Hybrid Flocking Control Algorithm with Application to Coordination Between Multiple Fixed-wing Aircraft
Flocking, as a collective behavior of a group, has been investigated in many areas, and in the recent decade, flocking algorithm design has gained a lot of attention due to its variety of potential applications. Although there are many applications exclusively related to fixed-wing aircraft, most of the theoretical works rarely consider these situations. The fixed-wing aircraft flocking is distinct from the general flocking problems by four practical concerns, which include the nonholonomic constraint, the limitation of speed, the collision avoidance and the efficient use of airspace. None of the existing works have addressed all these concerns. The major difficulty is to take into account the all four concerns simultaneously meanwhile having a relatively mild requirement on the initial states of aircraft. In this thesis, to solve the fixed-wing aircraft flocking problem, a supervisory decentralized control algorithm is proposed. The proposed control algorithm has a switching control structure, which basically includes three modes of control protocol and a state-dependent switching logic. Three modes of decentralized control protocol are designed based on the artificial potential field method, which helps to address the nonholonomic constraint, the limitation of speed and the collision avoidance for appropriate initial conditions. The switching logic is designed based on the invariance property induced by the control modes such that the desirable convergence properties of the flocking behavior and the efficient use of airspace are addressed. The proposed switching logic can avoid the fast mode switching, and the supervisor does not require to perform switchings frequently and respond to the aircraft immediately, which means the desired properties can still be guaranteed with the presence of the dwell time in the supervisor.
Hwang, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our