Decentralized Network Automation

In the last 20 years, the generation of electrical energy has moved from conventional, centralized large-scale power plants in the extra-high voltage to decentralized, renewable small-scale power plants (wind, PV, biogas) in the distribution grids. This has also changed the understanding of grid operation from a centralized, vertical energy distribution from top to bottom (from extra-high to low voltage) to a decentralized and partly horizontal distribution.

This now poses entirely new challenges, especially for low- and medium-voltage networks. After these network levels could be operated mainly passively and without in-depth metering infrastructure, they now have an active role with a very large number of decentralized generators and consumers. In particular, rooftop PV systems, storage systems, heat pumps and electromobility should be mentioned here. These generators and consumers harbor enormous potential to consume the highly weather-dependent, variable generation from wind and PV according to consume, to store it, or to feed the energy back into the grid during periods without renewable generation.

To accomplish this task, these grid levels must be made observable and controllable. In its work, the Decentralized Network Automation group is concerned with the various possibilities of optimally and efficiently using existing resources and devices and automating them in the sense of operational management.

With the use of modern AI methods, for example, network states can be determined with a fraction of the measured values of conventional methods. These then form the basis for coordinating the individual components of the system with each other. For this purpose, classical methods of optimization and control are used, but also innovative approaches with the help of AI and probabilistic methods.

Topics:

  • State estimation and prediction
  • Coordination and optimization of distributed generation and consumption units
  • Application-oriented use of AI and ML methods
  • Mathematical optimization and control methods
  • Automation of operational control