电力系统机电暂态过程主导动态参数的在线辨识

电力系统机电暂态仿真模型参数的不准确导致了仿真结果和实际量测的明显差异，限制了其在线应用。多年来,基于现场试验测试的动态参数辨识技术有了长足发展，但离满足工程准确性的要求还有相当长的研究路程。基于对电力系统实际动态过程量测的动态参数在线辨识方法是解决问题的一个途径。该文从发展在线机电暂态过程仿真-预测及安全稳定控制的需要和特点出发，研究了基于PMU/WAMS的主导动态参数在线辨识的系统算法。研究表明，应用Volterra级数降阶和模式分类及映射技术，可依据较少的量测信息，对系统中的主导动态参数进行在线快速辨识，弥补了现有电力系统动态元件辨识算法用于在线时要求量测数据丰富和计算量大的问题，为实现电力系统的在线实时辨识-仿真预测并进而实现在线实时控制系统打下了基础。文中给出了EPRI36节点系统和某省网级系统的主导动态参数在线辨识仿真结果，验证了所提出方法的有效性。

Research on the Identification of Dominant Dynamic Parameters of Power System

In power system transient study, the imprecision of model parameters will result in large discrepancies between simulations and real measurements, and thus greatly resitrict online applications of transient simulations. For many years, much progress has been achieved in dynamic parameter identification based on site trials. However, there is still a long way ahead to meet the high-precision requirments of practical engineering. Online dynamic parameter identification based on field dynamic measuerments may be one solution. In this paper, a systematic algorithm for dominant parameter identification based on PMU/WAMS is proposed and studied from the view of developing online transient simulation, stabiity predication and security. Our reseach shows that by using Volterra Model and pattern classification/recognition techniques, the dominant parameters can be identified online very quickly with only a few measurements The proposed algorithm overvomes the disadvantage of the existing identification algorithms, i.e., requirement of abundant measurements and high calculating burden. The method has laid a foundation for realizing online identification, simulation and assessment, and moreover, real-time control. Case studies have been conducted on the EPRI-36 power system and a provincial system. The results have verified the effectiveness of the proposed method.