Fall 2014 MnDRIVE Seminar Series


Nikolai Matni, California Institute of Technology
Regularization for Design: Connections Between Distributed Control and Sparse Reconstruction Theory

Oct. 6, 2014, 2:30pm; 3-180 Keller Hall


Both distributed optimal control and sparse reconstruction have seen tremendous success recently. Although the objectives of these seemingly unrelated fields are very different – the former is concerned with the synthesis and design of controllers, the latter with recovering some underlying signal – they are united through their quest for structure. In this talk, we first make an algorithmic connection between these two fields by showing how convex optimization, and in particular an appropriately chosen atomic norm, can be used for the simultaneous design (or co-design) of a distributed optimal controller and the communication delay structure on which it is to be implemented. We then make theoretical connections by developing approximation bounds and structural recovery guarantees for control co-design problems, once again through the machinery of convex optimization.


Nikolai Matni is currently a PhD student in Control and Dynamical Systems at the California Institute of Technology in Pasadena, California. His advisor is John C. Doyle, and his research leverages techniques from convex optimization, sparse reconstruction and statistical signal processing to solve difficult engineering problems, in particular in the context of distributed optimal control, although he also dabbles in computer vision and bio-medical applications.