Multi-Rate Design Of Networked Predictive Controller And Its Application For Consensus Of Networked Multi-Agent Systems With Delays

In this paper, a direct design approach of networked predictive controller (NPC) based on multi-rate sampling for networked control systems with delays is proposed, and a novel consensus protocol that uses multi-rate forward-channel NPC is provided, which actively compensates the delays to synchronize the states of the networked agents. The multi-rate controller costs less computation in the design, and the consensus protocol shows more suitable for distributed applications. Necessary and sufficient conditions for stabilization by the multi-rate predictive controller are derived, and requirements for system synchronization of the agent network are acquired. Numerical examples are provided to illustrate the effectiveness of the theoretic results.