A Real Time Event Detection, Classification and Localization Using Synchrophasor Data
A Real Time Event Detection, Classification and Localization Using Synchrophasor Data
复制标题
DOI：
10.1109/tpwrs.2020.2986019
复制
发表时间：
2020-11-01
影响因子：
6.6
通讯作者：
Amidan, Brett G.
中科院分区：
文献类型：
--
作者：
Pandey, Shikhar;Srivastava, Anurag K.;Amidan, Brett G.
随着极端事件数量、电网组件和复杂性的不断增加，电网控制中心观察到的警报也越来越多。控制中心的操作员需要监控和分析这些警报，以便在需要时采取适当的控制措施，以确保系统的可靠性、稳定性、安全性和弹性。尽管现有的警报和事件处理工具有助于监控和决策，但同步相量数据以及拓扑和组件位置信息可用于检测、分类和定位事件，这是这项工作的重点。在使用数据进行事件分析之前，还解决了相量测量单元 (PMU) 的数据质量问题。开发的算法包括基于统计、聚类和最大似然准则 (MLE) 的异常检测、用于事件检测的基于密度的噪声应用空间聚类 (DBSCAN) 以及用于事件分类的基于物理的规则/决策树。此外，拓扑信息、统计技术和图搜索算法用于事件定位。开发的算法已针对 IEEE 14 总线和 39 总线以及来自美国西部互连 (WECC) 的真实 PMU 数据进行了验证，获得了令人满意的结果。
With an increasing number of extreme events, grid components and complexity, more alarms are being observed in the power grid control centers. Operators in the control center need to monitor and analyze these alarms to take suitable control actions, if needed, to ensure the system's reliability, stability, security, and resiliency. Although existing alarm and event processing tools help in monitoring and decision making, synchrophasor data along with the topology and component location information can be used in detecting, classifying and locating the event, which is the focus of this work. Phasor Measurement Unit's (PMU's) data quality issue is also addressed before using data for event analysis. The developed algorithms include statistic, clustering, and Maximum Likelihood Criterion (MLE) based anomaly detection, Density-based spatial clustering of applications with noise (DBSCAN) for event detection and physics-based rule/ decision tree for event classification. Further, topology information, statistical techniques, and graph search algorithms are used for event localization. Developed algorithms have been validated with satisfactory results for IEEE 14 bus and 39 Bus as well as with real PMU data from the western US interconnection (WECC).