Structuring and Controlling Distributed Power Sources by Autonomous Virtual Power Plants

In future power grids, the number of distinct power plants will rise tremendously. At the same time, many power plants are no longer controllable but their output is subject to the current weather conditions. To deal with the complexity and unpredictability of these future systems, we propose Autonomous Virtual Power Plants (AVPPs) that self-organise into clusters of power plants. These clusters then autonomously plan the energy supply based on predictions made by the power plants and by the consumers. AVPPs react to changes in the load situation by adjusting schedules. They can also change their structure if circumstances make this necessary. Furthermore, trust in the form of reliability of the power plants and credibility of their forecasts is incorporated into the planning.