Data Driven Control
Within this group the focus is on the integrated design, analysis and decision making for large-scale (in physical size) multi-disciplinary dynamical systems. We address the fundamental question about what model complexity and actuator/sensor configuration is necessary for all individual system components in order to use these models for reliable and robust model based diagnostics, parameter estimation, monitoring, parametric system optimization and control. Uncertainty quantification and disturbance modelling are essential parts of the integrated design of these multi-disciplinary systems. Therefore, the combination of measurement data with multi-disciplinary system models is essential to enable reliable, robust and efficient decision making.
This fundamental framework makes it possible to develop robust integrated control systems for demanding industrial application fields, where there is a clear need for integrated system designs with embedded prognostics and diagnostics. Examples are: large-scale mechatronic systems, dynamic positioning systems, ocean-energy-harvesting systems, and wind-energy-harvesting systems.
Object-oriented modelling of multi-disciplinary dynamical systems. Uncertainty modelling. Linear-Parameter Varying, Hammerstein, and Wiener systems. Fault detection, isolation and control. Fixed structure robust and multi-variable controller synthesis. Convex optimization, convex relaxations, integral quadratic constraints, linear and bilinear matrix inequalities. Real-time control systems.
- Jan-Willem van Wingerden
- Kim Batselier
- Dimitris Boskos
- Riccardo Ferrari
- Manon Kok