Output Consensus of General Linear Networked Multiagent Systems With Unknown Disturbances via Cloud Computing

Abstract

For a general linear multiagent system (MAS) with multigroup suffering from disturbances, a cloud-based output consensus control scheme, including a cloud predictive control protocol and a disturbance observer, is proposed to achieve output consensus via cloud computing. In a cloud framework, agents in each group only send their own information to a corresponding cloud node, and cloud nodes transmit next control command for agents. Accordingly, the communication burden of an individual agent is reduced in each group. Moreover, the communication burden of each agent remains constant despite the scale of the network, and energy consumption is also reduced effectively from the perspective of one agent. In a cloud, delayed states of agents are reconstructed via delayed outputs received from a sensor, and a predictor is designed based on observed delay states. A cloud predictive control protocol is developed by dynamic variables and the predicted output to compensate for network delays from a cloud to an actuator, and a disturbance observer is presented based on predicted and local outputs. In theory, sufficient conditions for stability are presented and analyzed, and it is proved that the proposed cloud-based control scheme renders the MAS stable and to achieve output consensus. Finally, two simulations are presented to verify the proposed control scheme.