Tomasz Minko

Institut for Elektroniske Systemer - Aalborg Universitet

Friday, December 11, 2015, 11:00

Room 02-012, Georges-Köhler Allee 102, Freiburg 79110, Germany

Abstract: 

The power output of renewable energy sources may change rapidly. Denmark's ambitious energy plans (100 % energy from renewables by 2050) will cause increased dependency on intermittent energy sources, like wind energy. In case of lack of wind, turbines will not fulfil the grids energy demand, and transmission system operators (TSO) may need to compensate by pulling the energy from other sources. To remedy this, it will likely be necessary to draw upon services provided by the so-called Smart Grid. A supermarket is an example of a multi agent system, where each of its subsystems (HVAC, Heating, Solar panel, Refrigeration unit) is controlled autonomously by a local controller. We are examining the possibility of changing set points of a supermarkets local controllers, such as indoor temperature , supply air temperature  and mass flow  for HVAC system, display case air temperature , compressor frequency , condensation pressure  in vapour compression cycle (VCC), in order to decrease the systems total cost of operation. To this end, we introduce a model predictive control (MPC) based supervisory controller that according to the information received from TSO, aggregator, systems local controllers, sensors measurements and weather data, decides what is the cost optimal behavior of the whole system.