The problem of optimal choice of system parameters with respect to a given control quality criterion is studied. To compare systems with different parameter values, a quantitative estimation of controllability is introduced. This estimation is based on the average value of the function that defines the quality criterion. As an example, a very simplified model of an underwater vehicle is considered. The problem of finding an arrangement of its control propellers, in which either the movement time or the energy consumption of the vehicle is minimal is investigated. To test the used approach, the trajectories of the underwater vehicle are randomly generated. The energy consumption and the movement time along these trajectories of systems with different parameter values and different indexes are compared.