Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Control Design for Dynamic Virtual Power Plants
We aim to design distributed, linear feedback controllers for a group of power converters in a power system such that they do not exceed limits on current saturation and actuation for a collection of disturbances. We do so using a novel control design method to include state and input constraints.
Keywords: Distributed Control, Linear Feedback Control, State and Input Constraints, System Level Synthesis, Power System Dynamics, Power System Control, Converter Control, Faults, Nonlinear Disturbances, Ensemble Control
Power systems are witnessing a rapid transition from fossil fuel-fired generation to renewable generation which is interfaced to the grid via power converters. To improve reliability, we propose to combine a group of distributed renewable energy sources into a single dynamic virtual power plant (DVPP) to jointly provide frequency and voltage control. Unlike conventional generation, power converters have low inertia and can only sustain small increases in their internal current before protection devices operate to disconnect them from the grid. The combination of low inertia and disconnections caused by exceeding low current limits makes future power systems highly vulnerable to faults (temporary short circuits that cause large, nonlinear disturbances). Ensuring that current saturation and actuation limits are not exceeded in DVPPs amounts to control design of distributed, linear feedback controllers with state and input constraints. The challenge is to satisfy these constraints for each fault in the collection, while also not being overly conservative. We recently developed a novel control design technique based on system level synthesis (itself a recent control design method) which is able to simultaneously design distributed linear feedback controllers for a group of converter-interfaced renewable energy sources in a DVPP while ensuring that the bounds on states and inputs are tight for each disturbance signal considered. This is shown in the two attached figures for a test case DVPP consisting of a wind turbine (WT), photovoltaic (PV), and energy storage (ES) interfaced via power converters. To our knowledge this represents the first general linear feedback control design technique with state and input constraints that is non-conservative. This project explores applications and extensions of this technique to ensure safe operation for a collection of faults.
Power systems are witnessing a rapid transition from fossil fuel-fired generation to renewable generation which is interfaced to the grid via power converters. To improve reliability, we propose to combine a group of distributed renewable energy sources into a single dynamic virtual power plant (DVPP) to jointly provide frequency and voltage control. Unlike conventional generation, power converters have low inertia and can only sustain small increases in their internal current before protection devices operate to disconnect them from the grid. The combination of low inertia and disconnections caused by exceeding low current limits makes future power systems highly vulnerable to faults (temporary short circuits that cause large, nonlinear disturbances). Ensuring that current saturation and actuation limits are not exceeded in DVPPs amounts to control design of distributed, linear feedback controllers with state and input constraints. The challenge is to satisfy these constraints for each fault in the collection, while also not being overly conservative. We recently developed a novel control design technique based on system level synthesis (itself a recent control design method) which is able to simultaneously design distributed linear feedback controllers for a group of converter-interfaced renewable energy sources in a DVPP while ensuring that the bounds on states and inputs are tight for each disturbance signal considered. This is shown in the two attached figures for a test case DVPP consisting of a wind turbine (WT), photovoltaic (PV), and energy storage (ES) interfaced via power converters. To our knowledge this represents the first general linear feedback control design technique with state and input constraints that is non-conservative. This project explores applications and extensions of this technique to ensure safe operation for a collection of faults.
1. The student will become familiar with the novel control design method for non-conservative distributed linear feedback control with state and input constraints.
2. The student will become familiar with the formulation and design of dynamic virtual power plants (DVPPs).
3. This control design method will be applied to an existing test case in Matlab - which will be provided to the student - to ensure power converter current saturation and actuation limits are satisfied for a given collection of faults.
4. If time permits, the student may investigate conditions under which input and state constraints will not be exceeded for all convex combinations of the signals in the given fault collection.
5. If time permits, the student may explore further extensions and modifications to the novel control design method.
The project can be adapted on the run if new interesting research directions arise, and can focus more on computational or theoretical directions depending on the interests of the student.
**Corona Disclaimer:** This project can be done in person at the Automatic Control Laboratory, hybrid, or completely remotely, depending on the current ETH regulations. Most importantly, we can change between these forms whenever needed.
Finally, if the results are promising they can be turned into a publication.
1. The student will become familiar with the novel control design method for non-conservative distributed linear feedback control with state and input constraints. 2. The student will become familiar with the formulation and design of dynamic virtual power plants (DVPPs). 3. This control design method will be applied to an existing test case in Matlab - which will be provided to the student - to ensure power converter current saturation and actuation limits are satisfied for a given collection of faults. 4. If time permits, the student may investigate conditions under which input and state constraints will not be exceeded for all convex combinations of the signals in the given fault collection. 5. If time permits, the student may explore further extensions and modifications to the novel control design method.
The project can be adapted on the run if new interesting research directions arise, and can focus more on computational or theoretical directions depending on the interests of the student.
**Corona Disclaimer:** This project can be done in person at the Automatic Control Laboratory, hybrid, or completely remotely, depending on the current ETH regulations. Most importantly, we can change between these forms whenever needed.
Finally, if the results are promising they can be turned into a publication.