The capacity of renewable power sources and especially intermittent sources like wind and PV is steadily increasing. The existing balance between production and consumption is seriously affected by these new sources. Flexible demand for example in buildings is one solution to come back to a stable system. Flexibility in buildings can be achieved by using model predictive control (MPC) with optimized scheduling for the buildings' heating, ventilation and air conditioning (HVAC) systems. Two approaches for self-adapting building models are discussed in this paper as well as different algorithms for optimization of HVAC schedules. The two approaches for self-adapting models can be differentiated by their mathematical structure. The neural network (NN) approach is called “black-box” model. In contrast to that, the “white-box” model is a system of differential equations derived from building physics. Both models are developed to be used in model predictive control to forecast the building's thermal behavior. Once the thermal behavior is predictable, the optimal schedule at minimal costs for the HVAC systems has to be determined with respect to thermal comfort. A schedule for HVAC components contains the information, in which time step which component is on or off. Therefore, a binary integer programming problem has to be solved. (4 pages)