Hierarchical Control of Multiscale Systems for the Case of Resilient Future Electrical Power Systems

This project investigates control techniques for distributed systems consisting of physically coupled subsystems with nonlinear dynamics represented on different time-scales, and subject to time-varying constraints and goals. The main objective is to develop hierarchical control concepts which ensure consistent strategies on different layers even for transient behavior, to enhance resilience of system operation also in case of disturbances and faults. The motivating application underlying this problem structure, and at the same time the running example to investigate the methods within this project, are future electric power systems with a very high share of renewable energy sources. The operation of power grids is faced with the trend that synchronous generators with high inertia are replaced by inverter-interfaced renewable generation and storage units. This trend leads, in general, to faster dynamics, less damping due to lower inertia, higher volatility stemming from intermittent availability of wind and solar power, a higher degree of distribution with respect to generation, and more frequent changes of conditions and disturbances affecting the operation. These developments challenge the existing control and operation schemes which rely on clearly separated layers ranging from time-scales of sub-milliseconds to day-ahead planning with controllers that are tailored to static nominal operation and designed based on linear dynamics. To avoid harmful cascading effects over spatially distributed nodes and several time-scales, dedicated countermeasures are required. As opposed to existing measures, e.g., deployment of costly hardware devices or load shedding in extreme cases, the use of adapting and compensating control is investigated as an attractive alternative leading to the following objectives:

1. to develop hierarchic control schemes for networks with physical coupling such that dependencies with respect to time-scales and spatial distribution are consistently considered,
2. to ensure by autonomous adaptation of the control architecture that robustness and stability is achieved also for changing operating conditions and constraints,
3. to study in how far resilience of power system operation can be maintained by hierarchic control for increasingly higher shares of renewable energy and in presence of faults.

While at the last point is explicitly tailored to  power system applications, the trend of shifting system design to larger networks comprising differently fast dynamics can also be observed for other application domains. The ambition of the project is thus to make fundamental contributions to the field of hierarchic control of networks beyond the specific application of power systems.