【作者】 刘兆燕；

【导师】 曹一家；

【作者基本信息】 浙江大学 ， 电力系统及其自动化， 2008， 博士

【摘要】 基于同步相量测量和现代通信技术的广域测量系统(Wide Area MeasurementSystem,WAMS)可以提供电力系统多个节点的同步信号,这些广域信号均是实时测得的,具有较高的精度和可信度。广域信号在精度和速度上的提高给暂态稳定预测的实用化带来了新进展,出现了一些基于WAMS实测数据的电力系统暂态稳定预测的新方法。同时,WAMS也为广域控制带来了新的契机,然而广域信号的时滞对电力系统小扰动稳定性的影响不可忽略,确定电力系统保持小扰动稳定所能承受的最大时滞对于合理利用WAMS数据、评估广域控制效果具有重要意义。本文研究了广域测量信号在电力系统稳定性分析中的应用,重点在基于WAMS的暂态稳定预测和计及广域信号时滞的电力系统时滞稳定域评估两方面进行了研究。全文主要包括以下几个部分:简要介绍了WAMS的基本概念以及在电力系统中的应用,综述了电力系统暂态稳定分析的传统方法以及WAMS为电力系统暂态稳定分析带来的新思路。总结了WAMS的时滞特性以及广域电力系统时滞现象的研究现状。提出了一种基于WAMS的快速暂态稳定预测算法。该方法在从WAMS获得的实时广域测量信号的基础上,采用了一种在机器人领域被广泛应用的抓球算法,利用粒子群优化算法对该算法进行优化,可以快速预测故障后各发电机的转子角,根据发电机转子角偏离系统的惯性中心是否超过一定的角度可以判断该发电机是否与系统失去同步。预测的结果可用于在线失稳预警或就地控制,测试系统的仿真结果证明了该方法的有效性。研究了广域信号时滞对电力系统小扰动稳定性的影响。首先介绍了时滞动力系统的模型以及稳定性,将计及广域信号时滞的电力系统模型建模为时滞微分代数方程组,然后分析了时滞电力系统的小扰动稳定性,最后通过实例研究了广域信号的时滞对电力系统小扰动稳定性的影响。基于特征根聚类处理法进行了单时滞电力系统时滞稳定裕度的研究。首先介绍了具有单时滞反馈信号的电力系统模型和时滞稳定裕度的定义,然后引入了特征根聚类处理法来求解单时滞电力系统的时滞稳定裕度,该方法通过Rekasius变换将系统特征方程在虚轴上转化为多项式方程,利用传统的Routh判据得到了系统不稳定特征根个数在复平面的分布情况,进而得到了系统的时滞稳定裕度。该方法可以准确求解各种单时滞系统的时滞稳定裕度,最后用时域仿真验证了结果的准确性。针对具有单输入反馈控制的单时滞电力系统,提出了一种求解时滞稳定裕度的简便方法。该方法将含有指数项的特征方程在虚轴上转化为多项式方程进行研究,无需任何中间变量代换,处理方法简单,计算量小。利用该方法得到了测试系统的时滞稳定裕度,并且研究了励磁系统参数的变化对系统时滞稳定裕度的影响,最后通过与其他方法的比较和时域仿真验证了该方法的有效性。基于改进的特征根聚类处理法进行了多时滞电力系统时滞稳定域的研究。首先介绍了一种研究多时滞电力系统时滞稳定域的特征根聚类法,然后引入了Building Block的概念作为特征根聚类法的基础,简化了系统时滞稳定域的求解。对装设TCSC控制器的测试系统进行了多时滞稳定域研究,得到了系统在时滞空间内的稳定区域,最后通过对原系统进行的时域仿真验证了该方法是有效的。更多还原

【Abstract】 The Wide Area Measurement System (WAMS) is based on Synchronized Phasor Measurement (SPM) and the technology of modern communication. WAMS can provide synchronous signals of many nodes in a power system. These wide area signals are measured in real time with high precision and great reliability. The improvement of wide area signals in precision and speed brings about new progress in transient stability prediction. Several novel methods for transient stability prediction based on WAMS have been reported. Meanwhile, WAMS also provides new approches for wide area control. However, the effect of signal delays on power system small signal stability can not be neglected. The estimation of maximum allowed time delays that will not cause system instability is important in application of WAMS data and evaluation of control effects of wide area controllers.Transient stability prediction based on WAMS and evaluation of delay stability regions of power systems are studied in this research. The dissertation is organized as follows:The concept of WAMS and its applications in power systems are introduced. Traditional methods for transient stability prediction are summarized and new methods brought by WAMS are also reviewed. Time delay characteristics of WAMS and recent research on time delay phenomenon of wide area power systems are also introduced.A fast learning method to predict the transient stability of a power system is proposed. It adopts a widely used robotic ball-catching algorithm based on a continual stream of accurate generator rotor angle data from WAMS. The Particle Swarm Optimization (PSO) algorithm is used to perform multi-parameter optimization in this algorithm. This method can predict post-fault rotor angle of each generator. The rotor angle of a generator with respect to the inertial center of a power system can be used to judge whether this generator loses synchronism with the system. The prediction results can be used in on-line instability alarm and local control. Simulation results on two test systems demonstrate the effectiveness of the proposed method.The effects of time delays on small signal stability of power systems are studied. The model and stability concept of time-delayed systems are introduced. A power system with time delays is typically multi-delay dynamical and it can be modeled as differential-algebraic equations with delays. Small signal stability of power system with delays is analyzed. Time domain simulations on the New England Test System with a TCSC controller show that the dynamic performance of this TCSC controller deteriorates sharply with the feedback delay’s increase.Delay stability margin of a power system with a single delay is computed with Clustering Treatment of Characteristic Roots (CTCR) method. Model of a power system with a single delay and definition of delay stability margin are introduced. A novel framework based on CTCR method is presented to determine the delay stability margin of a power system with a single delay. The characteristic equation of single delayed power system is converted into a polynomial equation through Rekasius substitution. The complete distribution of number of unstable characteristic roots on the complex plane is obtained with traditional Routh criterion and the delay stability margin is also determined. Delay stability margins of all kinds of single delayed systems can be obtaind through this method. The validity of this method is verified by the time domain simulation results.A noval method to determine the delay stability margin of power system with single control variable is proposed. The transcendental characteristic equation is converted into a polynomial equation at the imaginary axis without any substitutions. This method needs less computation than the CTCR method. The delay stability margins of two test systems at typical equilibrium point are obtained through this method. The effect of exciter parameters on delay stability margin is also discussed. Time domain simulation results and comparison with other methods show that this method is simple and effective.An improved CTCR method to determine delay stability regions of a power system with multiple delays is proposed. Firstly, the CTCR method to analyze the stability of power system with multiple delays is introduced. Then, the ’ Building Block’ concept is introduced as the preparatory steps of the CTCR method to simplify the computation of delay stability regions. Complete delay stability regions of test systems are obtained through the combination of CTCR and ’Building Block’ concept. Time domain simulation results demonstrate the effectiveness and feasibility of this method.更多还原