Model predictive control of domestic heat pump

In this paper we consider a house heated by a domestic heat pump. We estimate the economical savings that can be achieved by moving the electrical heat pump load in time according to the time-varying value of electricity. Real data from an inhabited typical Danish house is used to estimate a simple thermal house model. Based on this model and based on the electricity spot prices, we design a control algorithm that seeks to minimize the cost of house heating. Via simulations on a model of a typical Danish single-family house, we show that economical savings in the magnitude of 7 % can be achieved during winter by allowing slightly higher variations in the indoor temperature. Further we show that these economical savings are upper bounded by 12 % for the particular case.