Event-triggered iterative learning control for output constrained multi-agent systems.

Area of Science:

• Control Systems Engineering
• Robotics and Automation
• Networked Systems

Background:

• Multi-agent systems require robust control strategies for coordinated tasks.
• Output constraints and nonlinear dynamics pose significant challenges in discrete-time systems.
• Event-triggered communication can reduce network load but requires careful design.

Purpose of the Study:

• To develop an event-triggered iterative learning consensus tracking control strategy.
• To address output constraints in nonlinear discrete-time multi-agent systems.
• To reduce communication frequency while maintaining tracking performance.

Main Methods:

• An estimated Pseudo Partial Derivative (PPD) algorithm was developed for system dynamics.
• An output observer was designed using the estimated PPD.
• A deadband controller and event trigger condition were implemented based on output estimation error.
• An event-triggered iterative learning control algorithm was constructed and its convergence analyzed using Lyapunov functions.

Main Results:

• The proposed strategy enables consensus tracking for output-constrained nonlinear discrete-time multi-agent systems.
• The event-triggered approach significantly reduces communication requirements by activating only when necessary.
• The control algorithm ensures consistent and complete tracking of the desired trajectory.
• Simulation results validated the effectiveness of the developed control protocol.

Conclusions:

• The event-triggered iterative learning consensus tracking control strategy is effective for nonlinear discrete-time multi-agent systems with output constraints.
• The proposed method achieves high tracking performance with reduced communication overhead.
• The strategy offers a practical solution for improving the efficiency of multi-agent systems.