Distributed Event-Triggered Control with Local Solution-Predictor Curves for Linear Multiagent Systems

For distributed control of linear multiagent systems, we propose an event-triggering rule having a user-defined exponentially decaying dynamic threshold for reducing local information exchange between agents. The key novelty of the proposed approach is that information exchange between agents are predicated on local solution-predictor curves when an event occurs. In contrast to standard event-triggering rules relying on sampled data exchange, these curves allow for a significant reduction in network utilization. A system-theoretical stability analysis is further provided for the proposed approach and an illustrative numerical example is included to show its efficacy.