传统的电力系统振荡分析方法具有模型参数依赖程度高、计算复杂、只适用于离线分析等不足。从Dirac结构出发,研究基于量测信号的电力系统振荡在线分析方法,尝试为此方法建立理论基础。首先介绍了能量结构的基本概念及其Dirac描述形式,给出了一般电力系统的能量结构;然后介绍如何基于能量结构分析特定端口外的未建模动态对系统振荡的影响;最后将提出的方法运用于不同精细程度的电力系统振荡源定位问题,以此验证所提出的方法的正确性、可行性和通用性。由于能量结构的概念适用于电力系统中的各种设备,因此文中提出的振荡分析方法具有一般性,并且该方法完全基于量测信号进行分析,所以计算快捷、可行性高,可应用于振荡源定位、振荡类型在线辨识和分散协调的振荡控制中。
Traditional methods for power system oscillation analysis have some disadvantages; for example, high dependence on power system models and parameters, complicated computation and restricted to only offline analysis. Starting {rom the Dirac structure, this paper attempts to study a measurement-based methodology for power oscillation analysis, and to establish a theoretical foundation for this methodology. Firstly, the basic concept of the energy structure of a control system and its Dirac description form is introduced, and the energy structure of a general power system is given. Secondly, the idea of utilizing energy structures to analyze the influence of certain un-modeled dynamics to the system stability is introduced. Finally, the proposed methodology is adopted to locate oscillation sources in power systems with different precise extent, so that the validity, practicality and generality of this proposed methodology are verified. As the energy structure concept is applicable to all kinds of devices in power systems, this methodology is universal. Moreover, it is totally measurement-based, so it is computationally efficient and practicable. It is promising in many areas such as oscillation source location, oscillation type online identification and decentralized oscillation inhibition.