Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of GraphSearch.
In this paper, the challenge of asymptotically rejecting sinusoidal disturbances with unknown time-varying frequency and bounded rate is explored. A novel data-driven approach for designing linear parameter-varying (LPV) con- troller is introduced, leveraging only frequency domain data from a Linear Time Invariant (LTI) Multi-input Multi-output (MIMO) system, eliminating the need for a parametric model. A two-step iterative algorithm is proposed involving convex optimization problems in the frequency domain. Closed-loop stability is ensured using Integral Quadratic Constraints (IQC) that characterize the bounded rate variation of the LPV controller’s scheduling parameters. Experimental validation is provided through results obtained on a hybrid micro-vibration damping (MIVIDA) platform tailored for space applications.
Alireza Karimi, Elias Sebastian Klauser
Alireza Karimi, Vaibhav Gupta, Elias Sebastian Klauser