Dynamic Event-Triggered Distributed Optimization Control for IoT-Enabled Euler–Lagrange Multiagent Systems
Abstract
This article investigates a communication-efficient distributed optimization control problem for a class of Internet of Things (IoT)-enabled Euler–Lagrange (EL) multiagent systems under resource-constrained communication links. In contrast to existing results that rely on continuous interagent information exchange, a nested inner–outer loop distributed control framework based on a dynamic event-triggered communication mechanism is developed, where each agent transmits its sampled generalized position only when the prescribed triggering condition is satisfied. Since the event-triggered mechanism (ETM) generates discontinuous sampled-and-held signals, a distributed smoothing filter is introduced to generate continuously differentiable reference signals for the EL inner loop. Furthermore, the proposed distributed control strategy requires only each agent’s locally measured objective gradient and the latest sampled-and-held generalized positions received from its neighbors. Under explicitly stated design conditions, all closed-loop signals are shown to be bounded, the agent positions exponentially converge to the global optimizer, and Zeno behavior is excluded. Finally, simulation results demonstrate the effectiveness of the proposed method in achieving distributed optimization while substantially reducing packet transmissions over communication-constrained networks.