Advanced Disturbance Recording and Playback enabled by a Distributed Dynamic State Estimation Including Bad Data Detection and Topology Change Identification

摘要

Previous work has demonstrated a distributed dynamic state estimator (DSE) that is implemented in the substation level using local data (synchronized and non-synchronized) and can capture in real time the dynamics of the system with speeds of 60 times per second. Due to the plurality of the available measurements in a substation, bad data is a common occurrence. This paper initially focuses on the algorithm based on which the DSE identifies and removes bad data. Topology changes can also be identified and the system model is automatically updated in the DSE procedure. The end result is a highly accurate and validated model of the system which can be efficiently stored and utilized offline, thus enabling a novel scheme for disturbance recording and playback. Numerical results on the bad data and topology identification, performed on the NYPA Gilboa-Blenheim substation, are presented. A demonstration of the disturbance recording and playback tool is also provided.