A determination of a sequence of actions is performed such that a sequence of states resulting from the sequence of actions is optimized using a prescribed optimization function. The optimization function includes a variable parameter with which it is possible to establish a risk relating to the resulting sequence of states based upon a prescribed state of the system.