基于广域测量系统和CELL理论的强迫振荡在线感知与定位

发布时间：2018-01-20 18:42

本文关键词： 强迫振荡扰动源定位广域测量系统特征椭球决策树　出处：《电力自动化设备》2015年02期 　论文类型：期刊论文

【摘要】：为了快速准确地定位电网强迫功率扰动源,根据"先感知,再分类,后定位"的思想,提出基于广域测量系统的空间特征椭球和决策树混合定位扰动源的新方法。通过对比分析不同强迫功率振荡信号,将实测的不同受扰轨迹信息映射到多维特征椭球,通过计算椭球的空间形状及其形态参数变化,实现强迫功率振荡态势的定量化描述;在抽取空间椭球特征参数的基础上,将不同扰动下强迫振荡瞬态阶段的特征椭球参数形成决策树样本集,利用C4.5算法离线训练,在线匹配以快速分类定位扰动源。算例结果表明,该方法可以在强迫功率振荡瞬态阶段快速分类定位不同扰动源,定位振荡主要参与机组和负荷的准确率很高。
[Abstract]:In order to locate power disturbance sources quickly and accurately, according to the idea of "first perceiving, then classifying, then locating". A new method for locating disturbance source based on spatial feature ellipsoid and decision tree is proposed. The oscillation signals of different forced power are compared and analyzed. The measured information of different perturbed trajectories is mapped to the multidimensional characteristic ellipsoid, and the quantitative description of the forced power oscillation situation is realized by calculating the spatial shape of the ellipsoid and the variation of its morphological parameters. On the basis of extracting the spatial ellipsoid characteristic parameters, the characteristic ellipsoid parameters in the transient stage of forced oscillation under different disturbances are formed into a decision tree sample set, and the C4.5 algorithm is used for off-line training. The simulation results show that the proposed method can be used to locate different disturbance sources quickly in the transient stage of forced power oscillation, and the accuracy of the location oscillation is very high.
【作者单位】：四川大学电气信息学院;阿尔斯通电网研究与技术中心;
【基金】：国家自然科学基金资助项目(51261130472) 英国ALSTOM公司国际科研合作项目(11H0207)~~
【分类号】：TM712
【正文快照】： 0引言随着以特高压电网为骨架的全网互联的推进,电网规模日益扩大,大型互联电网的高非线性与强动态耦合性导致近年低频振荡现象越来越频繁。强迫功率振荡作为特殊的低频振荡在实际电网中多次发生[1],严重地影响了电网安全稳定运行,降低了联络线输电能力。常规的负阻尼振荡控制

【参考文献】