Multiagent Models of Load Prediction

The current situation:

A major role in problems of operational, medium-term and strategic planning in power systems belongs to forecasts of the dynamics of main system state  parameters. One of the most important parameters is the energy consumption (load) of an individual consumer or groups of consumers of different sizes.

On the other hand, an important element of the smart grid concept of the future is "active consumers", who are participants in the retail electricity market and act as both consumers and prosumer.

Prosumers are directly involved in the management of the electricity network by providing distributed generation and storage capacity, direct load control capabilities, and demand response through price signals.

Existing methods of energy consumption forecasting do not take into account the new technological capabilities of the prosumer. More importantly, they do not take into account their involvement in the feedback loop, their ability to set goals and target-oriented behavior on the basis of their own predictive models.

Project objectives: development of algorithms for predicting the consumption of household, municipal and commercial consumers, taking into account their capabilities in distributed generation and consumption shift, the use of automated energy сontrol systems.

Approaches and research methods: multiagent simulation models based on consumer behavioral patterns

Prospects: universal multi-agent modeling environment for electricity markets, market brokers for microgrid markets.