A combined mixed integer programming and fuzzy control approach for movement of autonomous systems in formation

This paper presents a new combination of algorithms to allow coordinated movement of groups of autonomous systems in formation. A key feature of the work we will present here is a number of specific sets of fuzzy algorithms that, in combination, allow a movement in formation in 3D environments simply based on sensor inputs in a purely reactive manner. This approach is combined with a global optimization realized by using mixed integer programming to allow well planned actions of the formations with automatic avoidance of situations in which the stability of the formation could not be assured. For the combination of these two different algorithms and the possibility to extend these two with other algorithms to fulfill complicated mission tasks we we present a specific control architecture that allows to run several algorithms in parallel during a mission and to select the best performing one as the one that is controlling the autonomous vehicle.