基于PMU的励磁系统参数辨识研究

发布时间：2018-04-02 05:23

本文选题：励磁系统　切入点：参数辨识　出处：《华北电力大学》2014年硕士论文

【摘要】：同步发电机的励磁系统对电力系统的安全稳定及经济运行至关重要,其模型及参数是否准确关系到电力系统仿真计算的可信度,进而影响到以此为依据的实际工程应用。传统的励磁系统离线建模及参数测试方法已越来越不能适应电力系统发展的要求,而随着相量测量单元(PMU)和广域测量系统(WAMS)的发展,实时获取电力系统的同步相量数据已十分便捷,这为电力系统元件的在线建模及参数辨识提供了一个有效的平台和数据来源。本文利用相量测量单元PMU记录的电网扰动激励下励磁系统的动态同步数据,开展了励磁系统模型参数辨识的可行性研究。首先,从模型及激励的频域特征的角度,分析了励磁模型可降阶或转换的频域条件,并结合常见电网扰动激励的频率特征,深入研究了电网扰动激励下常用励磁模型及其参数的数值可辨识性,并用时域分析方法验证了所得结论,且所述分析方法有别于现有的可辨识性的理论分析和数值判断法。其次,分析了动态PMU同步数据质量对参数辨识的影响,一定程度上理论阐释了随机量测误差及系统输入输出数据的不同步对参数辨识影响的机理。再次,考虑到PMU数据会存在坏数据且实际量测误差难以确定,提出了基于输出误差准则的自适应抗差参数辨识方法,仿真表明所提方法是有效的且基于IGG权函数的抗差方法具有更好的抑制噪声和抗差能力。最后,归纳提出了计及非线性环节的参数分类辨识策略,并从励磁电压曲线拟合效果的角度初步仿真研究了基于大扰动PMU数据整体辨识计及非线性环节的励磁参数的可行性。
[Abstract]:The excitation system of synchronous generator is very important to the safety, stability and economic operation of power system. Whether its model and parameters are accurate or not is related to the reliability of power system simulation. The traditional off-line modeling and parameter testing methods of excitation system have become more and more unable to meet the requirements of power system development, but with the development of phasor measurement unit (PMU) and wide area measurement system (WAMS), It is very convenient to obtain synchronous phasor data of power system in real time, which provides an effective platform and data source for on-line modeling and parameter identification of power system components. In this paper, the feasibility of parameter identification of excitation system model is studied by using the dynamic synchronous data of excitation system under disturbance excitation recorded by phasor measurement unit (PMU). Firstly, from the angle of model and frequency domain characteristic of excitation, In this paper, the frequency domain condition of reducing or converting the excitation model is analyzed, and the numerical identifiability of the common excitation model and its parameters under the power grid disturbance excitation is deeply studied, combining with the frequency characteristics of the common power grid disturbance excitation. The conclusion is verified by time domain analysis method, and the method is different from the existing identifiability theory analysis and numerical judgment method. Secondly, the effect of dynamic PMU synchronous data quality on parameter identification is analyzed. To a certain extent, the mechanism of the influence of random measurement error and the non-synchronization of system input and output data on parameter identification is explained theoretically. Thirdly, considering that the PMU data will have bad data and the actual measurement error is difficult to determine, An adaptive robust parameter identification method based on the output error criterion is proposed. The simulation results show that the proposed method is effective and the robust method based on IGG weight function has better noise suppression and robust capability. A parameter classification and identification strategy considering nonlinear links is proposed, and the feasibility of global identification of excitation parameters based on large disturbance PMU data and nonlinear link is preliminarily simulated from the point of view of the effect of excitation voltage curve fitting.
【学位授予单位】：华北电力大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TM31

【参考文献】