Generation of Time-of-Use Tariffs for Demand Side Management using Artificial Neural Networks

This poster proposes a new method to generate individual time-of-use electricity tariffs to exploit the flexibility of energy prosumers while preserving privacy and minimizing communication effort as well as computational cost. Since an employed tariff structure may be impossible to derive analytically from a particular behavior of a prosumer, artificial neural networks may be used to learn the underlying mechanisms implicitly based on simulated household data. Using the acquired knowledge, such a network could be able to generate suitable tariffs to achieve a desired behavior.