ABSTRACT
This chapter addresses the challenge by developing a real-time hierarchical control system called Large-scale on-time collaborative heterogeneous Networked embedded systems (LochNess). LochNess decouples the estimation of evader states from the control of pursuers via multiple layers of data fusion. Multiagent systems comprise an important subclass of cyberphysical systems that involve communication and collaboration between interacting agents that locally exchange information. The chapter focuses on the solid foundation of adaptive control theory and multiagent systems theory in order to develop a new distributed adaptive control architecture that can foil malicious sensor and actuator attacks in the face of exogenous stochastic disturbances. The system-state information of different agents is exchanged through communication channels represented by a given graph communication topology, and local actuators of each agent utilize the information received from its neighbours for the control design protocol.
