Data-driven control includes several control theories and methods in which the controller is designed by directly using on-line or off-line input/output (I/O) data of the controlled system or knowledge from the data processing but not any explicit information from mathematical model and from the structure of the controlled process. The development of data-driven algorithms that can learn and construct control laws directly from (I/O) data would result in new learning-based control algorithms that can maintain consistent performance even in presence of complex unstructured uncertainties and dynamical changes.
Representative Themes
- Structure-independent control
- Uncertain Euler-Lagrange systems
- Adaptive sliding mode control
- Relaxing persistence of excitation conditions
- Unfalsified control
Representative Applications
- Control of unstructured robots (e.g. reconfigurable robots)
- Decision making in unstructured environments
- Iterative feedback tuning
- Simultaneous localization and mapping (SLAM)
Representative Results
- Roy S. and Baldi S., “Towards structure-independent stabilization for uncertain underactuated Euler-Lagrange systems”. Automatica, Vol. 111, art. 108775, 2020. doi:10.1016/j.automatica.2019.108775
- Baldi S., Michailidis I., Kosmatopoulos E. B., Papachristodoulou A., and Ioannou P. A., “Convex design control for practical nonlinear systems”. IEEE Transactions on Automatic Control, Vol. 59(7), pp. 1692-1705, 2014. doi:10.1109/TAC.2014.2309271
- Baldi S., Michailidis I., Kosmatopoulos E. B., and Ioannou P. A., “A “plug-n-play” computationally efficient approach for control design of large-scale nonlinear systems using co-simulation”. IEEE Control Systems Magazine, Vol. 34(5), pp. 56-71, 2014. doi:10.1109/MCS.2014.2333272