Model Reduction, Topology Identification and Distributed Control of Large Power Systems using Wide-Area Phasor Measurements
Abstract:
Recent advances in the wide-area measurement system (WAMS) technology using phasor measurement units (PMU) have given a new impetus to modeling and control-oriented research in large-scale electric power systems. One of the main challenges in the dynamic analysis and control of power systems is the development of analytical tools from limited measurement data. In this talk I will address this problem and develop methods for model reduction, topology inference and controller designs for large power systems using PMU measurements. The discussion will be broadly divided into two parts. In Part-1, I’ll present several novel coherency-based algorithms for constructing dynamic equivalents of multi-area power systems, defined over different classes of network graphs, using PMU data available from a limited number of nodes. I’ll also illustrate how probabilistic graphical models can be used for identifying the best feasible topologies of these reduced-order equivalents from covariance analyses of correlated PMU data sets. Results will be illustrated through real PMU data from disturbance events in the Pacific AC Intertie and other transfer paths in the US west coast power system. In Part-2, I’ll address the application of these equivalent models in wide-area monitoring with a focus on transient stability assessment. A brief note on how PMU’s can be used for damping control of these inter-area models via a so-called `control inversion’ framework with associated problems of optimal sensor placement, will also be discussed. The overall motivation of the talk would be to understand how the WAMS technology can help us in gaining valuable insight about the physical behavior of the North American grid, which is becoming more expansive, and, hence, more chaotic day by day.