Physical attackers pose significant danger to the operation of multiagent systems (MASs). Therefore, how to guarantee safety and performance of MAS in the presence of physical attackers is an important topic to be addressed. State-of-the-art literature on constrained multiagent system operations can only deal with constant or at best time-varying constraint requirements. Such constraint formulations cannot respond well to the dynamic environment and presence of physical attackers.
In this work, we consider a formation tracking problem for a group of unmanned aerial vehicles (UAVs) in the presence of a physical attacker. The safety/performance constraint functions are environment-aware and dynamic in nature, whose formulation depends on certain path parameters and presence of the attacker. The dependence on path ensures adaptation to the dynamic operation environment. The dependence on the attacker ensures swift adjustment based on the relative distances between the attacker and agents. UAV desired paths and desired path speeds can also be both path- and attacker-dependent. A framework where composite barrier functions are incorporated with path parameter timing laws has been proposed to address the safety and performance considerations. Adaptive laws and neural networks are used to approximate unknown attacker velocity, unknown system parameters and external disturbances are estimated by adaptive laws. The proposed formation architecture can ensure formation tracking errors converge exponentially to small neighborhoods near the equilibrium, with all constraint requirements met. At the end a simulation study further illustrates the proposed scheme and demonstrates its efficacy.
Dr. Xu Jin received a Bachelor of Engineering (BEng) degree in electrical and computer engineering (First Hons.) from the National University of Singapore, Singapore, a Master of Applied Science (MASc) degree in electrical and computer engineering from the University of Toronto, Toronto, ON, Canada, the Master of Science (MS) degree in mathematics from the Georgia Institute of Technology, Atlanta, GA, USA, and the Doctor of Philosophy (PhD) degree in aerospace engineering from the Georgia Institute of Technology, Atlanta, GA, USA.
Dr. Jin is currently an Assistant Professor at the Department of Mechanical and Aerospace Engineering, University of Kentucky, Lexington, KY, USA. He has authored/co-authored over 70 journal and conference papers with over 3400 citations. He has been a panelist for the National Science Foundation and a reviewer of over 20 journals in the areas of control systems and cybernetics. His current research interests include adaptive and iterative learning control, fault-tolerant/attack-resilient control, and constrained nonlinear systems control, with applications to intelligent vehicles and multiagent systems.