Performance limitations of large-scale networks with distributed dynamic feedback

Date/Location Information
Seminar Series: 
CCDC Seminar
2017b Spring
Talk Date: 
04/21/2017 - 3:00pm - 4:00pm
Webb 1100
Speaker Information
Speaker Photograph: 
Speaker name: 
Emma Tegling
Speaker Title: 
PhD Student
Speaker Organization: 
Speaker Department: 
Speaker Short Biography: 
Emma Tegling (née Sjödin) received her B.Sc. and M.Sc. degrees, both in Engineering Physics, from KTH Royal Institute of Technology in 2011 and 2013, respectively. Since 2014, she is a Ph.D. student with the Department of Automatic Control at KTH Royal Institute of Technology. Emma has also spent time as a visiting researcher at California Institute of Technology in 2011, the Johns Hopkins University in 2013 and the University of California at Santa Barbara in 2015. Prior to her doctoral work, she was a strategy consultant with Ericsson. Emma’s research interests are within analysis and control of networked systems, with a particular focus on smart power grids.
Talk Abstract: 

Networked control systems arise in a wide range of applications. These systems typically have a global control objective, while the control is distributed and relies only on local feedback from a neighborhood around each site. In this talk, I will address the question of what this implies in terms of limitations to the overall performance of such systems, in particular as the networks grow large.
We consider networked dynamical systems with double integrator dynamics, controlled with linear consensus-like algorithms. Such systems can be used to model, for example, vehicular formation dynamics and synchronization in electric power networks. We assume that the systems are subject to distributed disturbances and study performance in terms of H2 norm metrics that capture the notion of network coherence. In the context of power networks, we also show how such metrics can be used to quantify losses due to non-equilibrium power flows. With local, static feedback control, there are known performance limitations that cause these metrics to scale unfavorably with the network size. We discuss the underlying reasons for these unfavorable scalings and propose distributed dynamic feedback controllers, which, under certain conditions, alleviate the limitations of static feedback.