This is a talk in three acts. First, we provide a broad over view of what the field of control engineering entails (TL;DR: to make the flow of dynamical systems behave in a particular way) via some concrete examples. In the second act, we briefly are introduced to the notion of optimal control (TL;DR: calculus of variations, "impossible" to solve) and outline a common approach that relies heavily on mathematical programming for implementing (approximate) optimal controllers (aka, model predictive control). In the third act, we consider the problem of selecting the paramters of such controllers via a zeroth-order optimisation framework for non-smooth and possibly non-convex cost functions with matrix parameters that are real and symmetric. We provide complexity bounds on the number of iterations required to ensure a given accuracy level for both the convex and non-convex cases. We conclude the talk by showing how this framework is used for tuning the parameters in model predictive controllers, as this is a challenge known to be inhibiting industrial implementation of these architectures. We show how this can be applied to tune the control parameters of a diesel engine (I know it is not environmentally friendly, but the problem is hard).
Iman Shames is currently a Professor of Mechatronics at the School of Engineering at the Australian National University. Previously, he had been an Associate Professor at the Department of Electrical and Electronic Engineering, the University of Melbourne from 2019 to 2020 and a Senior Lecturer and a McKenzie fellow at the same department from 2012 to 2018, and before that he was an ACCESS Postdoctoral Researcher at the ACCESS Linnaeus Centre, the KTH Royal Institute of Technology, Stockholm, Sweden. His current research interests include, but are not limited to, optimisation theory and its application in control and estimation, mathematical systems theory, and security and privacy in cyber–physical systems.