New project funded by SNSF
Direct Data-Driven Control
In this project, direct data-driven algorithms are developed in time and frequency domain to compute fixed structure controllers that are suitable for distributed control of large systems. In these methods, the input/output data are used directly for controller design, with no assumption on the model structure or the need for a parametric model. The linear time-invariant systems are studied but the extension of the algorithms to some classes of nonlinear systems, such as sector nonlinearity, passive nonlinearity and linear parameter varying systems will be investigated. The results will be validated on a real time simulator of an electrical grid. Distributed controllers will be designed for distributed generation units that are connected to the electrical grid by electronic power converters.
The project is funded by SNSF for two PhD students.