Adaptive Dynamic Programming-Based Event-Triggered Robust Control for Multiplayer Nonzero-Sum Games With Unknown...

Area of Science:

• Control Theory
• Nonlinear Systems
• Adaptive Dynamic Programming

Background:

• Investigating robust control for unknown multiplayer nonlinear systems with input constraints and uncertainties is challenging.
• Existing methods often require full knowledge of system dynamics or struggle with complex interactions.

Purpose of the Study:

• To develop an event-triggered robust control strategy for unknown multiplayer nonlinear systems.
• To address constrained inputs and system uncertainties using adaptive dynamic programming.
• To ensure uniform ultimate boundedness of the system states.

Main Methods:

• A neural network-based identifier is used to estimate system dynamics from input-output data.
• A nonquadratic value function is designed to convert the stabilization problem into an optimal control problem.
• A critic network with a novel weight updating law and experience replay is employed to solve the Hamilton-Jacobi equation.

Main Results:

• The proposed event-triggered control method effectively stabilizes unknown multiplayer nonlinear systems.
• The approach relaxes the need for system dynamics identification and persistence of excitation conditions.
• Simulation examples demonstrate the practical effectiveness and robustness of the control strategy.

Conclusions:

• The developed event-triggered robust control method offers a viable solution for complex multiplayer nonlinear systems.
• The adaptive dynamic programming approach with neural networks provides a powerful framework for handling uncertainties and constraints.
• The method ensures system stability and boundedness, validated through simulation studies.