“…From the presented mathematical model, the following might be concluded: the optimization problem defined is mixed integer (0-1) non-linear, with non-linearities occurring due to startup and shutdown transient behavior of the components, since all the non-linearities are related to 0-1 variables, the model might be transformed to MILP by introducing additional 0-1 decision variables and inequality constraints, as suggested in [16] or [29], and solved by using classical techniques for this type of optimization problems (e. g. BBM), decision variables related to different observed time intervals are connected, because of start-up and shutdown energy, thermal storage and, eventually, some additional constraints, so the problem cannot be decomposed into separate sub problems for each time interval (hour), and if the vector of all the integer decision variables is predefined, the problem becomes a linear continuous economic dispatch problem, suitable to be solved using LP techniques.…”