Event-based distributed cooperative neural learning control for nonlinear multiagent systems with time-varying output...

Area of Science:

• Control Systems Engineering
• Artificial Intelligence
• Networked Systems

Background:

• Practical engineering systems often face performance degradation due to security constraints and network burdens from inter-agent communication.
• Existing control methods struggle with time-varying output constraints and communication inefficiencies in nonlinear multiagent systems.

Purpose of the Study:

• To develop a distributed cooperative learning control strategy for nonlinear multiagent systems with time-varying output constraints.
• To enhance system security and reduce communication resource usage through adaptive control mechanisms.

Main Methods:

• An improved output-dependent universal barrier function was designed to handle adjustable symmetric or asymmetric output constraints.
• A switching event-triggered condition based on neural network (NN) weights was developed for adaptive frequency adjustment.
• The Padé approximation technique was utilized to manage input delays and simplify controller design.
• Lyapunov stability theory was applied to prove system convergence and boundedness.

Main Results:

• The proposed control strategy ensures follower outputs converge near the leader's output while respecting output constraints.
• All signals within the closed-loop system were proven to remain ultimately bounded.
• The event-triggered condition adaptively adjusts NN weight updates, optimizing communication resource utilization.

Conclusions:

• The presented approach effectively addresses nonlinear multiagent systems with time-varying output constraints and communication burdens.
• The developed control strategy enhances system security and stability, validated by simulation results.