基于WAMS的电力系统扰动传播机理研究

发布时间：2018-02-04 10:46

本文关键词： 扰动传播 WAMS 分段均匀模型扰动到达时间扰动定位　出处：《华北电力大学(北京)》2014年博士论文　论文类型：学位论文

【摘要】：扰动是破坏电网安全稳定运行的直接诱因,如何分析扰动影响的范围、程度进而进行安全控制对于提高电网安全水平具有重要作用。随着电网互联节点数量的不断增加,传统分析方法基于电磁波传递的系统动态方程,不能直接给出扰动空间传播信息,并且难以准确预测多机系统稳定性的问题。针对此问题,论文将空间信息引入描述系统动态的微分代数方程中,用机电波理论刻画扰动的时空传播机理。在揭示扰动机电波传递机理的基础上,应用实测PMU相量轨迹信息,辨识系统扰动类型、位置、发生时刻及机电波传递参数,以针对实际扰动和辨识出的机电波传播参数,预测系统扰动传播规律,提前采取安全控制措施。本论文的主要工作和创新点为：提出了适用于扰动传播研究的链式电网模型与实际电网空间框架结构模型。构建了发电机的转动惯量作用域并提出非均匀链式电网的分段均匀介质模型,推导了扰动传播方程和对应的解析解。在链式电网的基础上,提取了实际电网机电波传播特征,提出了沿传播路径逐级对称分布的转动惯量处理方法,构建了具有分段均匀参数的电网空间框架结构模型,并证明了所提模型的正确性。从频率的角度揭示了扰动在电网中传播的幅值、相位变化规律。提出波动传递函数以定量刻画机电波波幅、相位在离散惯量模型中的变化规律；分析了不同频率扰动信号的传播特性,提出了扰动传播的转折频率,揭示了频散现象和局部振荡现象与转折频率的定量关系,最终为基于WAMS的扰动辨识奠定理论基础。提出了一种基于WAMS数据的扰动在线定位方法。揭示了系统稳态时频率的短周期波动特性及扰动到达后的单调升降特性,提出了基于滑动数据窗的机电波到达时间判别函数及判断准则,采用Floyd算法构建了由不同故障点扰动传播时间组成的参考时间矩阵,利用扰动到达的时序特征提出了有序最小距离分类器辨识扰动位置。本算法大大缩小了搜索空间,能够满足在线应用的要求。提出了一种扰动到达时间预测算法,为基于扰动传播理论进行电网主动安全控制提供了重要决策信息。构建了框架结构模型对应的无向赋权图和以各段线路中机电波传播时间为权值的邻接矩阵。提出基于Dijkstra算法的扰动到达时间及最快传播路径搜索算法。同时,进一步分析了扰动传播中的正、负反射现象对机电波到达时间的影响机理,并提出了实测扰动到达门槛值的选择方法。
[Abstract]:Disturbance is the direct inducement to destroy the safe and stable operation of power grid. How to analyze the range of disturbance. The degree of security control plays an important role in improving the security level of power grid. With the increasing number of interconnected nodes, the traditional analysis method is based on the dynamic equation of electromagnetic wave transmission system. It is difficult to accurately predict the stability of multi-machine systems. In order to solve this problem, the spatial information is introduced into the differential algebraic equations to describe the dynamics of the system. On the basis of revealing the mechanism of electromechanical wave transmission of disturbance, the PMU phasor locus information is used to identify the disturbance type and position of the system. The time of occurrence and the parameters of electromechanical wave transmission are used to predict the disturbance propagation law of the system according to the actual disturbance and the identified electromechanical wave propagation parameters. Safety control measures should be taken in advance. The main work and innovation of this paper are as follows:. A chain power network model and a real grid spatial frame structure model are proposed for the study of disturbance propagation. The rotational inertia domain of generator is constructed and the piecewise uniform medium model of non-uniform chain power network is proposed. The perturbation propagation equation and the corresponding analytical solution are derived. Based on the chain power network, the characteristics of electromechanical wave propagation are extracted, and the rotational inertia processing method is proposed, which is distributed symmetrically step by step along the propagation path. The spatial frame structure model with piecewise uniform parameters is constructed, and the correctness of the proposed model is proved. The amplitude of disturbance propagating in the power network is revealed from the angle of frequency. The wave transfer function is proposed to describe the amplitude of electromechanical wave and the variation of phase in discrete inertia model. The propagation characteristics of disturbance signals with different frequencies are analyzed and the transition frequency of disturbance propagation is proposed. The quantitative relationship between frequency dispersion phenomenon and local oscillation phenomenon and turning frequency is revealed. Finally, the theoretical foundation of disturbance identification based on WAMS is established. A disturbance on-line location method based on WAMS data is proposed. The short-period fluctuation characteristics of steady state frequency and the single disturbance after arrival are revealed. Adjust up and down characteristics. Based on the sliding data window, the time of arrival (DOA) of electromechanical wave is determined and the Floyd algorithm is used to construct the reference time matrix which is composed of the disturbance propagation time of different fault points. An ordered minimum distance classifier is proposed to identify the disturbance location using the time-series feature of disturbance arrival. The search space is greatly reduced by this algorithm. It can meet the requirements of online applications. A disturbance time of arrival prediction algorithm is proposed. This paper provides important decision information for active security control of power network based on disturbance propagation theory. The undirected weighted graph corresponding to frame structure model and the adjacent matrix with the weight of electromechanical wave propagation time in each transmission line are constructed. The Dijkstra algorithm based on the disturbance arrival time and the fastest path search algorithm is proposed. At the same time. Furthermore, the influence mechanism of positive and negative reflection on the arrival time of electromechanical wave is analyzed, and the method of selecting the threshold of disturbance arrival is put forward.
【学位授予单位】：华北电力大学(北京)
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TM712

【相似文献】