Cooperative Fault-Tolerant Output Regulation for Multiagent Systems by Distributed Learning Control Approach

Area of Science:

• Control Systems Engineering
• Distributed Systems
• Fault Tolerance

Background:

• Linear multi-agent systems are susceptible to actuator faults, compromising cooperative control.
• Existing methods often require knowledge of the exosystem's system matrix, limiting applicability.

Purpose of the Study:

• To develop a distributed learning control approach for cooperative fault-tolerant output regulation in linear multi-agent systems.
• To address the challenge of unknown exosystem system matrices in fault-tolerant control.

Main Methods:

• A distributed estimation algorithm with an online learning mechanism to identify the exosystem's system matrix.
• Introduction of an auxiliary variable to construct a data matrix for learning the system matrix.
• A novel distributed fault-tolerant controller design based on the developed estimator.

Main Results:

• Successfully identified the unknown system matrix of the exosystem for each subsystem.
• Enabled reconstruction of the exosystem state across all subsystems.
• Demonstrated effective fault-tolerant control and output regulation through simulations.

Conclusions:

• The proposed distributed learning control approach effectively handles cooperative fault-tolerant output regulation for linear multi-agent systems with actuator faults.
• The method's ability to learn unknown system matrices offers a significant advancement over existing techniques.