Talk 9.4: 3:15PM–4:15PM
| Session Name: | Advances in Control Theory and Applications |
| Session Time: | 3:15PM–4:15PM |
| Author Name: | Simon Lam |
| Author Email: | simon@control.utoronto.ca |
| Talk Title: | Control of Plants with Changing Dynamics using Switching Model Predictive Control |
| Slides: | 9-4.pdf |
| Abstract: | This presentation will introduce a new adaptive switching controller that can be used to control systems whose dynamics may drastically change. Typically for such a problem, the system cannot be described by one particular model and a fixed controller is hopelessly inadequate. One approach then is to use multiple controllers and a switching scheme to select which controller to apply. One such scheme that solves the so-called "family of plants" problem is presented. Like most switching schemes, however, this switching controller has the limitation of producing undesirable (large) transient responses, due to the fact that large control signals are produced during the switching or "learning" period. A new switching controller is presented that uses model predictive control to impose constraints on the control signals. It will be shown by simulation results that the new controller solves the "family of plants" problem with greatly improved transient behaviour. |
| Research Group: | Systems Control |
| Degree Program: | Ph.D. |
| Author Bio: | Simon Lam is currently a 2nd year Ph.D. student under the supervision of Professor Edward J. Davison in the Systems Control Group. He received his M.A.Sc. degree in 2005 and his B.A.Sc. in 2003, both from the University of Toronto. His current research interests include robustness control, decentralized control, and numerical methods. |