【作者】 刘倩；

【导师】 张恒旭；

【作者基本信息】 山东大学 ， 电力系统及其自动化， 2014， 硕士

【摘要】 目前,随着负荷的日益增长,负荷距离的增加,大型高压直流输电和灵活交流输电系统等的投入使用,当前电力系统的动态特性变得更加复杂。因此,对于电力系统动态稳定的分析就变得尤为重要。论文研究了轻型广域测量系统(WAMS Light)实测数据信号特征和滤波方法,基于WAMS Light提出了监视和辨识一体化的低频振荡在线监视方法,主要研究内容如下：WAMS Light所获得的电网实测信号中存在噪声,这会影响实测信号中有用信号的特性。故本文分析了现实中常见的信号噪声,并对实测信号中的噪声进行了分析,获知了电网实测信号中存在高斯噪声和脉冲噪声；并选取了目前普遍存在的几种信号去噪声方法进行优劣比较,以信噪比为主要性能指标。通过比较分析选取了适合实测信号的去噪方法,可有效去除信号中的噪声。本文所介绍的轻型广域测量系统(PMU Light)每天可采集大量数据,基于这些数据,能够进行低频振荡模态的辨识。而基于负阻尼机理,低频振荡可从数据的特征量上表现出来,故对于这些数据的统计学特征分析就显得尤为重要。本文基于WAMS Light数据,对电网实测数据在稳态特性和动态特性上进行了统计学分析。其中,稳态特性主要包括了实测数据的均值与方差、最值与合格率、实测数据异常事件、实测数据分布特征。本文首次全面揭示了我国各大区域电网的频率宏观运行特性,有助于掌握系统运行概况,提高系统调控策略,甚至提高电网安全稳定运行水平。在电力系统动态稳定性破坏前,往往有明显的轨迹特征表现。电力系统受扰后的响应轨迹包含了动态稳定性信息,即振荡频率、阻尼系数、振荡幅值与相位等振荡模态信息。基于获得的电网实测数据稳态和动态特性以及产生低频振荡的负阻尼机理,本文提出了一种基于WAMS Light的监测和辨识一体化的低频振荡在线监视方法。通过该方法,可以获知电网中某个地区发生低频振荡的时间以及振荡模态,并进一步获得主导振荡模态。同时也获知,同一天中在较大的增幅振荡发生前,会有小幅断续的振荡发生。可以通过一天中多个低频振荡阻尼比的值实现预警,对电网的安全、稳定运行具有重要意义。更多还原

【Abstract】 Nowadays, with the increasing of electrical loads and the distance between them, the application of large-scale high voltage (HV) transmission and Flexible AC Transmission System (FACTS) complicates the current dynamic characteristics of the grid. Therefore, it is more important to analyse the dynamic stability of power systems.In this essay, measured signal features and filtering method are analysed based on WAMS Light. A new method of online monitoring of low frequency oscillation with integration of monitoring and identification is proposed. The main contributions are state below.Measured signals obtained from WAMS Light have noises, which will affect characteristics of usable signals. Hence, in this essay I analyse common noise signals and noises in the actual measured signals and find the Gaussian noise and impulse noise existed in the actual measured signals of the power system. Moreover, several conventional algorithms of filtering are selected and compared and SNR is chosen as the main performance index. Finally, after comparing and analysing, I select a suitable filtering method which could remove noises effectively.A mass of data is able to be collected by PMU Light everyday. Low frequency oscillation can be identified based on these data. Moreover, low frequency oscillation is able to be identified from the feature of the measured data based on negative damping mechanism. Therefore, it is particularly important to analyse the statistical characteristics of these data. In this essay, steady and dynamic characteristics of measured data are statistically analysed based on data obtained from WAMS Light. The steady characteristics mainly include the average value and variance, extreme value and the qualified rate, abnormal events and frequency distribution of the measured signal. The macro features of frequency of power systems in China are firstly revealed, which is helpful to control the operation situation of power system, to improve the controlling strategy and even to improve the security and stability of power systems.Generally, there are obvious trajectory indications before the destruction of dynamic stability of power systems. Dynamic stability information, which includes oscillation frequency, damping coefficient, oscillation amplitude and phase, is contained by the response trajectory after the disturbance of the power system. According to the steady and dynamic characteristics of the measured data in power systems, and negative damping mechanism of low-frequency oscillation, I propose an online monitoring method of low frequency oscillation based on integration of WAMS Light monitoring and identification. By using this method, the time and oscillation modes of low-frequency oscillation occurred in any area of the power system can be obtained. It is known at the same time that a slight oscillation will occur before a large oscillation occurs on the same day. Oscillation warning can be realised through comparing the several damping ratios of low frequency oscillations within that day, which is significance for the safe and stable operation of power systems.更多还原