The Danish power-grid has to include more renewable energy. By 2050 the consumption
has to be 100 % based on renewable energy. This means more wind-power, and other
intermittent power producers will be incorporated in the grid. Since the energy-producers
are not well-suited to change their production, the demand-side has to be more flexible in
the future.

One way of making the demand more flexible is by gathering electricity consumers in a
Virtual Power Plant (VPP) and control when they are using power and not. This study is
focused on predicting the power of a VPP consisting of a pool of heat pumps.

A model has been developed, together with methods for estimating the parameters for
the model, and an approach to evaluating the performance of the model. The model takes
data from the “Control Your Heat Pump” platform to make the parameter-estimation. The
report demonstrates how this model is deployed by using Matlab and YALMIP.