Optimal Fixed-Time Control for Human-in-the-Loop Multiagent Systems With Actuator Faults.

Area of Science:

• Control Systems Engineering
• Artificial Intelligence
• Robotics

Background:

• Multiagent systems (MASs) require robust control strategies for complex tasks.
• Human-in-the-loop (HiTL) systems enhance adaptability but introduce control challenges.
• Actuator faults can degrade system performance and stability in MASs.

Purpose of the Study:

• To develop an optimal fixed-time tracking control scheme for HiTL MASs.
• To address actuator faults within the control framework.
• To ensure bounded system signals and efficient tracking in finite time.

Main Methods:

• An HiTL optimal control protocol based on simplified optimal control.
• Embedding synchronization error and leader input into the cost function.
• A novel reinforcement learning (RL) algorithm with exponential terms for fixed-time convergence.

Main Results:

• The proposed control scheme achieves optimal tracking control under HiTL supervision.
• The RL algorithm ensures fixed-time convergence of system signals.
• Actuator fault compensation is integrated, maintaining system performance.
• All closed-loop system signals are proven to be bounded within a fixed time.

Conclusions:

• The developed optimal fixed-time tracking control is effective for HiTL MASs with actuator faults.
• The reinforcement learning approach enhances convergence speed and energy efficiency.
• The control scheme guarantees robust and bounded system behavior in finite time.