the linearized system is an important approach toanalyzing small-signal stability of large-scale power systems.The characteristic of identifying eigenvalues of linear systembased on zero-input response is analyzed, and then theidentifiability is represented. This paper proposes a new methodfor online identifying the distribution of dominant eigenvalues.With the improved Prony algorithm, multi-channel outputresponses, which are measured at different sites followingcertain finite-duration disturbance, are approximated withProny signal-models. Thus the dominant eigenvalues of thestudied power systems are obtained. The proposed methodovercomes the drawback of conventional off-line small-signalanalysis, which is of low reliability because of variation ofoperating condition and uncertainties in model parameters.Based on the real-time data platform of industrializedWide-Area Measurement System (WAMS), the framework andflow for online monitoring eigenvalues of power grid isrepresented. Simulation results and field test data validate itsthe effectiveness.