Our inventors have created a novel rule for the distributed control of multiagent systems to reduce agent-to-agent information exchange. This improved method uses an adaptive event-triggering rule to replace continuous information exchanges with individual solution-predictor curves, allowing agents to approximate their own solution trajectories. Agent-to-agent information exchange is pared down further by the removal of standard norms; this helps filter events and prioritizes necessary information exchanges. Exchanging the standard sampled data exchange method for these solution-predictor curves will improve communication scalability, increase information security, and accommodate larger numbers of agents within an individual system. Applications of this technology include emergency/disaster response, military reconnaissance, task assignment, and traffic management.
An example multiagent system consisting of four agents, where agent-to-agent information exchange is predicated on event-triggering scenarios. (Here, the first agent stands for the leader agent and other agents stand for the follower agents.)