Coordinated Control of Networked Multiagent Systems via Distributed Cloud Computing Using Multistep State Predictors

This article studies the coordinated control problem of networked multiagent systems via distributed cloud computing. A distributed cloud predictive control scheme is proposed to achieve desired coordination control performance and compensate actively for communication delays between the cloud computing nodes and between the agents. This scheme includes the design of a multistep state predictor and optimization of control coordination. The multistep state predictor provides a novel way of predicting future immeasurable states of agents in a large horizontal length. The optimization of control coordination minimizes the distributed cost functions which are presented to measure the coordination between the agents so that the optimal design of the coordination controllers is simple with little computational increase for large-scale-networked multiagent systems. Further analysis derives the conditions of simultaneous stability and consensus of the closed-loop-networked multiagent systems using the distributed cloud predictive control scheme. The effectiveness of the proposed scheme is illustrated by an example.