Genetic-based optimization of temperature set-point signals for buildings with unoccupied rooms

In this paper we present a strategy, based on genetic algorithms, for the optimization of the set-point signals for the heating system of a residential building in case there are totally unoccupied rooms. In particular, the optimization algorithm determines a particular set-point signal for each room based on a simplified linear temperature model of the system. The methodology includes constraints on the temperature level in the occupied rooms based on the choice of the inhabitants. A genetic algorithm determines the optimal set-point signal that have to be applied both to the unoccupied rooms and to the occupied ones in order to minimize the energy consumption in the case of a radiators-based heating system. Simulation results demonstrates the effectiveness of the method.